AI makes it faster and cheaper to test a new venture. That shifts the critical question from whether you can build a product to whether you can win customers.
A few years back I wrote an article entitled “The Rise of QA.” I’ve said subsequently it could have equally been called “The Rise of Product Management.” The thesis is that the knowledge of how to write code was not the bottleneck, but rather the understanding of what the code needed to do. I stand by that, although for clarification, it’s not simply the writing of lines of code that matters (i.e., where to put the semicolon so the code compiles), but more importantly how the code should function, especially around edge cases, scaling, security, etc. In the same vein, this article is going to cover why sales has often become the critical and rate-limiting function for new ventures.
The examples are SaaS oriented, but the lessons apply to most businesses. I’ll note how the takeaways apply to other domains. Also, as we’ll see, “sales” should be taken broadly to include adjacent functional areas that serve a similar purpose.
While software has existed for decades, it was in the late 1990s that the industry took off. Back then, to launch a startup, you typically had to go to a VC to raise millions of dollars to hire up a team.
It’s important to understand VCs are some of the most risk-averse people you will ever meet. They wait as long as possible to bet on as sure a thing as they can. To be fair, every business wants to de-risk every project and venture, but, in my experience, VCs as a whole tend to be more timid than almost any other industry or functional area I’ve seen.
In the 1990s and early 2000s there was only so much risk that could be removed. VCs providing the first round funding of a startup had to take on three risks. First, would this market exist as suggested and would this product be well received (market risk)? Second, were the founders generally knowledgeable and competent (founder risk)? And third, would they be able to take the concept and deliver an MVP (initial execution risk)?
To the third point, for many tech companies of the era, the initial fund went to hiring engineers to build out a minimum viable product (MVP). Plenty of companies spent millions but missed the market (e.g., by omitting key features or building something customers didn’t want), faced delays or cost overruns, or simply failed to deliver an MVP.
Today, the VC refrain is “any idea can be built in a few weeks.” They’re not wrong that even non-technical people can build MVP software products in a matter of weeks. From a VC perspective, that removed much of the high-cost part of early execution risk. It should be noted that oftentimes these MVPs are functional, but may have security risks, scaling issues, etc. Still, they’re enough to get some early customers. That still leaves the other two risks.
With the risk of initial software construction significantly reduced, VCs now focus on the other two categories. The second is a pain, but in the VC mentality (arguably correctly), people can be replaced or supplemented. The looming issue is whether the product will be received by the market: will anyone buy it?
This now becomes the limiting factor. Who in the company solves that problem? Sales. Today, even seed-stage VCs ask, “What’s your revenue?” The PowerPoint deck that got funded twenty years ago is not enough. The proof of concept is not enough. VCs fund sales.
This moves the key person in the company from the CTO, who could get the company over the first hurdle by delivering software, to the salesperson who can deliver the first sales necessary for securing funding. (Disclaimer: I’m a CTO.) That may be someone with a sales title or it may be one of the founders; the point is about the task, not the official job title.
This is because in business there is only one metric at the end of the day: revenue. Certain aspects of a business may or may not be working well. If they are not, that can decrease sales. A company with weak or declining sales probably has a problem with someone in the company, be it product, marketing, sales, engineering, operations, or elsewhere. However, strong sales don’t necessarily imply everything is working well, just that there’s no problem so significant it’s visibly hurting sales.
There are some caveats. Simply hiring a good salesperson doesn’t magically make sales work. As noted above, the wrong product, marketing support, tech issues, or other problems can undermine sales. Even if those all work, if you, the product champion (often a founder at a startup) don’t know how to sell the product, you can’t just hire someone who doesn’t know the product as well and expect them to sell it.
While the argument is about the importance of sales, sales can come in many forms. Sales-driven products require inside or outside sales reps to contact customers and sell the product. Marketing-driven products have little or no sales and are driven by marketing, not salespeople. For example, very few people buy a house just by seeing a sales sheet; instead, they talk to the selling agent. On the other hand, you pick a drink like soda based on marketing; there’s no salesperson in the store walking you through the options.
