How to Think About Selling AI Products
Andrew ParkEditorial Lead, Heavybit
How to Actually Sell AI Products in Competitive Markets
AI in today’s news headlines is all about big numbers: High performance scores of models on key benchmarks. Millions in funding raised for billion-dollar startup valuations. But what about sales figures? There’s a great deal of noise being made about how well AI startups are doing, but not much in the way of specifics on how they’re successfully going to market.
EverWorker CRO John Roberts admits that the current sales environment for AI products is fraught: On one hand, organizations do want to realize the mythical productivity gains AI should be offering. On the other, many have been sold a bill of goods; low-rent GPT wrappers that aren’t exactly performant, or inscrutable products that no one bothers to learn how to use.
Below, Roberts outlines his suggestions for succeeding in AI go-to-market:
- Focus on Outcomes: Tech stacks are far less important than identifying valuable results for customers and actually achieving them
- Understand Changing Buyer Psychology: Buyers are no longer afraid of missing out. They’re afraid of f***ing up
- Provide Actual “Technical Support”: Your team’s job isn’t just to convince customers to buy; it’s to get customers successful
- The Strongest Path to Success is Success Itself: The strongest selling signal today is wildly successful and happy customers who won’t stop selling your product for you
Outcomes, Not Tech Features, as a Focus
While Everworker’s promise is providing “AI workers” that fit its customers’ use cases, Roberts suggests his team’s approach focuses less on tools and entirely on outcomes, whether for scaling outbound outreach, building sales playbooks, or running content and SEO programs. For revenue teams, part of the initial approach is diagnosing and triaging breakage points.
“If you can understand what your revenue target is, we simply walk back from there. Okay, that's your target. What about inbound leads? If you have them, how quickly are you following up with them? Industry standards being that if you're not following up within less than five minutes, your conversion rates are going to fall off a cliff. And if you are following up fast enough, what’s your conversion rate to meetings once you actually reach out?”
The Founder’s AI Checklist: Outcomes, Bottlenecks, Starting Small
Roberts outlines a plan that founders should build before buying a new AI product. “There’s definitely a checklist I would put in front of founders [looking to adopt AI products]. What’s most important is understanding the outcomes that they're trying to achieve and the bottlenecks they're facing today. If you don’t know either, it’s very difficult to propose a solution.”
“Compare this to the classic SaaS sales model, where teams buy a tool, then find themselves stuck six months later trying to figure out how to implement it. At EverWorker, I think our competitive advantage today is we actually started as a service-first organization. When we think about an engagement, we're always bringing in forward deployed engineers and business analysts that will actually help build the business case.”
“Make sure that the bottleneck you're thinking about today is actually the bottleneck that should be solved first, and then pass it over to the technology side to build a solution to solve that.”
The CRO also cautions startup founders to avoid going too big, too quickly. “Start small, get a quick win, and use that quick win to build buy-in internally, so that people feel secure that this isn’t something to replace everyone’s jobs.”
“I think that's the biggest concern that we have today: That CEOs might have already bought into the idea of being ‘AI first,’ but you get a level or two down into the organization and someone else on the team might say, ‘Wait! This thing is going to replace my job, so I won’t do it.’ The only way that you get over that is by starting small, seeing the success, letting people inside of the organization own that success, and then scaling from there.”
“What’s most important is understanding the outcomes that [customers are] trying to achieve and the bottlenecks they're facing today. If you don’t know either, it’s very difficult to propose a solution.” -John Roberts, CRO / EverWorker
“Our approach involves walking customers through a very prescriptive process of setting up. The people that were doing the job [manually] today need to be the ones that are auditing the output of the [EverWorker agents] before they give the all-clear to let the agents run free and continue to scale.”
Roberts notes that his team ’s mindset of adopting agents isn’t a “set it and forget it” approach, but rather, training customers to become zookeepers who manage the care and feeding of their agents while vetting outputs and measuring outcomes.
Why Selling Teams Are Becoming More Technical
“If you're focusing just on hiring, you need to have somebody that can understand the business process. That's the really important piece today.” The CRO muses that some of these ideal systems experts may have started at high-end consulting firms but were disillusioned by how corporate the environments were, and wanted to get their hands dirty and accrue deep domain expertise in technology.
“Those are the type of people that can really step in and say, ‘I can manage this agent. I understand the business context. I understand how this integrates and impacts other parts of the business, and I understand the outcome that I'm trying to drive towards. I can put all of that context into an AI worker and get the outcomes much faster without actually having to do the work.’”
Does that mean the future belongs to solutions engineers? “Look at how Microsoft had layoffs recently, and replaced those sales roles with SEs. People want answers. They want somebody that knows how to solve the problem. They don't care about any of the other stuff. Just solve my problem, provide the outcome and then move on.”
Why Selling AI Products Needs a New Playbook
The CRO suggests that today’s AI sellers are focusing too narrowly on the traditional Challenger Model, which focuses less on relationship-building and more on challenging buyers’ status-quo beliefs about software. “No one questions if the future is the place that they want to go. Everybody has bought into the idea that they need to change.”
“Where people are stuck today is part of a different story written by the same folks who wrote Challenger, which they covered in Jolt Effect. There's a lot of research on this, but basically, the fear among buyers is no longer the Challenger-style fear of missing out. It’s FOFU (fear of f***ing up). Everyone is stuck. They're frozen because they don't know what decision to make.”
“You need to figure out how to help customers overcome that fear. And there's a lot of things that you can do to help folks do that. But that's where deals are stalling today in the AI world. It’s why you start smaller, take risk off the table, give [customers] opt-out language, and lean in with services. You make sure that they understand that you're there to support them.”