The next wave of software companies sell AI services, not software. The unit of work is not a session, where AI makes a knowledge worker more productive, but a transaction or workflow: AI services software runs 24/7, is endlessly scalable and carries out tasks as they are queued up, with human intervention reserved for edge cases.

At Heron, we’ve found the biggest implication of this shift is who we sell to, and who should sell.

AI Services sells to smaller businesses and more senior buyers

There are two fundamental differences between selling B2B SaaS and selling AI services:

Firstly, the average US business spends 6-10x more on labour, outsourced or in-house, as it does on software. So, a company of a given size will spend 6-10x more on an AI-based service than on a B2B SaaS product. Put differently, to make an AI services sale of a given size, you may be interacting with a company up to 10x smaller than what would have been normal in B2B SaaS.

Secondly, buying AI services has much more far-reaching implications across the company than buying traditional software. Take the example of Heron: We sell workflow automation software that fully automates the intake of new applications for business lenders or insurers, allowing them to respond to an application in minutes, for any volume. Buyers have to reason through four different considerations:

1. Will this enable me to stop buying, or significantly downsize, outsourced labour spend?
2. If the work is currently done in-house, do I have more valuable tasks in other functions that I can direct on-shore employees towards?
3. How elastic is my distribution? Many customers constrain distribution not to overwhelm their Underwriting/Operations department: If that constraint is gone, how much more business could I accept?
4. How many more offers will end up as closed-won deals when I reply with offers in minutes or hours, not days?

All of these are cross-functional, C-level decisions.

Different business & buyer leads to a different sales motion

Because labour budgets are far larger than software budgets and cross-functional decisions are needed, the sales motion for a 60k ACV sale for an AI service looks fundamentally different to the same sale for a B2B SaaS company:

Selling $60k ACV…	B2B SaaS Sale	AI Services Sale
Company Size	>100 employees, >25M Revenue	30-50 employees, 5-10M Revenue
Buyer Profile	VP, Product Manager	CEO, Founder, Owner-Operator
Pain	Known business problem - buy vs. build decision; if buy - gather vendors	New solution to a pain that previously had no tech solution
Sales Cycle	3-4 months	1-2 months
Sales Motion	Bake-off against known competitors. Feature sale, slide decks, run MEDPICC	Strategic decision from CEO. Trial to ensure AI service matches human quality.
Close rates from demo	20-25%	>35% - new class of solution increase close rates; strategic decision required lowers it

Folks from non-sales backgrounds may fit this sale better than B2B SaaS sellers

This means that the type of person that succeeds as an AI services seller needs a different skill set and inclination than a traditional B2B SaaS seller. A B2B SaaS seller that regularly achieves quota definitely can switch to AI services and succeed, but it’s not guaranteed, and someone with a non-traditional background may outperform.

Compared to selling B2B SaaS, success in selling AI services requires sellers to:

1. Be able to hold their own with the founder/CEO of a (smaller) company. An owner-operator will not accept hand-waving on details; expects a directly comparable case study on the fly, and expects the seller to follow their own strategic logic in real time. If a seller can keep up, the buying decision can happen within that one conversation.
2. Comfort outside a corporate environment: A 50-person company sits in a suburban office park, not on the 38th floor. The seller needs to thrive on Dunkin’-fueled drives to meet customers where they are.

On the flipside, some traditional B2B sales skills do not pay the same dividends when selling AI services: Being able to work across >10 deals at once is not as crucial: Close rates are high and sales cycles short. Selling to smaller companies means showing up in person beats everything but the most sophisticated online approaches, so proficiency with the traditional sourcing stack is rewarded less. Multi-threading is not required until deals go into six-figure deal sizes.

If you have the right skill set, you can advance your career in months

The skill set required to succeed in selling AI services is much closer to that of a consultant or a semi-technical Deployment/Product person. At Heron, we have found success with ex-Palantir Deployment Strategists, former founders and tech PMs. The desire to compete and win on a sale is still required, but apart from that, traditional sales experience counts less than it ever has.

The good news here: If you have the skill set, you can speed-run the traditional sales track. We have folks on track to sell >1M ARR in their first year of ever doing sales, and other people tracking to sell >3M ARR in their second or third year of selling. Fundamentally, selling to smaller companies is faster and, if you have CEO buy-in, just easier, meaning you can sell much more than before. Compare these timelines to the traditional Sales track: 2 years as an SDR, then a year as a junior AE, then hoping to hit 1M ARR quota in the next year as a full AE. In AI services sales, careers can progress by years in a few months.

In tech, you’ll eventually be selling or building. You may as well sell now.

Even once folks realize this, many find the idea of a sales career unappealing - it feels like you’re running around conferences shaking hands with strangers and slinging slide decks at anyone who will listen. AI services sales looks nothing like that. It’s diagnosing a CEO’s business with them, designing the workflow that fixes it, and getting paid when it works. Closer to consulting than to selling. If you’re not going to be a builder, there has never been a cleaner on-ramp into sales. You may as well sell now.