Future of Work #4 — Enterprise Knowledge Management 2.0 + AI-Powered Search & Discovery = Team Intelligence

Photo by Mike Wilson on Unsplash

As a VC, I’ve been researching and tracking technologies broadly within the “Future of Work” space. Ask any VC and you’ll get different responses as to what Future of Work covers — be it collaboration, productivity, AR/VR, RPA, and more. So I thought I’d share some of our early learnings at Work-Bench, which are ever-evolving — we’d love to hear from you, and to continue to learn.

As an enterprise-focused VC fund, we invest in startups selling into large enterprise customers, and we engage regularly with Fortune 500 line of business buyers.

Trends Across Workplace Productivity & Collaboration, Search & Enterprise Knowledge Management

1. More than ever today, there is rapid proliferation of cloud apps at work, with everyday usage on productivity and collaboration platforms like chat, email, calendars, meeting notes, project plans, CRM, social, issue tracking, customer support tickets and more.

And many more productivity and collaboration apps…

2. With these platforms comes vast mountains of data — both structured and unstructured — across millions of messages, documents, and more.

3. As an end user, it becomes harder and harder to search, find, and locate knowledge within an org — be it a document, an email, or a team member.

4. Well-established players like Elasticsearch, Coveo, Attivio, Microsoft and Lucidworks lead in the enterprise search space, but are geared for larger company environments and specifically for search functionality.

5. Legacy internal wikis and intranets like Confluence and Sharepoint where teams input documentation quickly lose adoption because the manual maintenance of these wikis are painful. And as soon as the wiki is out of date, its utility goes dramatically down.

6. Enterprise knowledge fragmentation leads to huge loss of productivity, with IDC finding:

• A typical knowledge organization employing 1,000 knowledge workers wastes over $5.7 million annually searching for but not finding information.
• 36% of a typical knowledge worker’s day is spent looking for and consolidating information spread across a variety of systems. These workers can find the information required to do their jobs only 56% of the time.

Thesis: Rise of the Team Intelligence Layer

Enterprise knowledge management 1.0 — with legacy vendors like Confluence and Sharepoint — still require manual and painful input, organization, and maintenance.(As anyone who has ever been part of a team Dropbox folder can attest — it becomes sprawling chaos very quickly, unless there is tight discipline around file naming nomenclature and organization.)

Just as the legal industry now has legal search and discovery to find documents faster, eliminate down time, and be more productive, there is opportunity for a new generation of enterprise search & discovery, collaboration and knowledge management tools, that converge and combined with machine-learning capabilities, will create a new layer of team intelligence that:

1. Integrates across all cloud collaboration and productivity platforms
2. Uses this body of data to build an enterprise knowledge graph
3. Utilizes machine-learning algorithms to better extract meaningful and relevant information, which can then auto-pull, auto-populate, or recommend the right type knowledge or insight right when it is needed, either into a central repository or as an overlay or plugin right into the user’s workflow and application
4. Search will no longer be about just document search across systems — but also as a way to identify subject matter experts (SMEs) and people within an organization.
5. What is more exciting and still to come is that dynamic ML “team intelligence” layer that can derive insights from this one central knowledge base, for insights you can’t or don’t even know to search for now, that can:

• Surface the right content: For a customer support team to pull policy based on pattern matching, to cut down on response times.
• Offer predictive recommendations: For sales enablement, pull from email, CRM, marketing engagement touchpoints, and perhaps in the near future even voice call sources to correlate wins to actions taken (i.e. see Work-Bench portfolio company Dialpad’s recent acquisition of TalkIQ and their newly launched conversational intelligence platform, VoiceAI).
• Identify the right person in a large enterprise org: Be able to understand who on your team is best suited to sell to a prospect, based on their expertise and background; route requests to the right PM or internal subject matter expert; or best form teams based on competency or experience.

Anything that requires manual note-taking and input in wikis won’t be able to compete longer-term with companies that can auto-build a knowledge base that continuously and automatically updates; and that can sit as a seamless interface layer right on top of workflow.

Major vendors operating in this space include Google with Google Cloud Search, across all Google Apps, Slack with Slack Search for subject matter experts, and Microsoft tying its Team product to their extensive Office 365 integration.

Key aspects that will be critical for emerging companies building in this space to consider — cloud vs. on-prem search (while it may feel for all startups that our apps live on the cloud, the reality is that 65% of large enterprise workloads still live on-prem); with new advances in AI / ML, you may get a lot of different recommendations, but it may become hard to sort the false positives, recommendations targeted to a specific function (i.e. customer success or sales); and lastly, data security and privacy, with GDPR and worker concerns around data anonymization.

And as my teammate Michael Yamnitsky points out, he covered many similar companies from 2009–2010 at Forrester in this “personal cloud” space. I’d love to dig in more (and hear from you) why those companies did not take off, and if and why this new wave is primed to do so. With this convergence of search and ML-generated predictive insights…will there still be such a concept as “search” in the future? Or will our operating systems one day be able to preemptively surface us the information we need, right when and where we need it (and our kids will laugh that we ever needed a Google Search bar)?

Emerging Startups

Below are a select number of startups building in this space:

• Journal & Seva — All-in-one unified search platform across your business applications
• Notion — Building a 4 tools-in-1 workspace, across notes & docs, knowledge base, tasks & projects, and spreadsheets & databases
• Station — A wrapper around all of your desktop apps into a single window
• DotAlign — An Outloook plugin that uses your existing email, interaction data, and professional networks for relationship intelligence
• Honey — A knowledge hub rethinking the intranet, with easy-to-share links, content, and sharing.
• Worklytics — A YC-company that connects all productivity tools, and provides real-time analysis on org-wide engagement and collaboration
• Slite — Has raised $4.4M, with investors Index Ventures and YC, building a platform for teams to centralize content.
• Slab — Has raised $2.2M, with investors NEA, Charles River Ventures and Matrix Partners, building a modern wiki.
• Guru — Has raised $15.7M from Firstmark Capital and Emergence Capital, with a browser extension that surfaces information when/where customer service agents and sales reps need it.
• Swiftype — An enterprise search platform, acquired by Elastic
• Skipflag — An enterprise knowledge base, acquired by Workday

If you’re thinking about the future of work, I’d love to talk.