We are entering a new era of technology in which people will interact with computing across multiple interfaces — tapping, swiping and talking to phones, machines and disembodied virtual assistants — without skipping a beat. Not every application vendor that's dominant today will survive this transition.
Others will come from nowhere to lead this next generation of computing. One of the visionary leaders you may be hearing more from as this trend takes hold is Chuck Ganapathi, founder and CEO of Tact, a company that Microsoft brought onto its keynote stage for a demonstration during the opening session of Build 17 last month. A month earlier, Tact participated in Oracle's customer experience show. Salesforce has been a close partner since its earliest days.
These established giants of enterprise applications are scrambling to partner with Tact because it has something they don't. So when Ganapathi got in touch to suggest a phone briefing, I seized the opportunity to find out more about his plans to shake up the direction of enterprise applications. I wasn't disappointed.
As Ganapathi explains:
The shift we're seeing now in how humans use technology and interact with machines is unparalleled since the time we had the Windows GUI come out.
This is all prehistory now, especially to millennials, but we older folk still recall how transformative the advent of Microsoft Windows in the early 1990s turned out to be for the application software industry. An entire generation of early software leaders — Lotus, WordPerfect, Ashton-Tate — were wiped out because they failed to navigate the transition from the old green-screen, command-line interface of the original PC operating system to the graphical user interface (GUI) of Windows.
Transitions in the underlying computing have had similar effects. In the enterprise software world, SAP and Oracle were once the young upstarts challenging established mainframe software giants whose names are now forgotten (Cullinet, McCormack & Dodge and Management Science America, since you ask). The advent of cloud computing has created a new generation of challengers, led by Salesforce, Workday and ServiceNow.
Imagine then how transformative it would be to have both kinds of transition taking place simultaneously — a paradigm change in how we interact with computers that coincides with an equally dramatic upheaval in the underlying architecture of the applications themselves. That's what Ganapathi believes is in progress.
He's witnessed all of the prior changes, having started out in mainframe applications before moving to CRM pioneer Siebel and then joining Salesforce in 2005. But even though Salesforce dramatically changed how IT bought and operated CRM, he points out that the user interface hardly changed at all between 1999 and 2012. The fundamental application architecture is still the same today as decades ago, because it's anchored around owning and providing access to the data — essentially a database-forms paradigm.
Tact makes a fundamental break with this notion of an enterprise application being a system of record, says Ganapathi.
Because of these edge devices and AI, there's a big fundamental change happening, which is how do humans interact with machines?
We're not a system of record for anything. Rather than owning the data, what if we said we were trying to own the workflow of a particular type of person? We should be thinking about how applications help you in your working life.
With Ganapathi's background, it was natural to target the workflow of a salesperson.
This company was really about, how can we fundamentally rethink and reimagine our relationship with technology — and more specifically, the seller and the technology they use.
If you want to do this right, you can't have a vertical angle of CRM. The only way we're going to be successful as a company is if we fundamentally do two things:
One, get salespeople to want to use this technology, not be forced to; and
Two, build an architecture and a technology vision that starts with the person and then works outwards from there, as opposed to with the technology and more specifically the database.
When Ganapathi founded Tact in 2012, the concept was to create a machine equivalent of the executive assistant role that works alongside every senior vice-president at Salesforce. When Ganapathi hired his first EA, he no longer had to interact directly with his calendar, flight booking, expense filing and so on, and this person understood his habits, concerns and goals. Tact set out to use AI to create the same experience for sales people, says Ganapathi.
What if we stopped thinking about this world as a bunch of apps, and instead with the use of conversational AI, what if we could create the machine equivalent of my EA? But instead of doing it for executives we do it for sales people. If you made your sales people more effective that could hit your topline.
We think about the product as a digital assistant. We don't care about how you interact with the technology, we want to be able to insert the digital assistant into your workflow on all of the devices and interfaces you already use — touch, voice, messaging.
The digital assistant becomes an intermediary that connects the conversational applications into the systems of record that hold the data. The user interacts with Tact via a touch app, or by speaking to a voice interface such as Alexa, through text messaging or at-mentions in Slack. Tact then reaches into traditional applications ranging from email, calendar, Salesforce or Zendesk, to LinkedIn or Google Maps, to advance through workflow, or bring back information.
If all this seems too good to be true, Ganapathi points out that this is not a general-purpose intelligent agent. Focusing on the specific needs of sales people allows Tact to narrow the scope of the AI to make it more effective:
AI, natural language processing, loves constraints. The beauty of our case, because we're very focused on the sales person, we have a lot of context and a limited set of conversations. Because of the constraints and context, the AI is pretty great.
The company's unique technology lies just below the voice recognition or natural language processing (NLP) that's readily available from other platforms. Tact has concentrated on building a synchronization engine that brings the various applications together, along with a graph that "connects all the dots" between the different data sources and workflows. That combines with the team's institutional knowledge of the sales automation space:
We have unique knowledge of what works, what doesn't work. This whole world of AI and natural language is not an established field. What you see as working product is 5% of the effort. The other 95% is the experimentation that didn't work. We've just learned a lot.
The sauce comes from marrying well-understood concepts in NLP to very specific things in the life of sales that a general-purpose platform will not have. What's novel in our patents is taking this innovation in AI and NLP and applying it to this very specific domain.
That unique intellectual property is what's helping Tact form partnerships with many of the leading sales automation and CRM vendors. It's important to partner with all of them, says Ganapathi, so that Tact can be the glue that joins up a sales person's workflow across any combination of underlying applications.
For us to be successful, it's imperative that we position ourselves as Switzerland. We have to be an application-agnostic platform and vendor-agnostic. We can be more successful by being independent.
But don't they remember what happened when Microsoft changed the rules of the user interface and ensured its own applications became dominant in the new paradigm? If Tact succeeds in owning the sales person's workflow, their application partners will have ceded ownership of their own customers to the upstart. When I put this to Ganapathi, he agrees, but goes on to explain why he believes the incumbents are powerless to prevent being overtaken by this shift.
There's classic innovator's dilemma in play here. The core of their DNA comes from owning the data. Everything is based around the data. It is burned into their DNA and their biggest problem is that their own success is their own worst enemy.
Even if they know — they see everything that's happening — what can they do about it? Owning the system of record is their way to greatness. To say that's not the way, it's owning the workflow — the question is, what do I do about it? You have to destroy some of the institutions that made you great. To say HTML has no future in this world is anathema because you're basically saying what made me great is no longer the future.
An important new insight was Ganapathi's characterization of the established paradigm as database-forms. It's a new way of looking at the phenomenon of paper-based processes being baked into the inherent structure of enterprise applications. Before computers came along, every enterprise business process revolved around structured pieces of paper. Despite the best efforts of Michael Hammer to obliterate them, these paper-based processes were simply automated to run on computers, feeding back-end databases.
Those databases aren't going away, but these days there are better ways of feeding them. Tact has a head start on making sense of the AI-powered, conversational interactions that may prove to be the future of enterprise sales automation. It's early days yet, but this is one to keep an eye on.
[Updated — when first published, we incorrectly stated that Salesforce is an investor in Tact. It is an early partner. Tact is backed by Accel Partners, Redpoint Ventures, Upfront Ventures and Microsoft Ventures.]