Knowledge Management Series - Introduction to Viva Topics

A few weeks back, I was lucky enough to have been given the opportunity to try out the Lenovo ThinkSmart View, a standalone device dedicated to Microsoft Teams. Many of you may be thinking "erm, why would you want that?" and the simple answer (which I'll expand on in a minute) is performance. Too many times has Teams rendered my machine unusable and I spend a large chunk of my day in meetings. Would this improve my experience?

In this post, I'll cover what the device is, why you might need it, a run through of the set up, initial thoughts, longer term and then a wrap up.

## What is the Lenovo ThinkSmart View

Let's start with what the device is. It's an Android powered device with an 8" LCD, with a camera, speaker and microphone that allows you to offload the running of Teams to a separate device. It has a separate power supply and can run entirely independently of your PC but can also synch up so that you can still share screen on your PC even though the meeting is running on your device.

Overall, it is a great looking piece of kit that feels solid with a silky smooth surface and very stable when in either landscape or portrait mode. It includes a manual camera shutter so you can ensure that you are not being watched. For me, the speakerphone capability was hugely useful and like having a conference phone on your desk but you can also connect with bluetooth headsets if you are in a more exposed area like an open plan office.

You can see full details on the [Lenovo site](https://www.lenovo.com/gb/en/smart-devices/smart-office/thinksmart/Lenovo-CD-18781/p/ZZISZSCCD03).

## Why would you need it?

For me, the reason I was looking for a device like this was that my Surface Pro 6 often slows down massively when on Teams calls. I will admit that I am a tab hoarder and often have multiple profiles open in the new Edge but the amount of memory that Teams takes up for a call with a number of people on is huge. Too often, I was unable to do things on my PC during a meeting, hugely limiting my productivity.

I also like to have Teams front and centre and having a dedicated device showing a summary of my day and with easy access to chat was something I am keen on. Having a single screen showing your upcoming events while you are working on a document and comparing with a webpage means that you do not miss things coming up, especially if you are focusing and don't want notifications.

Finally, it is a good quality speakerphone meaning that you don't have to be wearing headphones in meetings all day. Being constricted for long hours really adds to the drain of remote working constantly.

## What was the setup experience like?

I will be honest that my initial impressions were hugely underwhelming. It only had meetings and calls available and felt very limited. This was surprising and not what I had expected from one of the Teams devices.

Thankfully, after a little digging, I found [this article](https://techcommunity.microsoft.com/t5/microsoft-teams-blog/introducing-microsoft-teams-displays/ba-p/1505437) on Teams Devices and confirmed that it had not been set up as a Teams Device. I logged it into my tenant and then went into the Teams Admin area - if you don't have access to the Teams Admin area, make sure you know someone who is as this is essential to support the device.

From Teams Admin, I could now upgrade to the latest version and also set it as a Teams Device. It took a few tries at doing this and I found it worked best by making it a scheduled update but later that evening, I finally had the device that I was expecting.

## Initial thoughts

The interface feels very slick and navigating takes place quickly. As well as having meetings and making calls, you have the ability to access the Teams Hub and can chat with people too. Chat can either be done on the touch screen or you can connect up to your PC and so use your PC keyboard. It's simple to understand and move between the different features and feels natural doing it.

An interesting area is how it interacts with your PC as well. When sharing screen, you can still connect with the PC client and share your screen from there but as it is not running the video, it uses less RAM there (one of the big reasons I wanted a separate device). It connected simply and worked well, giving that seamless experience.

What about the meeting sound and vision? The quality of the camera is very average and the angle of the device means that you are going to be getting that view of any chins that you have been hoarding during lockdown which isn't ideal. From the perspective of the person with the device, the audio is just right as a speakerphone but I did hear some comments about the volume of typing on the keyboard so you may need to make liberal use of the mute. The background blur also causes problems and stops the camera working at all in many cases so I decided against using that.

The biggest missing feature I found was tasks as would love to be able to pin my ToDos on the screen as a reminder of what to do that day alongside the calendar itself. Hopefully something that will come soon.

## Longer term

I've now been using the device for a couple of months with the comments above being for the first few days of usage. What are my thoughts now?

To be honest, I've started using the device less and less. It's nice to have there for my quick view of upcoming meetings but the quality of the video and audio became more of an issue. I like to be looking at people for meetings and having that separate device when I had content on screen gave off the view that I wasn't watching too much. The smaller screen was also hard to look at other people's content and I found myself having to move closer to see, giving people a lovely view of my forehead and not much else.

The link with the PC that I lauded at first soon became less stable and would often not be there, leading me to have to join the meeting twice. This would use just as much RAM on the PC and completely removed the point I was trying to overcome. I would also run Teams as typing on the device was so slow and clunkier than on a phone due to where it was and the angle. Which brings me to the one point that could start me using the device again more.

At the time of writing, the Cortana integration is only available in the US as a preview. To have a standalone device that I could ask when my next meeting was or what tasks I had left to do that day would definitely interest me. Would it work well? I don't know as I couldn't find anyway to enable it at all here in the UK. Hopefully this preview will reach a bit further and open up more of that voice control which is great in a home office with no-one to disturb. We'll see.

## Summary

So, the big question - would I recommend this device?

This is a really tricky one. It's great looking and I enjoyed using it - having that extra screen to see your chats, upcoming meetings and unread Teams messages is nice. Being able to not to have to use headphones and still get a clear sound was great as I have definitely been feeling distant from having headphones on all day long.

However, it's an expensive device just for these things. What would make it worthwhile is to be able to offload the full use of Teams from the PC and reduce that memory utilisation. For me, it just doesn't quite do this. The video quality isn't good enough and I don't want that up angle nose shot too much. The sharing of the desktop PC screen is too inconsistent.

The use of Cortana may tip things back into the firm recommend option but at the moment, I can't quite justify the cost for what it offers to others. I will be keeping it on my desk for now and keep an eye out for those Cortana updates.

Photo courtesy of me!


Do you need a standalone device for Teams? - Evaluating the Lenovo ThinkSmart View

I hit a strange issue this week but today had that massive buzz of solving it. Sometimes it is the simple things that you've missed but other times (like this one), it is something very odd. The problem was that I had a set of around 1000 migrated sites that had come from SharePoint 2010 and should have been set up as Modern sites. They appeared to be but the SharePoint Groups had retained the old names, i.e.

- Site Owners
- Site Members
- Site Visitors

I was expecting the groups to have the name of the site in there, i.e. if my site was called "My New Site" then the groups would be:

- My New Site Owners
- My New Site Members
- My New Site Visitors

As I was investigating, I noticed that some of the sites had started to get the new names. At this point, I thought I was going mad but as I had taken notes, I knew that something had changed and it wasn't just my goldfish level memory. I eventually realised that if I browsed to Site Permissions from a Modern Home page then something happened along the way that triggered the correction.

My next step was to check PnP PowerShell to see if the sites needed [Groupifying](https://docs.microsoft.com/en-us/sharepoint/dev/features/groupify/groupify-overview) but running the command [Add-PnPMicrosoft365GroupToSite](https://docs.microsoft.com/en-us/powershell/module/sharepoint-pnp/add-pnpmicrosoft365grouptosite?view=sharepoint-ps) told me that they were already associated with a Group.

Finally, I got developer-y and opened up the Chredge Dev Console to look at which REST calls the page made when it hit the Site Permissions page. This is when I discovered a POST call to /\_api/GroupService/SyncGroupProperties and this sounded promising. I fired up [Invoke-PnPSPRestMethod](https://docs.microsoft.com/en-us/powershell/module/sharepoint-pnp/invoke-pnpsprestmethod?view=sharepoint-ps) and made a call to a site that happened been manually fixed yet and by magic, it worked! This certainly saved a very dull day of manually clicking several hundred site permission links.

