Content marketing platform company Uberflip was recently on the verge of making an acquisition in the artificial intelligence space, before concluding it needed to develop its own capabilities, its CEO told the TechWeek Toronto audience on Thursday.
Speaking in a session on “data intelligence,” Uberflip co-founder Yoav Schwartz did not disclose the name of the firm his company had considered acquiring, other than that it was a machine learning startup.
“We realized we no resources internally that could analyze the validity of their software. That’s when we took a step back and decided we need that level of sophistication internally,” he said.
This was one of the key topics in the TechWeek Toronto panel: to what extent can AI capabilities be farmed out to a third party, and to what extent it needs to be home-grown. For Uberflip, which makes software to help marketers deliver content across portals and other channels, the answer has become clear.
“(AI tools) are really the ingredients to the salad that we make,” Schwartz said. “It’s absolutely critical to have that in-house. No one outside is going to know enough about your business to know what needs to be in your salad.”
The problem is that many B2B startups are small, and aren’t necessarily in a position to hire a room full of PhDs to put AI and data science into action. Consider Vantage, an e-commerce company based in Toronto which tries to help merchants boost their sales and marketing success on platforms such as Shopify and BigCommerce. According to Brandon Kane, the firm’s CTO, Vantage employs eight people and one data engineer, even though it boasts more than 21,000 customers and manages 65 million unique customer profiles, including order histories and page views.
“We also have one customer success person. That means a lot of the analysis has to be done by a machine in order for it to work at all. It’s a question of scale,” he said, adding that many of the underlying algorithms around machine learning are now available through open source, which means it’s not a case of reinventing AI or data intelligence.
“You do have to have a story around AI and machine learning. Without it, we would just be another way to create Facebook ads, which wouldn’t be great,” he said.
Even firms such as Ratehub, which deal with consumers, need to think about leveraging AI and analytics that brings in data from other businesses. Alyssa Furtado, the firm’s CEO and co-founder, said the company is particularly interested in using data intelligence to better understand why banks decline 60% of the credit applications RateHub receives. Financial institutions aren’t typically forthcoming about their underwriting criteria, she added, but analyzing trends from its customer base may yield some clues.
U.S. competitors to RateHub such as CreditKarma and NerdWallet are already assembling sizable teams to take on such tasks, Furtado added, but its Canadian competitors haven’t yet.
“It’s an opportunity for us now, the quicker we’re able to adopt and invest in (analytics and AI),” she said. “The first step for us is making sure we have the resources in place to understand conversion metrics, then getting people using and accessing the data.”
Ratehub has toyed with hiring someone at the director level who could oversee a program versus an analyst who could executive, and settled on a more senior analyst who will work with a consultant as needed, Furtado said. Uberflip decided to hire a PhD, Schwartz said, but they needed to have an engineering background.
“As great as someone could be with the data, if they couldn’t action it immediately we and had to train our engineering team there would be no point,” he said. The company is now hiring its second data scientist.
Latest posts by Shane Schick (see all)
- Why are COOs the forgotten B2B marketing target persona? - November 19, 2019
- ActiveCampaign VP predicts its machine learning features will offer an alternative to A/B testing in e-mail marketing - November 15, 2019
- InsideView product exec says Data Integrity will address B2B firms’ cleansing and visualization needs - November 14, 2019