InsideSales Unveils Collective Intelligence Concept at Accelerate NYC
To InsideSales’ chairman, CEO, and founder, Dave Elkington, there’s no underestimating the importance of machine learning and AI. “There are winners in sales transformation and there are losers. The winners, we believe, are using data and AI,” he told the audience at the company’s Accelerate NYC event this week. But to Elkington, the key component of machine learning and AI is not the algorithm but the data. “The math has been around for over 60 years,” he said, citing AI innovator Arthur Samuel, who became known for his work in computer checkers. “What’s different today than 60 years ago? It’s data.”
He then went on to introduce the concept of “collective intelligence,” which he says leverages a “co-op of data” to “allow you to learn from the network, learn from everybody at scale.” He cites GPS navigation app Waze as an example. “Waze shows me how many Wazers—people who are participating in the co-op of data—are within [for example] a five-mile radius of me. There’s a Wazer that is two minutes ahead of you, there’s a speed trap, slow down, and there’s construction up on the left. It’s going to recommend a reroute based off of that. The idea is it’s actually learning from the network; it’s learning from the people around you.” Compared to other navigation apps, he said, “the difference is collective intelligence.”
InsideSales’ chief strategy officer and chief customer officer, Dave Boyce, and its chief operating officer, Chris Harrington, elaborated on the concept of collective intelligence in an interview with CRM magazine.
For Boyce, today’s enterprise software architecture hampers collective intelligence. “If you were to rebuild enterprise software architectures today, you’d build them more like the consumer Internet,” he says. “In our personal lives we expect the cloud to know about us—I expect it to know about someone else like me, so [for example] when Netflix recommends movies to me, it’s because it knows people like me might like that, and when Amazon recommends products to me, it’s people who bought X also bought Y. I expect it to know where I fit into the universe, and based on my patterns and everybody else’s patterns start to steer me to things that are going to be useful. Enterprise data architecture doesn’t allow for that. There’s only a certain amount that you can do by analyzing your own history. If you’re constrained to what’s within the four walls of your own personal history, you can analyze and turn over lots of stones [regarding] what happened in the past, but you don’t know what the best thing is that could happen in the future because you can’t see any other possibilities other than what you’ve already experienced—unlike the consumer Internet, which is structured in a way where we benefit from the collective intelligence of everyone out there experiencing things.”
He goes on to say that this structure is hard to undo because of existing privacy agreements. “Once you’ve made those promises, I will not open or look at your data, it’s hard to unmake them. If you had to do it over again, you would structure it in a way where everything is protected and anonymized, but it is comingled, normalized, analyzed, and used for the collective good,” he says.
He cites the Open Data Initiative between Adobe, Microsoft, and SAP as an example of this. “That’s a valiant attempt—let me declare a canonical data model that everyone can subscribe to and once I do then I can contribute data, it can be contextualized, and then I can get contextualized insights back. I think that’s great—if you think about Adobe as marketing and SAP CRM as sales and Microsoft as the productivity suite, conceptually, that’s amazing. The problem is for anyone to participate in that, they have to proactively opt in, and if a customer has already signed an agreement [for example] with Microsoft that says, ‘Thank you for promising hands off my data,’ now they have to go un-sign that agreement so that they can contribute data to that data co-op.”
For this reason, he asserts that it is better “to begin with the end in mind” with respect to data rights. “You have to architect it from the beginning in the way that you want to go. MapQuest couldn’t have gone back and Waze-ified their offering because it was already out there and already had user agreements, but Waze could come in and build it from scratch,” he says. “The benefit of Waze is not that they know better where the roads are—they had an agreement with all the users when they signed up, and the agreement is that they’ll contribute a tiny fraction of value and get a whole bunch of value back from the collective intelligence.”
Harrington asserts that collective intelligence is challenging even within organizations, saying that “being able to truly effectively correlate marketing, sales, customer service, and financial data within your own walls is a challenge because it is protected, it is siloed, and it’s designed that way.” He adds that often, departments will be hesitant to share their data: “‘I don’t want you to have access to it because you’re in marketing and I’m in sales’—that’s truly what happens out there because too often today we use data not only as a currency, but as a proof point for our business. You end up with your analysts who are your data gunslingers, and I have mine, and we’re each trying to prove our viewpoint as opposed to building the point.”
He adds that the transition to using data in a collective way “is going to be even more complicated than the transition from software to SaaS.”