I should say that I cannot find this API documented anywhere so this does not appear to be a fully supported API but if this post saves someone a day or several days of clicking on sites then I'll be a happy person.

Photo courtesy of [Toa Heftiba](https://unsplash.com/@heftiba) via [Unsplash](https://unsplash.com)


Issue with Microsoft Groups not syncing for SharePoint Site

Last week, myself and my fellow co-hosts of GreyHatBeard were joined by Luise Freese, Vesa Nopanen and Karoliina Kettukari at Collab Days Munich/Vienna as we shared our thoughts on the etiquette of Microsoft 365. It was a fun format with plenty of chance to laugh at the same time as attendees picking up some great ideas as we ran through three questions that each person had 90 seconds to answer.

- What breach of good etiquette most frustrates you in Microsoft 365?
- What do you wish Microsoft 365 would do to encourage good etiquette?
- What is your top tip for tip top Microsoft 365 etiquette?

So what did our etiquette experts recommend?

## What breach of good etiquette most frustrates you in Microsoft 365?

### Al

People booking meetings without checking the scheduling assistant to check if I have time available for the meeting. See [this video](https://support.microsoft.com/en-us/office/video-use-the-scheduling-assistant-a7bf1aee-bee6-46d6-a126-194ed04fbe09) for an example on how to do it.

### Luise

Saying "hello" and then writing nothing but forcing me to ask "whats up". You disrupt people from what they are doing and force them to wait. Instead say hello and what you are after in one post. For a great link explaining this, see [https://sbmueller.github.io/nohello/.

Also, when people keep with old working habits in new tools: cascading folders, signatures in chats, adding people to a chat to cover your ass. There's no need!

### Vesa

Not showing up on time for Teams meetings or having to leave in the middle of the meeting without telling the schedule issues in advance. The worst is when people show up late and then ask for the full review of things missed, shouting out a few quick opinions and then leaving.
Also, when using Microsoft 365, don't make all content personal. Instead of files on OneDrive, make use of SharePoint. Instead of personal flows and personal Power Apps, make them team flows and shared apps. Start as you mean to go on, all the way from draft.

### Karoliina

When people are complaining about notifications, teams and channels but not caring to go through the settings to configure them. There are settings that can be configured with a bit of time.

In fact, even further, complaining about everything but not caring to learn how to use the tool properly. There are so many ways to learn how to do things that you should investigate.

### Garry

Not using subject lines for new conversations. These help introduce a subject and give people a summary.

Also, ignoring presence status - do not disturb is not I'm available and those chats coming in will break someone's focus.

## What do you wish Microsoft 365 would do to encourage good etiquette?

### Vesa

Analyzing the agenda and not letting someone send the invitation until agenda is properly filled if there are more than 10 people invited.

Marking a meeting "cameras will be on" and turning off such a meeting cameras on automatically.

Reminding people putting documents to silos/hidden areas such as OneDrive that "this seems to be a document that would be better to be saved to a shared location".

Bring back Clippy as a Microsoft 365 version and annoy people until they comply!

### Al

When sharing a screen on Teams put a message up with the name of a person who has put their hand up and allowing me to click it so their hand is lowered.

Put the names of people with hands up in order so I know who put their hand up first and can allow them to ask their question first.

### Garry

Allow setting of default messages for presence statuses. For example, you could suggest alternative contact methods.

Letting people automatically know when you are next free based on your calendar, e.g. "Garry is busy right now but will become available at 14:00?"

### Luise

Mute – raise hand – unmute. Please, dear Microsoft Teams, help people to lower their hands when they have made their point. EVERYONE forgets to lower their hands!

### Karoliina

Smart tooltips when writing a message or making a calendar invite (like you get in Outlook).

Suggestions to make a team instead or to discuss in an existing team when you create a groupchat with the same people. Also suggesting NOT to mention a whole team or a channel.

## What is your top tip for tip top Microsoft 365 etiquette?

### Karoliina

Teams is for teamwork & sharing is caring & all that glitters are unicorns.

Learn to use the tools together with your team as working together will help that combined understanding.

### Al

Use scheduling assistant to ensure people are free and if you cannot find a time when everyone is free, use FindTime to offer options to all attendees.

Make sure that you respond to meeting invites and add messages to responses. Just leaving things hanging leaves people unclear on what is going to happen.

### Vesa

Respect others – don't take over the meeting. Keep your mic off unless you speak. Keep the camera on and pay attention to the meeting (don't multitask – I am guilty on this one).
In Microsoft 365, work together, put files and apps already in version 0.1 to a shared location. This can make it clear that it is in draft but is available for viewing straight away.

### Luise

Don't let people jump from meeting to meeting. Schedule meetings that everyone can take a break, especially in this current world where everyone is at home.

### Garry

Use Subject lines in conversations, you can search messages by subject line in Teams and it makes them more discoverable.

Don't just throw documents into Files and expect people to know that they are there. Follow up your upload with a post in the chat area to let people know that are there. For imporant documents start a new conversation with subject line and shared file, this helps discoverability of the document and also a good starting point for document conversations.

Share the way you work in email footers and your status message and encourage others to do the same to set expectations for how they expect you to engage.

## Wrap-up

We all had a great time during the session, despite my buzzer adding that frisson to the atmosphere. Vesa and Luise won through to the final where they had to defend the indefensible, showing their ability to justify things that probably shouldn't be.

- Luise - What we need is a new service for tracking tasks in Microsoft 365
- Vesa - Agendas in meeting invites are a waste of time

Both were very articulate and had me persuaded that they were great ideas (outrageous!) but Vesa just managed to take the win by mentioning whiskey!

I would like to give my heart felt thanks to [Luise](https://twitter.com/LuiseFreese), [Vesa](https://twitter.com/vesanopanen), [Karoliina](https://twitter.com/kettukari), [Garry](https://twitter.com/garrytrinder) and [Al](https://twitter.com/al_eardley) for all their great tips and for being so entertaining with them. If you would like to hear more of these events, let user group and conference organisers know and we'll see what we can do!

Photo courtesy of [My Life Through A Lens](https://unsplash.com/@bamagal) via [Unsplash](https://unsplash.com)


Sharing the etiquette of Microsoft 365

Microsoft has been speaking for some time about the power of Microsoft Search and the ability to have one place to search for all your content. This has gradually been coming more and more true but one of the areas that has not been widely available has been the ability to add your own custom content to the index. However, that is now possible and allows developers to bring content into search alongside other content. In this post, I will run through two ways of doing this but both will require development experience, including a sample from the graph team and then my own example using Logic Apps to consume Twitter content.

## Overview of custom connectors

Microsoft Search has the ability to connect to standard systems through the Microsoft 365 Admin Centre for some standard systems including SQL Server, Azure Dev Ops, MediaWiki and the others mentioned in the image below. These allow users to follow a wizard guide to enable content to be added to the index so where you have data in SQL such as a table of custom application data, this can be added to your Microsoft Search.

However, if you have a third party application or content in something like Cosmos DB, you need to use the interfaces in Microsoft Graph to add content to the index. This broadens the capabilities to connect without having to first move data to one of the systems below.

![Current Search Connectors](/images/2020/09/DefaultSearchConnectors.jpg)

## Adding custom data with .Net Core

The easiest point to get started with the custom search connectors is to use the [sample code](https://github.com/microsoftgraph/msgraph-search-connector-sample) from the Microsoft Graph team. This has a great guide in the readme to get started using a console app to connect. The process that it follows is:

- Set up the app registration in Azure AD
- Grant permissions for External Items to have Read/Write
- Create an External Connection giving a unique ID, name and description
- Create a schema for the index content and assign to the connection
- Add items to the index
- Confirm connection set up in Microsoft 365 Admin Centre
- Create a Vertical at the tenant or site level
- Create a Result Source for the layout of the results (using https://searchlayoutdesigner.azurewebsites.net/)

Once these are set up (following the great [readme](https://github.com/microsoftgraph/msgraph-search-connector-sample/blob/master/README.md) in the sample), the results show up in the main search results as a vertical below where the sample Parts vertical is set up. The console app sets up each of the steps above through number prompts for the user but is designed as an example not as a real solution to use. Which brings me to using Logic Apps.

