The problem with “selling the work”
David Peterson

Many startups these days are founded on the premise that, to use Sarah Tavel’s phrasing, they can “sell the work,” rather than software. (See the recent splashy fundraise announcement from 11x.ai as the latest example.)

And, indeed, selling the work has some critical advantages over selling software.

It opens up new vertical opportunities. Despite software eating the world since 2011, there are still countless niches that software has yet to touch. By selling the work, rather than the software, companies are able to wedge into verticals that previously were challenging, or simply unprofitable, to penetrate

It aligns incentives. The shift from seat-based pricing to usage-based pricing has unlocked growth for many software categories because usage-based pricing aligns incentives between the company and the customer. “Selling the work” achieves that same alignment across a whole new set of software categories.

It’s easier for customers to buy. Turns out that it’s easier to sell a piece of completed work (whose price and value can be compared to the cost of the human that is currently completing that task), than a “productivity improvement” (which is just about the best pitch you can usually make with most seat-based SaaS).

But I think the business model might have a fatal flaw.

I was chatting with the founders of an AI-powered company targeting the legal sector last week, and they shared an interesting insight from their time in Big Law. Turns out that pretty much every big customer was showing up to their renewal conversations saying basically the same thing: “I know you’re using AI to automate your work. So I’ll need a 30% discount.”

As you might imagine, the law firm caved. As is true for many service providers, any pricing power they have is based on their brand alone, and that isn’t enough to protect them from the deflationary impact of AI. Most of the benefits their customers imagined they were getting from AI (whether real or not!) got transferred right to their customers.

So what happens next? Is this the new equilibrium in the legal industry? Or is this just a transient dynamic on the way to something radically different?

My bet is on the latter.

The continued competitiveness of open source models suggests to me that most of “the work” itself will eventually become a near-commodity. If that’s the case, then the clearing price for that work will continue to fall. In that world, only business models that don’t rely on making money by selling that work will succeed.

Law firms, themselves, will have to evolve or die. The biggest firms with the best brands will call in consultants to build their own internal tools to automate a lot of this work. Others, realizing they need to change their cost structure entirely, will re-launch as vertically integrated “AI-enabled” law firms that give away the commodity work at cost (essentially zero) as a loss leader for winning the more profitable business that still requires a human touch. And some enterprising founders will realize they can disintermediate the service provider entirely, “sell the work” at cost as a way to win business, and monetize in other ways.

Regardless, it’s hard to see how a startup that is simply “selling the work” within legal will survive. And because most service providers do not have pricing power, I imagine we’ll see this exact scenario play out in pretty much every other service sector that’s being impacted by AI as well.

My point is not to celebrate the (hypothetical) flaw of a popular, new business model. But more to highlight just how disruptive - and deflationary - AI actually is. What seems disruptive now may easily be disrupted soon. What looks like a new equilibrium may simply be a momentary pause as business models evolve. The only thing that’s clear is that it’s nearly impossible to see more than a few steps ahead. It’s a good reminder that this era will reward the founders who are nimble enough to stay upright as the ground continues to shift below their feet.

David

Am I Thinking About AI the Right Way?
Gil shares the four AI themes and questions he's thinking about.

The Venture Apocalypse
The venture world is deeply debating its future, but core principles remain unchanged.

No Sleepwalk to Success
Engineering success in a technical startup.

Revenue Durability in the LLM World
Everything about LLMs seems to make revenue durability more challenging than ever.

More blogs

Germany / Hardware. Quantum Systems raised $100m, led by Notion Capital and Porsche SE, to accelerate internationalization and scale up the production of its Drones.

Germany / Energy. Marvel Fusion raised $70m, led by HV Capital, for its novel approach to “inertial confinement fusion” with cutting-edge laser technology.

UK / SaaS. Raycast raised $30M, led by Atomico, for its command-bar based shortcut platform.

UK / FinTech. Apron raised $30m, led by Zinal Capital, with participation by Tony Fadell (inventor of the iPod) and existing investors, to continue scaling its SMB-focused business payments platform.

