Enterprises aren't buying AI tech

Startups should become an expert in applying AI to one industry vertical

A desperate and disheveled AI salesman can’t get customers to pay attention to him.

The problem

A lot of B2B AI startups are having problems selling into the enterprise. One venture capitalist’s view is here.

HUGE problem for startups selling AI products to enterprises - buyers have no idea what is going on you think evals are a problem in research, but way more so in the real world: noisy, snake oil marketing, top-down pressure, lack of talent, focus on credibility as safe signal

Many of these startups are pursuing a horizontal customer acquisition strategy, in which they view all enterprises as potential customers. I think that pursuing a horizontal customer acquisition strategy is likely to fail. Selling AI to enterprise requires a more calculated, vertical approach, in which you get close to customers in one particular niche, become the recognized AI expert for that particular niche, and use AI to solve one specific and concrete problem for that niche. Of course once you acquire a steady pipeline of customers, then you can expand to adjacent niches. Land and expand, as the saying goes.

Let me unpack this argument a bit. A lot of B2B SaaS companies, such as Salesforce or Hubspot or Gong, have pursued a horizontal customer acquisition strategy. They pursue customers who operate in any vertical. And, to some extent this makes sense: all enterprises need CRMs (Salesforce or Hubspot) or revenue intelligence solutions (Gong). And, yes, Salesforce was one of the first¹ companies to adopt a SaaS model, and so it was revolutionary when Marc Benioff first started hunting around for customers. But—and this is key—Salesforce sold standard, deterministic software, which solved a discrete and concrete problem for companies.

AI isn’t like that. According to the effective accelerationists, AI is a general and generalizable technology which will change everything everywhere all at once. In their mind—and that of the salivating venture capitalists who see billions in profits to be had—this means that any and all enterprises need only plug AI into their operations, and…well, something will happen. AI, in this narrative, is magical pixie dust that an enterprise merely has to sprinkle around.

Back here on Earth, the people who operate the enterprises to whom these startups want to sell their AI pixie dust must contend with more quotidian concerns. They have demanding customers, hungry competitors, limited time, and little to no AI-native talent. They don’t have the resources or knowledge to evaluate the claims being made by enterprise AI sales teams. And, if it ain’t broke don’t fix it holds a lot of time-tested wisdom.

Here’s how Matt Beane describes the situation:

Ask around, and this is still mostly what you’ll find. Leaders at companies might have heard of AI, but they don’t know much about it other than what they overhear as their kids “cheat” on their homework. But - and I mean this seriously - that’s the beginning and the end of their exposure. And their decision making isn’t going to change. If you’re Marcus, the guy (aka research subject) I know who owns and runs a mom-and-pop third party order fulfillment company on the eastern seaboard, you saw a Super Bowl commercial about AI, and you kind of shrugged and went back to trying to get orders, manage costs, and keep your warehouses productive for your customers. It’s a brutal business, and the only time you adopt a disruptive new technology is… never. That’s right. Never. You just don’t. Your margins are razor thin, your customers are hyper-demanding, and your workforce is a dynamic problem. If someone can sell you some technology to solve a real problem, reliably, starting yesterday, you’ll buy it, but the case has to be crystal clear and even then they have to call three times.

Now, you may read that, and think to yourself, But I’m not selling to mom-and-pop third party order fulfillment companies! And, fine, sure, maybe you’re selling to larg(er) enterprises that have hundreds or thousands of employees. You’re a real B2B AI company, selling to real companies with real revenues and, importantly, you think, the cash to pay for your amazing AI pixie dust. But none of this matters if your target customer can’t make any sense of what you and a dozen of your competitors are selling.

This is why I think that AI startups selling AI solutions need to pick a particular industry vertical, get close to the companies that operate in that vertical, and solve one concrete and discrete problem for those customers. Become the recognized industry expert in applying AI to that industry, and that industry’s companies will beat a path to your door. AI is no doubt great technology, and, as the saying goes, today’s AI is the worst you’ll ever use. However that by itself does not move the needle for the enterprise. They need a reason to buy, and that reason has to solve a problem for them, and do so in a way that doesn’t threaten their job. No one ever got fired for buying IBM is a saying for a reason.

The solution

Here’s a more detailed breakdown of the argument I’m making, and the rationale behind adopting a vertical strategy:

1. Complexity and ambiguity: AI solutions are often marketed as general-purpose tools, which creates ambiguity for potential buyers who lack deep technical knowledge. Without understanding the precise value that AI brings to their business, decision-makers struggle to justify the investment.

2. Overpromising and snake oil marketing: Due to overhyped marketing, enterprises may lump all AI solutions into the same category of unverfied tech. This makes it difficult for them to distinguish genuine products from those offering little value.

3. Top-down pressure: Decision-makers often feel pressure from executives or the market to adopt AI due to its buzzword status, even if it’s unclear how these tools will impact their specific industry.

4. Niche expertise: By focusing on a particular vertical, AI startups can become recognized experts in solving that industry’s problems. This focus allows them to tailor their product and messaging to directly address common pain points.

5. Building credibility: When a startup demonstrates successs in solving concrete problems within a niche, it can leverage this credibility to expand horizontally into adjacent markets, following the “land and expand” strategy.

6. Customer-centric approach: Understanding a vertical’s unique challenges enables startups to align their products with the strategic goals of their customers, making it easier to justify the purchase.

1. Silicon Valley lore asserts that Salesforce was the first company to sell software as a service, but this isn’t really true. Bloomberg sold software subscriptions long before Salesforce was a mote in Benioff’s eye, and other software like AOL, and Compuserve and GEnie before it, were also software services sold as subscriptions. Interestingly, ChatGPT parrots the Silicon Valley line.