![Parts Search Results](/images/2020/09/PartsSearchResults.png)

## Pulling index results using Logic Apps

Taking the ideas in the repo above to a more usable setting, I decided that pulling a set of tweets from Twitter would be useful to be searchable across Microsoft 365, using the great information that gets shared regularly on the TweetJam's run by [Christian Buckley](https://twitter.com/@buckleyplanet) under the [#CollabTalk](https://twitter.com/search?q=%23collabtalk&src=typed_query) hashtag. There are some great conversations that take place and useful information shared and making this available within the help guides could be useful.

The use of Logic Apps meant control over how the ingestion of the indexes can be triggered. Rather than the historic way of running nightly imports, the trigger can happen straight away and have the content made available. They also provide an easy way of viewing the history of imports both through the run history and through easy integration with App Insights.

So what Logic Apps are created? All the Logic Apps are available in the [Mcd79TwitterSearchConnector repo](https://github.com/kevmcdonk/Mcd79TwitterSearchConnector).

- [Setup](https://github.com/kevmcdonk/Mcd79TwitterSearchConnector/blob/main/Mcd79TwitterSearchConnectorSetup.json) - configures the connection and required schema, more on this below.
- [Search](https://github.com/kevmcdonk/Mcd79TwitterSearchConnector/blob/main/Mcd79TwitterSearchCollabTalk.json) - Retrieves up to 99 tweets to pull a range of information.
- [Trigger](https://github.com/kevmcdonk/Mcd79TwitterSearchConnector/blob/main/Mcd79Twitter-CollabTalk.json) - Regular trigger for any new tweets with the hashtag.

The first Logic App creates a connection with unique ID, name and description to identify the index items. It checks first to see if it exists and adds it if not. The same then follows for the schema which outlines the model for the index and defines which properties can be held. This is a relatively simple schema with datatypes but no collections or sub-schemas available.

![Setup Logic App](/images/2020/09/SetupConnectorAndSchema.jpg)

The schema itself defines the properties to be captured with the name and type as well as four other values from the [schema definition](https://docs.microsoft.com/en-us/graph/api/resources/schema?view=graph-rest-beta):

- isQueryable - whether the property can be used in search queries using KQL
- isRefinable - whether the property can be used to refine the query with filters
- isRetrievable - whether the property can be returned in the search results
- isSearchable - whether the property is searchable

The schema that I created for the Tweets can be seen in [the repo](https://github.com/kevmcdonk/Mcd79TwitterSearchConnector/blob/main/schema.json).

In both the search and trigger Logic Apps, the key element is the part that creates the index. This is done by making a PUT request against the connection and using a unique ID for each index. In the tweets case, I am using the Tweet ID. The body of the PUT request includes four key areas:

- Type - the type of object, usually microsoft.graph.externalItem
- [ACL](https://docs.microsoft.com/en-us/graph/api/resources/acl?view=graph-rest-beta) - the security required on the index item which can be a user, group, everyone or everyoneExceptGuests.
- [Content](https://docs.microsoft.com/en-us/graph/api/resources/externalitemcontent?view=graph-rest-beta) - the text content of the index, in the tweet it is the tweet contents.
- [Properties](https://docs.microsoft.com/en-us/graph/api/resources/externalitem?view=graph-rest-beta) - JSON object that must meet the schema defined

You can create up to 2,000,000 indexes in your tenant across both the custom connector and the standard connectors. Once you have added the index, you can follow the same guide as in the search sample [readme](https://github.com/microsoftgraph/msgraph-search-connector-sample/blob/master/README.md) to set up the Verticals and the Result Source to show how it will be rendered.

Once done, you can see the results appear in your search results as below.

![Tweet Search Results](/images/2020/09/TweetSearchResults.jpg)

So there you have it, the content from CollabTalk appearing right there in your search results. Obviously, I have picked this as an example but that content could be from your custom applications, Salesforce and anyone one of the many other Connectors that exists in Logic Apps. The only struggle I had was about getting the Graph Auth in place (I originally tried using Managed Identity as in [Garry Trinder's great post](https://garrytrinder.github.io/2020/09/implement-microsoft-graph-app-only-calls-the-easy-way-using-azure-logic-apps-and-azure-managed-identity) on using it with Logic Apps but found issues that I'll look back at) but otherwise, it was set up in an hour.

While this is a developer task and not through a wizard, it is an easy way to extend your search results and get access to data where you need it and render how it appears.

Photo courtesy of [John Barkiple](https://unsplash.com/@barkiple) via [Unsplash](https://unsplash.com)


Creating a custom Microsoft Search connector with Logic Apps

With Microsoft Ignite 2020 (Sep 22-24) came more Knowledge Management news and the biggest one of these (at least as far as most people were concerned) was the first widely available service from Project Cortex. Rather than release one big service to organisations, Microsoft has listened to feedback and will be releasing different services under the Project Cortex banner and from October 1, you will be able to add SharePoint Syntex to your E3 and E5 licenses - but what is it?

> This is part five of the Knowledge Management series. For other parts, see links below:

> - [Part One - Introduction](/2020/08/12/knowledge-management-series)
> - [Part Two - Considering where knowledge is stored](/2020/09/10/knowledge-management-series-considering-where-knowledge-is-stored)
> - [Part Three - Organising knowledge](/2020/09/14/knowledge-management-organising-knowledge)
> - [Part Four - Creating a culture of knowledge](/2020/09/17/knowledge-management-series-creating-a-culture-of-knowledge)

## What is SharePoint Syntex?

After the hints and suggestions around [Project Cortex](https://aka.ms/ProjectCortex), at [Ignite](https://myignite.microsoft.com) there was details on what people would actually be able to do. [Jeff Teper](https://www.twitter.com/jeffteper) announced that it wouldn't just be about the single service but people would be able to pick the services that they wanted and the first available would be SharePoint Syntex, from October 1. This first service would be focused on automating the process of getting your metadata out of your SharePoint content using [Machine Teaching](https://blogs.microsoft.com/ai/machine-teaching/). This is slightly difference to machine learning in that you can gather intelligence using a smaller set than you would use in machine learning and use experts to provide guidance, reducing the size of initial data set that would be required.

Syntex is suggested as being used in three areas:

- Content Understanding - extracting metadata from documents and setting classification automatically
- Content Processing - categorise content and build in to automated processes
- Content Governance - manage to content to apply appropriate security and governance

Using a set of models, you can teach Syntex how to process different types of documents to automatically extract the metadata in three key ways (more on these later in this post):

- Document understanding - recognise patterns in documents such as resumes, statements of work and contracts to extract key info
- Form processing - using AI Builder, extracting information from images and other document forms that are structured or semi-structured such as printed contracts or feedback forms
- Digital image processing - automatic tagging of images by recognising the content and extracting handwritten content from images

Syntex also appears to be wrapping in some of the Microsoft Search capabilities with Graph Connectors so that you can pull content outside of SharePoint and OneDrive into this as well. For larger organisations who have invested in other ways to store their knowledge such as ServiceNow or even MediaWiki. From what I can tell at this stage, this means that you don't need to first mgirate your content into SharePoint/OneDrive to get your metadata, you can also connect it at source which fits with my earlier post of finding [where knowledge is stored](/2020-09-10-Knowledge-Management-Series-Considering-where-knowledge-is-stored.html) and [organising that knowledge](/2020-09-14-Knowledge-Management-Series-Organising-knowledge.html) without having to move it.