Finland / Software. Distance Technologies raised $11.2m, led by GV, for its glasses-free mixed reality app ecosystem targeting the automotive, aerospace, and defense markets.

Austria / SaaS. DeepOpinion raised €11m, led by Red River West and AIpha Intelligence, for its AI-driven disaster claims automation platform.

Israel / TechBio. Enzymit raised $10m, led by Grove Ventures and Khosla Ventures, for its cell-free production platform that aims to make bioproduction faster, simpler, cost-effective, and scalable.

UK / Chemicals. FluRok raised $9.8m, led by BGF, to commercialize its platform that can produce fluorochemicals without using hydrogen fluoride.

ENTERPRISE/TECH NEWS

Will the AI platform players move up the stack into B2B? Sarah Tavel of Benchmark wrote a brilliant and thought-provoking piece on why she believes OpenAI and other foundational model players will be forced to move up the stack. “It seems inevitable that as the underlying foundation models become more powerful, the LLM players will seek to justify the enormous investment that has gone into training their models by moving "up the stack", and evolve from an API or chat interface, to async agents….As an example, imagine a future enterprise buying decision: Why buy a specialized AI application that lets you automate internal IT ticket response, when the foundation model companies offer an AI agent that if you point it in the right direction with a job spec, will read your knowledge base, build its own integrations to connect to your existing systems of record (e.g., Jira), and then handle all the internal requests automatically? Some might laugh at this scenario, but I’d suggest that if you are a B2B founder building an AI-native application, you NEED to do the thought experiment of assuming it over the next 3-5 years as you consider the strategy for your company. Not just because of the risk of this scenario happening, but because any progress down the path of the scenario will meaningfully increase competition for your company (as I describe in #3 below). So how do you future proof your B2B AI application layer company?” To stay relevant and valuable, she argues, a software business will need to rely on at least one of three strategies: (1) a network effort, which she believes will be more important these days than ever; (2) some proprietary data; or (3) killer execution in an overlooked market. The piece is controversial, but it’s well worth a close read.

The race to beat OpenAI. As if in response to Sarah’s piece, the Information covered wrote about how traditional SaaS companies are racing to deploy agents. “In the last month, Salesforce, Microsoft, Workday and ServiceNow—hoping to head off similar products from OpenAI—have unveiled products they said will let workers automate ordinary business tasks, such as placing an e-commerce order or scheduling a sales meeting. But the companies will have to overcome skepticism among customers before the agents can take off.” The reason for the rush? “ServiceNow, Salesforce and Microsoft may believe they have little option but to jump into the agent market early. OpenAI has its own plans to launch even more capable agents that can decide for themselves how best to carry out tasks in a web browser or application, like booking travel or transferring data to a spreadsheet. OpenAI CEO Sam Altman has said OpenAI’s technology will soon let organizations automate work currently done by large swaths of employees.” Maybe Sarah has a point.

The SaaS era is over. Sam Lessin struck again with his unique ability to craft pithy, accurate, and radical takes on the state of technology and venture. On Twitter, he declared the “SaaS is dead” He made a forceful argument that as software rapidly commoditizes, the future of software will often involve buying operating assets in non-software businesses and leveraging software to drive increases in efficiency and value. You can read the whole argument on Twitter.

Patience is a virtue. Aileen Lee wrote a very thoughtful and important piece this week on why - particularly at a moment of technological inflection - it may take time for the winners to emerge. Even more so, the winning strategies for both tech companies and venture capitalists. “It’s been almost two years since chatGPT was released. This means the stage is set for long-term successful AI-native companies to rise over time. In the web and mobile, many “first movers” who attracted early funding, revenue and adoption were surpassed by later entrants. Could there be 1st mover disadvantages in big platform shifts? And, will some of today’s frontrunners in retrospect become the Netscape, MySpace, Chemdex or Motorola of the AI boom?” Later in the piece she writes, “if you’ve been thinking it’s too late to start an AI company given the hype, or that it’s too late to join an AI-native company - it’s not. History indicates we’ve got decades of exciting innovation ahead. And, today’s leaders may just be paving the road for other companies to take the wheel. Long term success will play out in the coming decade(s).”