At still other companies this might be more business development. At some companies that’s for partnerships, including channel sales. At others, the line blurs between business development and sales.
Even customer success (or under similar names like account management) may be crucial for upselling or retaining customers. Forward deployed engineers have become popular at many software companies because they help deliver the product to get value. So while it is sales that is now key, the form of it (and title of the role) can vary. More broadly, it’s any activity that can attract and retain customers, but while every role can affect sales in some way, it’s sales-focused roles that have the most impact.
The example above was for a software startup. It applies to internal ventures as well. You may not be pitching to VCs, but the project sponsor will want to see results. Gone are the year-long projects that may quietly (or not-so-quietly) go off-track. Companies will want to see results quickly and that means building the MVP quickly and then seeing if the market will bite. Sales answers that question.
There are a few exceptions that should be noted. Companies with massive tech debt still can’t move nimbly. They’ll need to pay down the tech debt before they can quickly deliver if it’s about updating a legacy product. Also, highly regulated industries will have other gating factors, e.g. passing FDA clinical trials.
Software is the poster child of AI’s efficiency (although it is somewhat overhyped at times). Operations, marketing, and other teams are also being automated allowing less time to deliver. For example, first-pass marketing collateral, from website to pitch decks, to brochures and images can now be made by someone with limited experience using an LLM. As with the software, it may have issues, but it’s good enough for the MVP stage. Hooking up AI agents and chatbots to a knowledge base provides first-level account support. And in the past few decades HR and finance can easily be handled by business process outsourcing services. All this takes us to the open question of sales.
This applies to non-SaaS companies, too. Large consulting firms like McKinsey would have armies of analysts on a project. What you bought was “the process,” how they approached the problem, and the quality of the people (firms like McKinsey often hired from top schools). What used to take twenty people at such a firm may soon be done by five. With hundreds of mini-consulting firms (companies of one to fifty people in size) all armed with the same LLMs with the same process and ability to parse the same information, what’s the differentiating factor? Sales (with the support of marketing in some cases).
In this case it comes down to the fact that AI-backed knowledge work will be a bit commoditized. This has been the case with law firms for years. All law firms have access to the same statutes and same prior rulings. The choice to use one law firm over another when both have expertise in the same area and charge roughly the same is a question of which firm you trust more. There will still be product differentiation, but the more a product or service relies on the same underlying LLMs, the more trust, stemming from sales and marketing, will matter.
There’s a lurking risk in all this. VCs will now focus on sales as proof of concept. However, companies have gotten wise. We’ve seen circular financing writ large with AI companies (see “What Are Circular Deals?”). It happens on a small scale, too. I recently heard of a startup whose customers were all long-time friends of the founder. The hapless salesperson showed up to grow sales only to learn they were a mirage. This is nothing new; years ago one of my clients double-counted some revenue (internally), making the business unit look more successful than it was.
Sales revenue seems like a simple yardstick. However, wiser VCs and potential employees need to dig deeper.
Years ago, MVP execution was the big risk on a new venture, be it a startup or company initiative. Did you have the right knowledge and experience to navigate a project and get it to the market on time and on budget? LLMs have significantly reduced that risk in many industries (although not removed it entirely).
This has now shifted the risk question from “Can you build it?” to “Will anyone buy it?” It falls to sales to answer that question, putting them in the driver's seat. Be warned though, as I discuss in my next article “The Decline of Sales,” LLMs will alter the nature of sales in the future.
It’s critical to learn about corporate culture before you accept a job offer but it can be awkward to raise such questions. Learn what to ask and how to ask it to avoid landing yourself in a bad situation.
Investing just a few hours per year will help you focus and advance in your career.
Groups with a high barrier to entry and high trust are often the most valuable groups to join.