The final area is for me one of the most interesting and that is the extension to the recent Content Services in SharePoint that will allow administrators to see how the metadata is being used. This has been an area lacking in SharePoint for a long time with Administrators spending plenty of time structuring their metadata but having no idea if it was being used in the right way without using a third party tool or some heavy scripting. Making this available and consumable will open up a far more analytically driven management of the metadata and understanding the depth to which people are assigning content. For example, do people tag content with HR or Human Resources as a department, do they just tag everything as Public rather than Internal or Confidential, etc. It will also look at widely used Enterprise Keywords (an open set of metadata with little governance) to see commonly used terms that probably should move to a more governed set of metadata. This will also include management of Content Types to Hub Sites to help push out standard content types more easily - I truly hope that this works better than the current Content Type Hub!

## Document Understanding

There are several third party tools that have added capabilities for tagging metadata available but this is the first from Microsoft. Some of those tools have used Azure and the AI capabilities in Cognitive Services but not Microsoft has wrapped these up into their own tools. The Document Understanding allows licensed users to take a sample set of documents and process them by:

- Identifying a Content Type to hold the metadata
- Teach the model what should match this Content Type using at least five different documents
- Train the model to extract entities using phrases, positions and others
- Match the model to labels
- Assign to Content Type Hubs

You can watch a great [Click through demo](https://aka.ms/SharePointSyntex/demo) of this which gives an idea of how it works. I was lucky enough to see an early preview of this late last year and at the time, felt that it was a little hard to just get going with it - it felt like it would involve a lot of training to get started. However, Microsoft have really learnt from the early previews and this demo shows that it is easy to pick up common patterns. I look forward to being able to trying this out with some real content like our requirements docs and statements of work that can make it easier to match the content. Imagine being able to load all your project definitions into one place and then have the document be tagged with the Project Name, department and key stakeholders so that it shows up in the project site, department site and makes it easy for stakeholders to see all their definitions. It could even track common themes in the project so you could see all Finance projects, C# projects, customer projects, etc in one place. The models created can be re-used as well and published to other libraries as well so they can be duplicated across areas for common patterns.

![Syntex Extractor Example](020/09/SyntexExtractorExample.jpg)

## Form Processing

For those who follow the Power Platform, you will likely already know about AI Builder and what could be done to process images of forms and cards to extract entities automatically. I have even used this with my son in a Power App to take photos of Pokemon cards and extract all the stats for each. This logic could be used to process tickets or physical ID cards quickly and easily. This has been integrated into SharePoint for a while but had multiple hoops with licensing and connecting to metadata. With Syntex, this can now all be done for you in one go. It will be interesting to see how much this gets above the standard AI Builder over time.

![Form Processor Example](020/09/FormProcessorExample.jpg)

## Digital Image Processing

Similar to Form Processing, this seems to be similar to functionality to what is already available. Text and handwriting are extracted from image files and indexed already and you can get a recommended alt text for uploaded photos. Syntex takes this further in that you can set your own fields with these values and get more control over how it is used.

## What is new?

There were certainly some people who were disappointed with Syntex, hoping for more to be announced around the knowledge hubs and new information as well. I personally think we'll see those pretty soon and certainly before the end of the year but let me quickly what I cover exactly what is new:

- Document understanding really builds on the Azure Cognitive Services to make it easy to apply models - yes, you could develop your own in Azure but this makes it easy for end users to understand and use.
- Compliance and security are closely linked in to all of this so you can add more intelligence to your labelling processes.
- Bringing it all together into one place and integrating the services, making it easier to get less technical users to make the most of the services available.
- Analytics on how the content is actually being used, allowing people to act on real data not just hunches.

You can read more about the details in the [announcement blog](https://techcommunity.microsoft.com/t5/project-cortex-blog/announcing-sharepoint-syntex/ba-p/1681139) on SharePoint Syntex.

Will these new capabilities and ease of use for existing services be worthwhile? Much will depend on the final details on the licensing to be announced on October 1. Microsoft have said that Syntex will be an add-on to [E3/E5 licenses](https://www.microsoft.com/en-us/microsoft-365/compare-microsoft-365-enterprise-plans) but I have not yet seen how many license you will need, i.e. will it require everyone in an organisation to be licensed or just those creating models? October will reveal all but I would hope that it will just be a smaller subset of power users who need to be licensed. How much? [ZDNet think](https://www.zdnet.com/article/microsoft-launches-sharepoint-syntex-to-automate-content-categorization-and-build-a-foundation-for-knowledge-curation/) that it will be $5 per user per month which I think sounds a great amount for those power users to set content. If it's for all who view the content then that could be a harder sell to most organisations.

## What is next?

We have [already been told](https://techcommunity.microsoft.com/t5/project-cortex-blog/announcing-sharepoint-syntex/ba-p/1681139) that this is the first service in Project Cortex and that more will be generally available by the end of the year but what will those other services be? There have been details of Knowledge Hub sites that act as a landing page for knowledge on specific entities such as projects in an organisation as can be seen in [this demo](https://demobuilderwebcpptxz.blob.core.windows.net/microsoft-365-knowledge-sharing/startdemo.html?ot=false&lan=&guidemodeenabled=false&audioenabled=false). In addition, Project Cortex has talked about having knowledge experts who will be able to define specific knowledge on top of the automatically tagged info. There is also many comments about Cortex being the join of People and AI to ensure that it is useful and relevant to what people need.

I would like to see more around the use of Bots to surface Knowledge automatically to people that is relevant at the right time and based on what they are doing, especially with a greater integration to Microsoft Teams which is now the hub for most people doing work. It would also be interesting to see whether taxonomies can start to be recommended rather than just being defined by experts up front.

What is coming on October 1 looks very useful and the positioning of a set of services feels like a great approach to open this up to more people. I'll confess to being slightly disappointed that more wasn't released yet but will hold on to that for the end of the year!

Photo courtesy of [Edu Grande](https://unsplash.com/@edgr) via [Unsplash](https://unsplash.com)


Knowledge Management Series - Introduction to SharePoint Syntex

Culture - something many people are accused of lacking and something organisations strive to have. It's a hard thing to actually define though. When it comes to knowledge management though, it's a little simpler as what I am talking about is does your organisation encourage the sharing of information? Are people rewarded for working collaboratively and retaining the knowledge for the good of the organisation? Or are people encouraged to focus on themselves, hoarding information and not understanding the culture of sharing?

> This is part four of the Knowledge Management series. For other parts, see links below:
>
> - [Part One - Introduction](/2020/08/12/knowledge-management-series)
> - [Part Two - Considering where knowledge is stored](/2020/09/10/knowledge-management-series-considering-where-knowledge-is-stored)
> - [Part Three - Organising knowledge](/2020/09/14/knowledge-management-organising-knowledge)

The value of a sharing culture means that your knowledge management can become more valuable, leading to it being more used and building the benefits of productivity, innovation and togetherness. Staff can get information more easily and at the right time, allowing them to focus on getting done what they need to get done and inspiring them to build on the shoulders of others rather than recreating every single time. They can feel that they are working together with others to make the organisation the best that it can be, keeping employees engaged and excited about their job.

It seems a no brainer when put like this so why isn't everyone doing it? Well, some don't see the value in being so open, often because it can take time to be able to see the benefit to the bottom line and in some cases never seeing any improvement at all. Spending time to share things properly can feel like it is eating away at time that could be sent focused purely on the job. Thankfully, companies like AirBnB and Microsoft are showing the value in being open in what they are doing and demonstrating that organisations can be successful in doing it.

So the other reason why people don't? Because it is hard to know how to do it and it can take a lot of time and investment to change the culture of an organisation. I am going to try and approach a way that organisations can start to change that.

