Gamma's average payer is the wrong unit for an enterprise sales decision

Gamma reached a reported $100 million in ARR with 600,000 paying subscribers and, for much of that journey, no sales team. Divide those two figures and you get roughly $167 per paying subscriber per year.

That arithmetic is useful. The next inference needs more care.

SaaStr's account of CEO Grant Lee's remarks describes inbound requests to buy for whole teams or departments that went unanswered until Gamma built a sales function. The reported company-wide average does not tell us what those prospective contracts were worth.

I would not use $167 to prove that Gamma could never have afforded sales. I would use it to ask which purchase needs a salesperson and which should remain self-serve.

The denominator changes when the buyer changes

A subscriber is not necessarily an account, and an account is not necessarily a single-seat purchase. Several people can use a product independently before their organisation considers a broader agreement. The average across all payers can hide the part of the business for which assisted buying makes sense.

As an illustration, imagine I am advising a hypothetical AI document product. Its individual plan can be bought without a conversation. A department then wants central administration, a security review and a coordinated rollout for many users.

The economic comparison is no longer a salesperson's cost against one individual's subscription. I would compare the incremental contribution from the department agreement with the cost of qualifying, closing and supporting that agreement. Those inputs need to be estimated and then replaced with observed results.

A larger contract does not automatically make sales profitable. Lengthy procurement, custom commitments and heavy onboarding can consume the apparent upside. Equally, a low average subscriber price does not establish that every prospect is uneconomic to assist.

The useful segmentation begins with the buying task. A customer who can choose and pay immediately should not be forced through a meeting. A customer who cannot complete an organisational purchase through the existing flow needs a clear route to someone who can resolve it.

Test the unanswered demand

Lee's account is particularly useful because the signal was already arriving. Potential customers were asking for a buying motion the company had not yet staffed adequately. That suggests a practical experiment before committing to a large sales organisation.

I would take a bounded sample of genuine team-purchase requests and assign an owner to handle them. Record why the customer needed help, the time spent, the outcome and the ongoing obligations created. Keep unsuccessful opportunities in the review, including those that never had a plausible budget.

Then compare the assisted path with what the product could solve directly. If most questions are straightforward billing confusion, improving checkout might be the better investment. If the requests involve repeatable organisational requirements, a focused sales and onboarding motion may earn its place.

This also changes how I read ARR per employee. It is an outcome of a particular product, customer mix and operating model. Copying the ratio into another company's staffing target without those conditions risks treating a result as a universal design principle.

The Gamma story does not require choosing between admiration for self-serve growth and criticism of missed enterprise demand. The two motions can serve different purchases inside the same customer base.

Judge a sales motion by the purchase it enables, not the average subscriber it never needed to call.