lost and found box
PHOTO: Paul Gorbould

The unification of the Microsoft productivity suite into a single Office 365 cloud platform was supposed to reduce the age-old challenge of finding important business information quickly. Paradoxically, the flood of Office 365 apps with overlapping functionality has made the problem more acute. For example, to find account information buried in customer communications, workers need to sift through Outlook, Teams, Yammer, SharePoint, OneDrive and possibly even Skype for Business.  

Is help on the way?

Microsoft Graph to the Rescue?

Microsoft Graph is a technology that runs behind the scenes of your Office 365 experience, capturing and logging the content you create, and linking it to the people with whom the content is shared.

The best way to visualize how the Graph works is to open the Microsoft app, Delve. Delve is a billed a way to “discover and organize the information that is most likely to be most interesting to you right now — across Office 365.” It is basically a visual representation of Microsoft Graph recommendations, presented as a series of cards displaying documents, emails and calendar invitations created by people you work with. The concept behind the Graph and Delve is that if you are most likely to be interested in other items being created and shared by close colleagues.

It’s as if you own a store and Microsoft is sending you a newsletter about what all the other stores in your area have on sale this week. While some of this newsletter content may be interesting, it’s likely a portion will not be, because most of the local stores are not your competitors. Finding the valuable information nuggets requires some cognitive effort. And what you really want to see is what your competitors across town are selling, but the newsletter doesn't include this information.

In a similar fashion, what Microsoft Graph/Delve surfaces are what workers around you are doing. While this might be interesting, this approach possesses two deficiencies: First, your colleagues are working on many projects that don’t concern you. Second, people outside your work circle are toiling away on the kind of contracts, proposals and projects that you are — but alas, the Graph doesn’t capture this information.

These two deficiencies have led to Delve’s demise. (Rumor has it a new and improved Delve will be coming later this year. Stay tuned.)

So, if the Microsoft Graph is not the key to bringing together the information you need to work from all your various Office 365 apps, is there another solution on the horizon? 

Related Article: Microsoft Graph Data Connect: The Glue That Binds Office 365 and Azure 

Microsoft Search to the Rescue?

Last month, Microsoft Search was released to the general public. The purpose of Search is to provide “personalized insights surfaced by the Microsoft Graph … across all your applications.” Is it possible to query the Microsoft Graph and get more relevant results than what it presented with Delve?

The answer seems to be a resounding …. sort of. Today — not really. In the future — maybe. What Search today provides is a consistent user experience across Office 365 apps on the desktop, in the browser, and on mobile devices. Translation? The search window appears in a consistent location across your apps and devices. And with Search, you can look for files, conversations and people. Results will incorporate the relationships between people and content as captured in the Microsoft Graph, but no more. If you want, you can integrate Bing searches to return content from outside the organization as well.

But since the relationships captured in the Graph are what Delve presented, search ultimately encompasses the same weakness that was present in Delve. But it’s better than Delve, because at least Search won’t display irrelevant files and conversations from people within your work circle, only those that are relevant to your search term.

The key to surfacing relevant information quickly across all the apps is reliant on something still missing from Office 365. What’s missing are relationships between people and the products, projects, and services on which they are working, what is often called the semantic view of organization activity.

Related Article: Looking at the Evolution of Microsoft Search

Enterprise Knowledge Graph to the Rescue?

The semantic view of an organization is captured in an enterprise ‘knowledge graph,’ which is a construct that collects and integrates inputs from a wide variety of data sources, not only Office 365. In our scenario, inputs could be emails, documents, calendar events, app notifications, as well the people who created, viewed, edited and shared these items. What’s unique about the knowledge graph is it captures the reason why specific people and items are connected. For example, a group of geographically dispersed employees might be connected because they are working on similar projects for different customers, or documents created by different business units might be connected because they relate to the same compliance directives. In both cases, a worker querying a knowledge graph would be able to uncover highly relevant set of results because the items are connected in a meaningful way.

This type of connectivity is missing from the Microsoft Graph. Apparently, there is a research project at Microsoft to build a knowledge graph called Enterprise Graph, but it seems years away from productization.

Related Article: 6 Mistakes to Avoid With Microsoft's Hybrid Search

Where Does That Leave Us?

The knowledge graph concept is not new. Google introduced a knowledge graph in 2012, for the consumer market. With it, when you perform a search, Google’s knowledge graph is at work under the covers, connecting you to the most relevant information via a collection of related topics that are updated continuously.

For the enterprise market, the situation is more complex. Unlike the consumer space, where Google has devoted massive resources to solving the problem of connecting many different data sources, each organization would have to do the heavy lifting themselves; including capturing its own unique taxonomy, its complete set of applications, and a definition of the business relationships between functional units. This information would have to be fused into the enterprise graph. Today, you need to connect the dots yourself or pay someone to do it for you. Presently, this is a herculean effort, undertaken by only the most information-intense organizations.

Don’t lose hope. We are starting to see the seeds of a knowledge graph revolution. In 2018, Gartner added knowledge graphs to its hype cycles for AI and for Emerging Technologies, meaning it is gaining traction among early adopters. And Microsoft’s research project is also a positive sign. Plus, there are specialty-firms like Cambridge Semantics, AllegroGraph and Kenome, who already provide building blocks for creating your own enterprise knowledge graph.  

But make no mistake — we still have a way to go.

In the meantime, workers will continue to struggle to piece together information from multiple apps, in Office 365 and beyond. In commemoration of the 50th anniversary of the first moon landing, we can summarize by saying that the Microsoft Graph and Microsoft Search represent one small step for man rather than one giant (knowledge graph) leap for mankind.