For VCs, Artificial Intelligence is Table Stakes

A speech bubble emerges from Darth Vader’s helmet, announcing Klue’s three alliterative tenets: “Fast, Flowy, Friendly”.

In the company’s brightly-lit Vancouver office, a cut out of the helmet – along with its surprisingly unmenacing utterance – is pinned on the wall a step away from a window sill holding Geoffrey Moore’s Crossing the Chasm, the classic business book outlining the model of innovation adoption.

Both the book and the character – part man, part machine that he is – are inadvertently appropriate for the company that just raised $4 million to apply machine learning to the collection, curation, and distribution of business intelligence for enterprise.

Machine learning is a branch of artificial intelligence, which may have now become the most overused term in the startup universe, as ever-more ventures add dot-ai to their names and craft investment pitches touting their use of learning algorithms.

When Jason Smith, the CEO of Klue, did the rounds presenting his vision to investors, he took a more grounded approach to communicating about the technology.

Smith says, “The relief that I saw wash over some of the VCs when I said, ‘Look, we know the hype is above the reality, and here’s how we’re addressing it today and how we expect to do so in the future. It just completely level set the conversation in a way that now they were open to how we were advancing it.”

If you’re building a tech company today and looking for funding, AI is a huge deal – but it’s not differentiation.  In fact, it’s become the cost of entry.

John Ruffolo, CEO of OMERS Ventures which led the investment in Klue, says, “We’re at the point – the way we evaluate it – [AI] is table stakes. When we are looking at a pitch and we ask, so how are you using AI, and they say, ‘Well, you know, we’re thinking about it,’ to be honest, it’s a big strike.  Why are you not using it would be the big question.”

He’s not alone in this thinking.

“In a few years,” starts Frank Chen’s blog post, “no investors are going to be looking for AI startups.”  Chen, a partner at Andreessen Horowitz, continues, “no investor will be funding startups calling themselves AI-powered startups…because investors will assume the startup is using the best available AI techniques to solve the problem they are solving.”

In many investors’ minds, the term ‘AI’ has become the equivalent of ‘mobile-first’ or ‘cloud-native’.  In other words, it’s not compelling.  It’s required.

“Every single company that we look to invest in must imbed AI, because if they don’t, they will be a high cost producer, and they will be unsustainable,” Ruffolo says. “If you’re not thinking about AI this very moment, you will be in the very near term, and you will get blown out because [your competitors] will be able to get greater insights, greater speeds, and lower costs.  Then my question will be: ‘How do you think you’re going to compete effectively?’”

Peter Misek, partner at BDC IT Venture Fund, which also invested in Klue, agrees: “There’s not a single one of our companies that we haven’t pushed to rethink their businesses in that lens.  Not one.”

Artificial intelligence is enabling cost reductions from the removal of human intervention, the ability to ‘make predictions cheap’ (as Rotman School of Management professor Ajay Agrawal has explained), and the potential to discover new revenue opportunities.

Applications and examples related to all three abound.  Chatbots, which simulate humans in conversations with customers, are expected to save companies $8 billion per year by 2022, according to Juniper Research.  In its August earnings call, Shopify credited machine learning when they announced that they “nearly doubled the number of capital advances in Q2 over Q1” in its Merchant Cash Advance program.  Scotiabank has partnered Toronto-based AI software company Layer 6 AI to create a product recommendation engine that personalizes marketing of its products to holders of Cineplex Scene loyalty cards.

These ubiquitous headlines are informing founders, which often place the technology at the center of their pitches.  Ruffolo says, “It’s a little bit like the whole dot-com days where everybody decided to put the word .com to tell you they were on the web, putting on the ‘.ai’.”

But he asks, “Are you really an AI company that’s selling AI or are you just leveraging AI?”

In his post, Chen says, “nearly all the startups we see these days are dressing themselves in AI clothes (powered by machine learning! uses the latest RNNs and GANs! post-deep learning architecture!)”

Very shortly, he writes, “software without AI will be unthinkable.”

When Klue’s Smith presented his company to investors, he was prosaic about the technology, and “it felt more honest and real, and at the end of the day, I think VCs appreciate the honesty.”

But, he’s quick to say, while the focus on AI in the pitch should be measured, the technology had still better make an appearance.

“There should not be a startup out there that does not have some sort of machine learning aspiration, if not technology,” says Smith. “It’s a must.”