When self-serve needs a sales team, and whom it calls first

Sell first to accounts where usage is already spreading and sticking. A founder can work those deals early; a quota-carrying team waits until price point, deal friction and usage data show self-serve can't close them.

• Price point and deal friction set the timing. Brian Balfour says revenue per user has to pay for the acquisition channel, and once deals pass roughly $10K a year, buyers ask for POs, security reviews and procurement. HubSpot's $25-per-user-a-month Sales tier couldn't cover inside sales costs; a $50 tier is when the sales machine started working. Hila Qu looks for several active users from one domain hitting team or admin limits, and for deals stalling at security review, procurement or seat expansion.
• Start with a founder, scale with a team. Casey Winters calls Hila's PQL signals a late trigger for starting sales and would let a founder or generalist work deals while the user base is still growing. Hila accepted the distinction: her "too early" applies to a quota-carrying team. Gustaf Alströmer would start only once the accounts with the deepest usage are still active a few months later, since reps on churning accounts just buy churn faster.
• Sell to accounts already in the product. Casey points to Dropbox, which emailed a company's CIO once roughly 20% of that company used the product, and would sell to the buyer who owns the risk of that usage. Hila would map users to teams to accounts, then rank accounts by usage depth times company fit. Brian would aim reps at accounts where usage has spread and change one variable at a time. Separately, he recalls that HubSpot's first Sales buyers were existing customers whose marketing leader had influence over sales.
• Keep the first segment narrow. Gustaf cites Slash, whose single wedge, a credit card for sneaker resellers, reached $5M ARR in 11 months before it expanded. Brian would write down the target persona and use case before hiring reps: at HubSpot, refusing commission on wrong-fit deals changed rep behavior right away. For the first hire, Gustaf recalls a YC interview in which Linear's founder described sales as part of the brand experience; for a PLG company, he says, that means a first hire who sells the way the product feels, fast and technical, without forcing a demo.
• Build the plumbing before the first AE starts. At GitLab, Hila says, generating PQLs was easy and fitting them into a sales process built around MQLs and SQLs was hard. She would settle that before hiring, with a warehouse, reverse ETL, product analytics and PQL scoring so reps see usage in the CRM, and would fit the PQL score to accounts that retained and paid rather than a guessed seat count.
• Draw the line between self-serve and sales. Hila would split the two motions by segment from day one, each on its own metric, so sales doesn't take deals self-serve would have closed. Lauryn Motamedi, who ran self-serve growth at Airtable, says leadership has to set that line explicitly, and would keep risky pricing and onboarding tests away from accounts in an active sales cycle. To see whether self-serve feeds sales, Brian would count how many of the last 20 closed deals had prior product usage; Casey would track that as a standing metric.