## Build Trust

The most important thing when trying to change a culture and, in this case, build the culture of sharing is that you need to have trust. If people take the time to share then they need to know that they are not going to lose out by it. Creating good content that is then taken by others as their own with no acknowledgement is a great way to lose the trust and lead to people not sharing. Chastising staff for spending time in capturing their knowledge instead of getting on with things is another great way to lose that trust.

Connected with this is the building of a no-fail culture. Allowing people to share and capture the lessons learnt when things have not worked means that people learn together. Understanding why when you skip the testing or allow that technical debt to build, things fail and don't work well further down the line helps to build the support to do things properly. It helps to justify people sticking to their guns and insist that things are done properly. Without the trust in place to allow people to learn from failure without fear of reprisals, then people will play safe and let the problems hit later.

If you want to read more about learning from failure, [this post](https://medium.com/@ShaneLester2016/5-best-books-about-learning-from-failure-db1fe8b2fac1) has some great books to read on the subject (with the [Scott Adams one](https://www.goodreads.com/book/show/17859574-how-to-fail-at-almost-everything-and-still-win-big) already on my "to read" list).

## Highlight influencers who are sharing well

As you build trust by not attacking failure and supporting the sharing of knowledge well, you can also take the positive aspect of highlighting the successes, especially specific people with influence within the organisation. In the modern culture where there is often a flatter structure to organisations and internal social networks allow people to influence through what they do as well as their role, this means that not everyone who can influence is a manager. By celebrating those that ["work out loud"](https://www.allthingsic.com/wol/), you encourage those behaviours and help others to learn how to do similarly themselves. Creating culture is about celebrating the things done well and promoting them widely.

## Top down sharing

Your influencers are key but organisations are still heavily influenced by their hierarchies and messages from senior management. To encourage others to share, having senior managers share their own knowledge, information and even thoughts will massively increase the likelihood that others do as well. Being open with what is happening and keeping the channels of communication going will show that it is not just what people are being told to do but what is the way that things will be done as well. "Do as I say, not as I do" isn't a phrase that helps in this situation.

There is a challenge that it must feel natural as forced sharing will often come across like that so allowing each manager to find their own path to this will help with the longevity of it and the acceptance from others. If sets of blogs and communications from the management team feels the same then the suspicion will arise that they are all coming from comms and aren't real. Allowing different personalities to come through will encourage diversity across the staff as well.

## Learning programmes

Now much as I love informal learning and learning by example from others, there is still absolutely a place for more formal programmes to help others learn and to encourage the sharing of knowledge. While there is a backlash for the use of brownbag lunch sessions (the feeling being that people should still be allowed a proper lunch break too), the benefit of having sessions where you share knowledge and understanding and have the opportunity to see what others are doing really help to build that understanding across different areas and to allow for new ideas to pr`opogate out.

My first fulltime job was in a graduate role and the sessions where we were introduced to what others across the organisation did really opened up my eyes. It allowed me later to see the true benefit in SharePoint across everyone because I could have discussions with different areas about their needs with a good initial understanding of what they did. This isn't just about technology or the specific knowledge associated with different areas of a business but understanding how things are done in different areas. Do they have projects or just get things done? How does each department interact with others internally as well as with other departments. All this knowledge can help when shared.

## Break the organisation hierarchy

I am not talking about starting civil unrest in the canteen at breakfast but encouraging communities of interest and purpose to be able to work together regardless of which department they sit in. Agile techniques can work across operations, HR, engineering, finance and IT so people with interest in that should work together to share what works for them, new ideas for being more productive and efficient and ways of doing things.

The sharing of knowledge through communities is a great indication of the ability of an organisation to be agile as it encourages the more matrix approach and ability to flex to challenges and opportunities that are faced. [Team of Teams by General McChrystal](https://thearmyleader.co.uk/team-of-teams/) highlights the benefits of smaller cross-functional teams and for those to function well, using the knowledge shared between different teams in the required skills. By being able to move people into different areas, you can help the understanding of different areas and build that culture of sharing.

## Make it easy

The last advice is to make it easy for people to share, help people fall into the right buckets. If they have hoops of approvals to go through to get knowledge out to others and it takes time that people rarely have then people won't do it. The natural inclination is to make sure that people capture all the information and metadata needed but this should be kept as easy as possible to do.

How do you make it easy? Using simple forms with as much filled in automatically as possible. Use context to provide the information rather than requiring people to do it themselves. Look at how modern AI tools can help automatically pull knowledge from existing areas (yup, I'll mention [Project Cortex](https://aka.ms/projectcortex) at this point too, it's been a while). Create easy to use tools that work within tools that people are using already like Microsoft Teams. Allow people to do what they were doing before but with that knowledge being stored too.

## Summary

It is not an easy task to change the direction of an organisation and to build a new culture, however big the organisation is. It takes not only time but the enthusiasm and desire to make that change from all involved. However, for knowledge management initiatives to succeed, it is key that people are central to it and that they are encouraged and supported along the way to help it happen.

Photo courtesy of [Chris Lawton](https://unsplash.com/@chrislawton) via [Unsplash](https://unsplash.com)


Knowledge Management Series - Creating a culture of knowledge

Organising can seem such a boring word. Organising is the opposite of spontaneous, it's the planned dinner you can't be bothered with instead of the quick pint after work that turns into an epic night out. It's what the young you is terrified that you will turn in to. It is also what makes things just work. It is what makes you be able to be productive, to be able to get on with what you need to do and be the most effective you can be. It's the front-row seats at the O2 after hours spent with multiple windows to get the tickets. But yeah, it is still a bit boring.

> This is part three of the Knowledge Management series. For other parts, see links below:
>
> - [Part One - Introduction](/2020/09/03/Knowledge-Management-Series.html)
> - [Part Two - Considering where knowledge is stored](/2020/09/10/Knowledge-Management-Series-Considering-where-knowledge-is-stored.html)

## So why do you need to be organised with your knowledge management efforts?

In the [last part](/2020-09-10-Knowledge-Management-Series-Considering-where-knowledge-is-stored.html) of this series, I talked about working out where your knowledge is held and in many cases, that can be pretty large areas. The example I gave for the etiquette of Microsoft 365 includes Twitter and that is certainly a big place to find information. Being organised means that you can extract/use useful information at the right time from those larger sources.

How though? How do you sort out the huge volumes to be useful? The starter point is to define your taxonomy of information, define how it is that you can categorise knowledge. To be able to understand your knowledge, you need to be able to have ways to classify it into areas, whether it is curated manually or defined with articificial intelligence. A central taxonomy helps to:

- bring standardisation across your organisation
- navigate knowledge
- reduce confusion of common terms
- reduce redundant information
- match to analytics on usage

Creating the taxonomy is not an easy task and is something that should be done with as broader a group of people as possible.

- Consider the areas that you would apply to knowledge and how people will use that knowledge.
- Consider the areas containing knowledge outlined in your knowledge map and how that would be categorised.
- Consider specific pieces of knowledge and how they could be categorised.
- Consider how a hierarchy of the terms could be created.
- Consider your taxonomy to be iterative and plan for change.

With those tips ringing in your ears, it is time to start building out a taxonomy for the etiquette of Microsoft 365.

## Building a taxonomy for the etiquette of Microsoft 365

The taxonomy can take several forms. For the etiquette, the natural place to start is the navigation that is already in place in the [repo](https://greyhatbeard.github.io/m365-etiquette/). These are:

- By service - a more technical approach to the different services in Microsoft 365
- By approach - a grouping of different ways of using the services
- By scenario - specific scenarios to consider when using services

These have naturally evolved from the contents used and so these make sense to start with. This can be extended with a similar notion of which groups of user it is relevant for, i.e. would it be standard users or power users or perhaps just for admins to consider. Another area added is for the strength of the advice to help guide people consuming the information as to how much they should consider the recommendations. Finally, the source of the knowledge was added which will allow for analytics to be run on the content at a later date.

