I still meet now and then IT managers that don't get the value of using metadata to manage unstructured content such as text documents, images, and video. They claim metadata is the old approach, and the new approach is using search with machine learning and artificial intelligence to manage unstructured content.

In iTunes, how would you find Rock'n Roll songs from the 70s, but not 80s, without metadata? You could search for song title and artist, but you would then only find the song. You could search for Rock´n Roll, but you would without metadata only find songs with this in the title. Metadata is key to finding the music you want, - this could genre, artist, composer, etc.

It´s the same for Office 365. You get some out-of-the-box metadata such as author, date, and file format, but you will NOT be able to identify business area, country, process, information type, etc. without the use of metadata. Metadata helps you search and find the information you are looking for (e.g. business unit, process, country), but also manage this over time (e.g. retention, access). Machine learning and artificial intelligence will help you extract and apply metadata, but you still need metadata with master-lists of business area, process, information type, etc.

If you have a corporate metadata model, then you have a standard way to describe unstructured information. If you want to improve findability, but also break down information silos, then the metadata values should be standardized using master data sources or controlled vocabularies (see below image).

Metadata is an investment in the future findability of information. The more you invest time in developing the information, the more important it is to ensure you and your colleagues can find and reuse the information. You need therefore metadata to manage unstructured information, but you may leverage machine learning and artificial intelligence to extract and apply metadata (often called entity extraction).

Metadata is key to automating records management with SharePoint Online and Record Labels. Office 365 can be set up to auto-apply a record label based on metadata. If users classify a document as an HR file and set document status to final, then this can be automatically declared as a record based on the metadata. Document type can be a mandatory metadata field, which also help users to search and find different types of information. Document status can be draft by default, and changing it to final will help users search and find draft vs final information.

Success is always 80% about people, 15% about process, and 5% about the technology. Metadata for Teams and SharePoint Online has to be of value to users. They have to see the value of it, - and it should not require them to manually add a lot metadata. The key to this is automation, e.g. information automatically inherent metadata based on storage location. Don´t force users to add them all.

How do you implement metadata for Teams and SharePoint Online?

When you create Teams, Planner, Yammer, and Stream, a SharePoint Online site is automatically created to store files. Teams now support metadata, - users can can see SharePoint metadata directly in the Teams client. Here are the high-level steps to implement metadata for Teams and SharePoint Online to improve search and automate records management.

Step 1: Establish a corporate metadata model with master data sources

Determine what metadata you need to find and manage the information. Some metadata may be mandatory for all Office 365 sites, like business area, country, and process area. Others may be optional like facility, department, product, employee number, etc. Identify possible master-data sources for each metadata element, e.g. list of all facilities. Standardizing the metadata elements AND values will make it easier to identify and manage information across sites.

Step 2: Connect SharePoint Term Store with master data sources

Use standard APIs to connect SharePoint metadata store with your master data sources. If this is not possible, you then have to establish a process to upload the lists to the SharePoint term store on a regular basis.

Step 3: Site Configuration / Site Provisioning to set default metadata on new sites

When new SharePoint Online sites are created manually, or automatically when creating Teams, Planner, Yammer, and Stream, you want to ensure these sites get the relevant default metadata. All information stored on the site will then automatically inherent the metadata, which will improve search and allow you to automate records management with SharePoint Online. Two ways to achieve this:

• Option A: Site provisioning with a central service that set up the site with content type for users with relevant metadata.
• Option B: Self-service site configuration with content type that forces site creators to pick and set relevant default metadata. This could be done using out-of-the-box SharePoint Site Design with Logic App.

Users are then empowered to create Teams and SharePoint sites with the required default metadata. This makes it possible to search for information per business area, country, etc. Records management with SharePoint Online may be set up to be automatic based on mandatory metadata, e.g. document type (e.g. HR file) and document status (draft vs final). The site configuration/provisioning should also include a site directory to make it easier to track site owners and purpose, e.g. what metadata has been set on sites.

Step 4: Get metadata on pre-existing SharePoint Online sites.

Develop and run scripts to set metadata on pre-existing sites and documents. This means all your sites will have governance, not only the new sites created with the new site configuration / site provisioning.

Step 5: Configure corporate and site search to show metadata as progressive filters

Metadata on documents will now make it a lot easier to find information in Office 365 based on business area, country, process, document type, document status, etc. Basic search can show metadata as progressive filters on the search result page, and advanced search can allow users to search only using metadata.

Step 6: Establish communication and training resources

Get business leaders to communicate the value of this to staff and create simple 2-minute videos explaining how to create a site, how to classify information, how to find information, how to archive information. All communication and training should start with the WHY. Explain first the benefit, then how to do it.

Step 7: Establish roles and responsibilities for ongoing improvement

The business will always change, and so will the requirements for new metadata elements and values. Establish therefore processes with roles and responsibilities for ensuring new or incorrect metadata can be quickly added or corrected.

This approach creates value to the business and users. It will make it easier for them to search and find information with automated records management with SharePoint Online and Record Labels. Site creators set the default metadata for sites that information automatically inherent when being stored on sites. Users only select document type (unless also set as default on site) and changes document status to final (unless set default on folder, e.g. Final Reports) to automate records management with SharePoint Online.

Feel free to contact us if you need help implementing metadata in Office 365 to improve search or to automate records management with SharePoint Online.

Atle Skjekkeland is CEO of Infotechtion is a vendor-independent boutique consulting firm specializing in improving and automating information protection and governance in Office 365 and beyond.