HOW TO STARTUP

AI companies are growing faster. The FT published data suggesting that AI companies are generating revenue at a very rapid pace compared to previous generations of companies. “The AI start-ups in the cohort took a median 11 months to hit $1mn in annualized revenue after their first sales on Stripe, compared with 15 months for the previous generation of SaaS companies, the data showed. AI start-ups that have scaled to more than $30mn in annualized revenue achieved the milestone in 20 months — five times faster than past SaaS companies.”

Precocious unicorns. According to an interesting dataset published by SVB, AI unicorns seem to be minted earlier than their non-AI peers. This is hardly surprising given the heady valuations many of these companies are achieving relatively quickly. More concerning, however, is a graph that shows that around 30% of all unicorns are “unprofitable and shrinking,” around double the average of recent years. “While most VC-backed tech unicorns are well capitalized and have seen improvements in profitability, a growing minority are unprofitable and face declining year-over-year sales.”

One founder’s journey. The story of how Ofer Ben-Noon found inspiration from his mother, and mentorship from a tech legend which combined to lead him to two exits totalling over $1B in combined value. “Argus, his first startup, was founded by Ben-Noon in 2013 with two friends from Unit 8200, Oron Lavi and Yaron Galula. "At first, we spent about ten months working on cyber protection for smartphones, but it wasn’t going well—we faced technological challenges and weren’t sure the world needed our product, as there were already several startups in that space. Zohar Zisapel wasn’t enthusiastic about it either. He and my mother had just returned from a trip to the East, where they met people from tech companies. He called me and said, 'You need to focus on cyber protection for cars—that’s the future: connected cars and smart cities.' "At that time, no car company had SIM cards in vehicles, and only General Motors had cellular modems. But Zohar saw two steps ahead and said, 'When cars are connected to the internet, they’ll need cyber protection.' I wasn’t sure at first, but he told me to investigate. He even connected me with General Motors Israel and Gil Agmon (Delek Automotive Systems), which I couldn’t have done without him." Zisapel’s advice and connections paid off—along with his initial investment, which earned him an 18% stake. Within four years, Argus developed numerous patents, raised $30 million, and was sold to the German automotive giant Continental for $450 million, 15 times the original investment. Argus remained an autonomous company within Continental, and Ben-Noon continued to manage it until 2020, when he felt ready to move on to a new venture in digital health.”

HOW TO VENTURE

Getting precise about artisanal venture. Samir Kaji tweeted that it’s no longer accurate to think of the majority of VC firms as artisanal shops. “The definition of venture capital (VC) has evolved from being artisanal, early-stage financing to a broad spectrum ranging from inception to pre-IPO. Essentially, VC now represents minority investing in private technology and life science companies. The VC landscape has become as fragmented as the buyout space, now encompassing small, mid, and large-cap segments. Each of these has distinct risk/return characteristics. Smaller funds offer substantial cash-on-cash upside but come with significant volatility. In contrast, large, reputable brands are increasingly akin to private equity (PE) risk/return. Small funds represent the potential for venture alpha (with associated risk), and large funds offer venture beta. Both can play a role in a diversified portfolio.”

Firebolt’s CEO Eldad Farkash joined Judy Khan Shaw on NYSE FloorTalk to share how Firebolt is transforming the cloud data warehousing space and helping companies solve data processing challenges.

CruxOCM’s CEO Vicki Knott is disrupting the status quo of one of the oldest “boys clubs”, the oil and gas industry, pushing for better safety standards and company values that allow workers to thrive. “In an industry made up of only 15% women, as a female CEO, Knott has been a “misfit” in her own right—something she takes pride in. Recognizing this industry won’t be able to make real diversity, equity, and inclusion (DEI) strides without a fundamental shift in culture, she’s taking a wholly different approach to leadership than many of her peers.”

Aquant
Senior Enterprise Account Executive (New York)

CruxOCM
Engineering Manager (Remote)

SOURCIX
Data Scientist (Tel Aviv)

Viably
VP Marketing (Raleigh)

More jobs