![Taxonomy for the Etiquette of Microsoft 365](/images/2020/09/EtiquetteTaxonomySketch.jpg)

The next step (not one that I've undertaken quite yet) is to start mapping samples of your content itself using this taxonomy. Identifying exceptions that don't fit this taxonomy is important as it helps you to work out whether you need to extend the taxonomy or allow some bits of content to not quite fit. Looking at [Limitations of Private Channels](https://greyhatbeard.github.io/m365-etiquette/By-service/teams/) in our repo, there's some great advice but it doesn't feel quite like it fits the must, should, might level exactly and hard to say if it would be for power users or admins. We will need to make a decision as to whether we update the content there or adapt the taxonomy. We will certainly need to break many of the pages into smaller blocks as parts of the pages have a real mixture of content. This shows a good example of working out whether to change your taxonomy, change your content or perhaps to have a transformation process into a central repo is the right thing to do.

## How to store your taxonomy

This is always a tricky one to answer. The important factors to consider are:

- Visibility to all
- Ability to consume with APIs
- Clear ownership
- Ability to request additions and changes

There are many services that can be purchased or databases built with APIs to support these factors. Coming from a SharePoint background, my natural position would be to store these in the taxonomy service and I will dig further into that in a future post in this series, considering how to make it more visible and easier to request additions and changes for all staff.

For the etiquette of Microsoft 365, all the content is held currently in a [YAML file](https://github.com/kevmcdonk/m365-etiquette/blob/master/docs/metadata.yml) with indents to show the taxonomy within that. It's an early beta so not yet extensive but gives an idea of how it will look when past the OneNote. The downside with YAML files is adding additional metadata such as alternative names capturing owners in the way that you could in the SharePoint Managed Metadata service. A future change could be to create a JSON file to hold this with the additional details as attributes. But this works nicely for now and for this series.

<script src="https://gist.github.com/kevmcdonk/60dd92ffc914348edf8a226ddcbd67cb.js"></script>

## What about an onotology?

Some of you reading this may have heard about ontologies as well as taxonomies. For me, the simple difference is that taxonomies will have a single hierarchy for each item in the taxonomy. Ontologies allow for more complex relationships to be put in place such as where departments and functions of an organisation may overlap. A taxonomy would be described in a tree view whereas an ontology is more like a web diagram linking different nodes.

If you are unsure, then stick with the simpler taxonomies as these are easier to implement with services like SharePoint Managed Metadata. As your information becomes more complex, then you could consider moving to the more complete ability to map relationships with an ontology instead.

For more information on the differences, the following articles may help:

- [Forbes - taxonomies vs ontologies](https://www.forbes.com/sites/cognitiveworld/2019/03/24/taxonomies-vs-ontologies/#79e714f57d53)
- [Earley - What is the difference between taxonomy and ontology](https://www.earley.com/blog/what-difference-between-taxonomy-and-ontology-it-matter-complexity)
- [Pedia - difference between taxonomy and ontology](https://pediaa.com/difference-between-taxonomy-and-ontology/)

## Driving change and adoption

Once you have begun the process of planning your knowledge, it is vital that you have a broad set of contributors, that you have opened up to people who will be using the metadata to make sure that you have the right information in place. For example, if your metadata includes an organisational structure but your HR systems use a different approach to your Finance systems then you have already excluded a large part of your potential user base. Engaging well and early with your key users of the metadata will help to ensure that the organised knowledge will be used - it may make things take longer as you adapt to different needs but it will be time well spent for later.

Broadening out to more users initially can lead to a much higher scope of work so be prepared to engage with areas that you may open up to later, be prepared to disappoint some people initially so that you can gain momentum instead of being bogged down in never ending requirements definition instead of earning some value. Don't be afraid of adapating an agile/continuous improvement approach to your knowledge metadata, being open with progress.

You should also make sure that you consider how the metadata will be managed:

- How will people request changes?
- Who will have ownership of different areas?
- Will there be an automated approval process?
- How will you know what is being used and what isn't?

The SharePoint Managed Metadata service, as I mentioned above, isn't great at these things so it is important that you consider how you are going to handle the requests, how you will monitor usage up front. It may be that from day one that you won't but do plan to get scripts in place to help with this or engage with some of the great third parties that are out there.

## Wrap-up

As with many collaboration systems, governance is key to Knowledge Management and organising your knowledge to ensure easy access and understanding will make sure that it is used effectively. Building up the metadata so that you do not always have to pull your data into one place gives you the flexibility needed for modern agile organisations while also giving that standardised structure to keep things consistent.

Engaging with others is key and part of that change and adoption is building a culture where people do understand the value of sharing knowledge and people are encouraged to actually do it. Building that culture is the subject of the next blog post.

Photo courtesy of [Faris Mohammed](https://unsplash.com/@pkmfaris) via [Unsplash](https://unsplash.com)


Knowledge Management Series - Organising knowledge

There are many ways to start a knowledge management journey such as starting in one specific area of your organisation, focusing on a set of key stake holders or just building a business case for the value of a large scale program of work. For this series, I will focus first on where your knowledge is stored, considering the different areas and ways that information can be held. This will help you later to consider how you make use of that information to help the right people at the right time.

> This is part two of the Knowledge Management series. For other parts, see links below:
>
> - [Part One - Introduction](/2020/09/03/Knowledge-Management-Series.html)

So where would you find knowledge in your own organisation? Your are probably thinking of formal locations such as your Intranet, support guides, wikis and formal documentation. You may also be thinking of less structured areas such as your departmental shared drives or your Share Point Team Sites for collaboration. But what about all those conversations people have in Teams and in email, chats that take place in meetings and all that incredible experience stuck in people's heads? I bet that you can think of many times where you have read a formal document but still needed to contact the author or even someone else that you know with the experience. With even the best Knowledge Management programs, huge swathes of knowledge will still be stuck in people's heads. So what should you do?

## Types of knowledge

Knowledge can be broken down into two key areas:

- Explicit knowledge - knowledge captured for the purpose of holding knowledge
  - Structured content - e.g. Record management systems
  - Unstructured content - e.g. File shares, wikis, meeting minutes
- Tacit knowledge - knowledge held informally that can be hard to share with others
  - Discussion forums
  - Enterprise social media
  - Emails

A third area sometimes cited is implicit knowledge which is the knowledge that is primarily in people's heads. If you think back to when you joined a new organisation or a new team, then there is always that time where you pick people's brains about what to do and how to do things. If you are lucky, then there is a page or document that lists things out but there is always that implicit knowledge not written down even if it is as simple as the best time to beat the crowds for the coffee shop.

You can think of these three areas as a scale with explicit knowledge in structured areas being the hardest to capture and maintain but the easiest to search and navigate knowledge and implicit knowledge being the hardest to find but easiest to capture.
![Types of knowledge](/images/2020/09/TypesOfKnowledge.jpg)

## So where is your knowledge held and what types of knowledge do you have?

Do you have formal record management or document management? The first steps to getting your knowledge into a usable state is to build out a knowledge map which helps you to capture the areas with knowledge at that point in time. It is important to realise that it will be an iterative process as you start to think about the knowledge, discover locations for that knowledge and then realise other areas of knowledge. If you are looking for more assets on knowledge maps then APQC ([American Productivity & Quality Center](https://www.apqc.org/)) have a lot of tools available, some for free but most for members. Much of what I have learnt recently has come from one of their books, [The New Edge in Knowledge](http://www.newedgeinknowledge.com/) which I would recommend even though it is a little old now.

The video below is a replay of a mapping of the Etiquette of Microsoft 365 knowledge, considering where useful knowledge could be held to learn more about etiquette insights similar to those we have captured for GreyHatBeard. You can see that I think about the main areas and then start extending that to specifics where known and building out the connected areas. It can also be useful to start considering at this point how much knowledge you have in each area and how much you would like to have.

![Mapping etiquette of Microsoft 365](/images/2020/09/MappingEtiquetteOfMicrosoft365.gif)

You may be looking at this and thinking that this is far too basic for an organisational knowledge map. That is certainly true and this lower level example would work better for an individual team or area. [APQC](https://www.apqc.org/) include a set of examples of knowledge maps for enterprise, cross-functional and process/role based maps in their [Getting Started with Knowledge Mapping](https://www.apqc.org/system/files/K06334_Getting_Started_Knowledge_Mapping.pdf) document.

The important consideration is not about the technical systems themselves but the knowledge in them and how it is used. That is the key really - how is it used. Which brings us neatly on to the next in the series on how you can organise your knowledge.

Photo courtesy of [NASA](https://unsplash.com/@nasa?utm_source=unsplash&utm_medium=referral&utm_content=creditCopyText) via [Unsplash](https://unsplash.com)

NB - if you noticed that the animated gif was from a OneNote and wanted to do the same, this video has info on how I made it - [OneNote Playback for Meetings – Modern Workplace Tips](https://www.youtube.com/watch?v=3dDlo1qAhPA)


Knowledge Management Series - Considering where knowledge is stored

Have you heard about Project Cortex? The upcoming service from Microsoft that will help organisations to better manage their internal knowledge by using the best of breed content services such as SharePoint, One Drive and Microsoft search.

> This is part one of the Knowledge Management series. For other parts, see links below:

> - [Part Two - Considering where knowledge is stored](/2020/09/10/knowledge-management-series-considering-where-knowledge-is-stored)
> - [Part Three - Organising knowledge](/2020/09/14/knowledge-management-organising-knowledge)
> - [Part Four - Creating a culture of knowledge](/2020/09/17/knowledge-management-series-creating-a-culture-of-knowledge)
> - [Part Five - An introduction to SharePoint Syntex](/2020/09/26/knowledge-management-series-introduction-to-sharepoint-syntex)
> - [Part Six - An introduction to Viva Topics](/2021/02/06/knowledge-management-series-introduction-to-viva-topics)

Your current reaction probably sits in one of three places if you live in the Microsoft space.

- "I'm so excited" - you can't wait for darter and all the benefits it will bring to ease access to information,
- "It's going to be too expensive" -you might be excited about it or not but either way, you just don't think you'll get the budget To get anywhere with it.
- "Erm, what is Cortex?" it may be that you haven't heard about Cortex and the buzz that started at Ignite in 2014 -checkout https://aka.ms/ProjectCortex as a starter or
  listen to our first Grey Hat Beard show

Regardless of your current state of mind or Cortex,
the aim of this series of posts is to get you thinking about knowledge in your environment.
Where is it stored, how do your store it, formal
informal knowledge and how can you get access to that knowledge for the right people at the right Time. Much of this will be made easier by Cortex but you can gain huge advantages by implementing some of these suggestions Without Cortex and if you will be
implementing Cortex then you will still need to
plan for many of these things.

• Add something about what is knowledge management
• aim is to connect those with knowledge to those who want (need knowledge
To help this series to be practical as well
• theoretical, lam going to demonstrate
some of the practices using the Greybeard
etiquette of Microsoft 365 content. the
ones will contain:
-Considering where knowledge is stored -Organising knowledge
-Creating a culture of knowledge
-Records management us. knowledge might
• Al driven access to knowledge
• Searching and browsing through knowledge

Photo courtesy of [Patrick Tomasso](https://unsplash.com/@impatrickt) via [Unsplash](https://unsplash.com)


Knowledge Management Series

Microsoft Search was announced as the single search for finding items across the whole of the Microsoft 365 world back in 2017 during the Ignite that year. One backend to reference so that no matter where you search from, you get the same results back. That's the promise and that is gradually coming to fruition. I will cover more on how this looks in SharePoint but this post will look in to the Dev side - how can Microsoft Search be used in a SPA (aka Single Page Application - sorry, no massages here)?

TL;DR - take a look at the code at [https://github.com/kevmcdonk/ms-search-spa](https://github.com/kevmcdonk/ms-search-spa) to see how it's done.

That fateful conversation. "I really like this but can it look a bit less like SharePoint?". It's a hard one as Modern SharePoint offers so much to people but there are those that have been bitten before and just don't want the flexibility that SharePoint brings and don't want the space grabbing that it can take as well. Usually, I try and push harder but in this case, there was little that the client wanted outside of the ability to easily search enterprise content. They wanted a way to get to their other web applications and other content faster. What did they like? Microsoft Search.

The power that Microsoft Search offers to index content from a multitude of places has increased the interest. It allows users to search content from across Microsoft 365 in a single way, knowing that results will come back the same regardless of where you search from. It also now allows you to pull content from outside of Microsoft 365 so that your single search can be aligned to all of your content. I spoke more of this at the [GlobalCon1 event](https://events.collab365.community/globalcon1-agenda/) run by [Collab 365](https://collab365.community/) that you can sign up for the [All Access Pass](https://partners.collab365.community/13990/41362) to view, including an eBook on the subject.

What was announced at Ignite in 2019 was that this search would now be available using the Microsoft Graph. There are only certain areas available at the moment such as searching files, events, messages and external items - you can read more in the [Graph Docs](https://docs.microsoft.com/en-us/graph/api/resources/search-api-overview?view=graph-rest-beta). The sample that I have put together makes use of these APIs as well as some others to retrieve SharePoint info. Sadly, not quite all the needs are available in Graph so I've had to make some SharePoint REST calls too but we'll get to that.

## How it looks

The UI focuses on the ability to search for contents with a large search box as the focus. Below that, it shows the trending news and then follows up with other trending items using the Graph and SharePoint APIs.

![Main page](/images/2020/08/mssearchspa-1.jpg)

![Second half of main page](/images/2020/08/mssearchspa-2.jpg)

The search bar returns results straight away and uses the Adapative Cards to render the results although these could do with some improving to look a bit better!

![Search results](/images/2020/08/mssearchspa-3.jpg)

## Framework

The overall app utilises the Next JS framework to create a static app that can be hosted on Azure Storage. This makes it very cheap to run and relatively simple once the code is compiled. Azure DevOps can build and deploy all the files and get the site running in minutes. Which is nice.

Next JS is an opionated framework focused on Single Page Applications with React. It handles the routing and used WebPack to run dev locally with the Hot Modules - all the things you would expect from a modern framework. Big thanks to [Garry Trinder](https://www.twitter.com/garrytrinder) for getting this set up as it makes things a lot smoother than my original, pure JS version. He also put plenty of work in to the other elements so huge kudos to that man.

I used MSAL to utilise Azure AD for authentication and Office UI Fabric for the UI. The final part to help with rendering the search results is to use Adaptive Cards and Adaptive Templates. For the rest of this post, the focus is on the API calls (Graph and SharePoint) with a little Adaptive thrown in too and most of these are located in the [Services](https://github.com/kevmcdonk/ms-search-spa/tree/master/services) folder.

## Calling the Microsoft Search Graph APIs

The [Search.js](https://github.com/kevmcdonk/ms-search-spa/blob/master/services/search.js) file shows how the Search API is queried. It receives a token that is retrieved from MSAL and accepts a searchQuery, a string that contains the query that is comes from the search box the user types in. The call is pretty basic with the only addition being that you can control the number of results to return and whether to skip results, allowing you to page through records.

<script src="https://gist.github.com/kevmcdonk/5cb7f9a6a72f58628dd57e826af49489.js"></script>

The limitations really are that you can only return one entity type at a time. To return all events and messages, you would need to make two separate calls and merge them. Same with files (aka Drive Items) which slightly differs from the one search message at the moment. It does simplfy the response though as the entities returned are just those that you search for.

## Adaptive Cards and Templates

The results returned from search are nicely rendered from the Graph using Adaptive Cards. These are used in the [Graph Explorer Preview](https://developer.microsoft.com/en-us/graph/graph-explorer/preview) to show how the results can be rendered which demonstrates the direction of importance that these are taking.

The best place to find out more about Adaptive Cards is at [https://adaptivecards.io](https://adaptivecards.io). There you can view samples as well as use the Designer to build your own cards.

The card templates are used in the [search.js component](https://github.com/kevmcdonk/ms-search-spa/blob/master/components/search.js), rendering the responses back from the Graph or SharePoint calls. The templates themselves are held as JSON files in the [cards folder](https://github.com/kevmcdonk/ms-search-spa/tree/master/cards) and added in to the component. The key packages used are adaptivecards, adaptivecards-fabric and adaptivecards-templating which combine to allow you to render the UI in a Fabric friendly way. You can see below that can add some config like the default font (good old Segoe UI in this case) and that you can create an action for when actions take place with the card.

The templating is still in preview so approach with caution and be ready for some changes although I found that it worked pretty well once you find the right samples. However, if you have used previously, it is worth noting the breaking changes as of May 2020 noted at [https://docs.microsoft.com/en-us/adaptive-cards/templating/](https://docs.microsoft.com/en-us/adaptive-cards/templating/).

<script src="https://gist.github.com/kevmcdonk/288e71d02b60e49889f982530a273119.js"></script>

## Getting News from SharePoint

To get the news from SharePoint, I would love to use Graph but at the time of developing it, there was not enough support for the Search of news articles so I went back to using the SharePoint REST API. I added a filter for the Page content type and also for PromotedState as 2 to ensure that it was just News pages and not other standard pages.

<script src="https://gist.github.com/kevmcdonk/09a4dd977f7f4fec418d927d1ea82518.js"></script>

## Trending Items

For the trending items, things were a lot easier as I could make use of the trending insights functionality in the Microsoft Graph.

<script src="https://gist.github.com/kevmcdonk/aa229abc623037431860a33fe6d80999.js"></script>

## Using SitePages API

One of the areas that could be a little controversial in this code is the use of the SitePages API. This is used in SharePoint itself for actions on SitePages but is not formally supported as there is no documentation for it. I made use of it to return specific news pages filtered to only return files (the filter of FSObjType=0). However, the risk of using this is that it may stop working at any point so should be used cautiously.

<script src="https://gist.github.com/kevmcdonk/59924e7072bb4286cb7fa8ec3a84f302.js"></script>

## Next steps

While I was happy with how this works, I would love to implement a little more to help with testing. Firstly, I would extend the existing Cypress tests to do a full UI test based on some of the great work that Elio Struyf has done.

Next, I would improve this with deployment scripts that set up test data to use in a fresh environment wiht sample applications set up to be searched. It would also be good to get additional services to understand better which applications are in use and show those as the first option.

Any other ideas you have, I'd love to hear them via [Twitter](https://twitter.com/kevmcdonk) or as suggestions in the [GitHub issues](https://github.com/kevmcdonk/ms-search-spa/issues?q=is%3Aissue+is%3Aopen+sort%3Aupdated-desc).


Using Microsoft Search with a SPA

Have you ever read a blogpost telling you that you should focus and just laughed? Do you really see the point of focusing on sticking to one thing or do you just think you need to get on with all these different things as they are all a priority? What is the point of focusing?

Well the honest answer is that we are finding out right now. We are seeing people able to focus on one thing because that is the thing that is front of most people's minds – Covid-19. What if you have to invent a new ventilator to try and save more lives? [People have done it](https://www.bbc.co.uk/news/business-52309294). What if you had to find a way to inspire thousands through a national singalong or clapping along to support the NHS in the UK? [People have done it](https://www.bbc.co.uk/news/av/uk-52319191/coronavirus-uk-residents-clap-for-our-carers). Why have things like this not been happening all the time, why has it taken a global crisis for this to happen? The answer is focus. When you focus on one clear thing, you can get it done. It’s not always easy to focus but the power of it can be seen as we look around and read the news right here right now.

## TL;DR

Have a listen to our [GreyHatBeard podcast](https://www.greyhatbeard.com) on the subject or read on.

<iframe src="https://open.spotify.com/embed-podcast/episode/0uhfXIgSBvdf9EGcNsRJDl" width="100%" height="232" frameborder="0" allowtransparency="true" allow="encrypted-media"></iframe>

## So how do you focus?

As with many solutions, there’s always a harder question to answer. How do you dedicate your time to one thing when there are all sorts of challenging priorities around you? When the kids need you to help them with their homework, your wife is asking what shopping is needed for the weekly trip, your colleagues are asking how to solve something and HR is asking whether you need to take that leave and all at the same time, how do you focus on getting just the one thing happening?

If you try and juggle all those different things, you are unlikely to get any of them done and context switching between the different things will mean you are less effective on each of them. By the end of the day, you will be exhausted and your to-do list won’t have changed. You will finish with substandard results for each task, unhappy family, agree work life and probably no leave to recover.

## Prioritise

So the first step has to be to list and prioritise. Sorting out which is more important is hard but first you need to get everything down so that you can. For me, I like to use Microsoft ToDo as it will already pull my tasks from Planner but you could use a piece of paper or post-its too (these are better as you can then sort). Work through the list and decide then on what has to be done first.

## One thing at a time

Now that you have a list of things to work through, you can look at one thing at a time. Shut off your notifications, ignore you email and work on that one thing until done. If it's too big, then break the task down a bit but stick to the one thing at a time.

Leave some time for emails and other social media but don't keep swapping between different things. Just get one done at a time.

## Use your environment and equipment

To stop yourself from being disturbed, use technology like noise cancelling headphones and think about where you sit. Pick places where you won't be disturbed. Use different desks (or rooms in the house at the moment due to lockdown) for different tasks so your mind is in the right place for it. Some places for documentation, another for meetings, somewhere else for coding perhaps.

Also use different music, podcasts and audio books for different types of work. High tempo for coding works for me but lower intensity for writing documents and reading articles. For repetitive tasks, having an audio book or podcast in the background works.

## Where can I find out more?

There are so many things out there on focusing, amusingly enough to really distract you. Great books such as [The Organized Mind](https://www.amazon.co.uk/dp/B00M8PUC8K) or [Getting Things Done](https://www.amazon.co.uk/Getting-Things-Done-Stress-free-Productivity-ebook/dp/B00SHL3V8M) will help. Or take a listen to [GreyHatBeard](https://www.greyhatbeard.com/2020-04-17-show8part2/) - we'd love to know what you think!

<iframe type="text/html" width="336" height="550" frameborder="0" allowfullscreen style="max-width:100%" src="https://read.amazon.co.uk/kp/card?asin=B00M8PUC8K&preview=inline&linkCode=kpe&ref_=cm_sw_r_kb_dp_mkhNEbFBDP5D0" ></iframe>
<iframe type="text/html" width="336" height="550" frameborder="0" allowfullscreen style="max-width:100%" src="https://read.amazon.co.uk/kp/card?asin=B00SHL3V8M&preview=inline&linkCode=kpe&ref_=cm_sw_r_kb_dp_ilhNEbHQ0726J" ></iframe>

Image by [JESHOOTS.COM](https://unsplash.com/@jeshoots) from [Unsplash](https://Unsplash.com)


Covid 19 shows the power of focus

Blog

My involvement in the Microsoft Community