PLG versus sales-led
Choose self-serve, sales-led or a hybrid by setting the CAC each motion produces against the margin its contract returns in the first year.
On this page
Product-led growth (PLG) lets the product do the selling: a visitor signs up, uses it, and puts a card on file without speaking to anyone. A sales-led motion puts a person in the path: a lead books a demo and a rep closes the deal. The choice is argued as culture and decided by arithmetic. Each motion has a CAC, each contract has a margin, and a rep costs thousands of dollars a deal, so the contract has to be large enough to carry one.
Two motions, priced
The self-serve CAC is the one this module has used throughout: $50,000 of paid media for 96 paying customers, $521 each, from CAC, LTV and payback.
The rep-assisted CAC comes from module 0's constants in the arithmetic a marketer carries. Demo bookers come from search at module 0's $2 click, not at the funnel's $0.50 average, which is mostly broad social traffic; one in a hundred books a demo, so a demo costs $2 ÷ 1% = $200 of media, and one demo in five closes, so media costs $1,000 a deal. A rep costs $150,000 a year loaded and closes 30 deals a year, which is $5,000 a deal. In all:
self-serve CAC $50,000 ÷ 96 = $521
rep-assisted CAC $1,000 media + $5,000 of rep = $6,000
The rep's time is five-sixths of the second figure, and it does not shrink when the contract does. A rep spends roughly as long closing a small agency as a large one, so the cost per deal is set by deals per rep, not by deal size.
The mistake: a rep for a $1,200 contract
The invoicing app adds a Team plan for agencies: $100 a month for up to five people, $1,200 a year. Agencies feel like bigger customers, so the founder hires a rep to sell the Team plan on annual contracts paid up front.
At 80% gross margin an annual contract returns $960 in its first year against a $6,000 CAC. Contracted payback is $6,000 ÷ $80 a month = 75 months, and that is the flattering version, because it assumes every agency renews forever. Assume 85% of annual contracts renew. Lifetime margin is then $960 ÷ 0.15 = $6,400, an LTV/CAC of 1.07, and cumulative margin, 960 × (1 + 0.85 + 0.85² + …), first reaches $6,000 in year 18. In its first year the rep costs $150,000 plus $30,000 of media and brings in 30 × $960 = $28,800 of margin.
The same Team plan sold self-serve has a $521 CAC and $80 a month of margin. At 5% monthly churn its cumulative margin is 1,600 × (1 − 0.95ⁿ), which passes $521 in month 8. The product is the same; one motion pays back in eight months and the other in eighteen years. A $1,200 contract cannot carry a $6,000 CAC.
The contract size at which a sales motion pays
The course's payback target is 12 months. On an annual contract paid up front, the first year's margin is collected whatever happens at renewal, so the target becomes: first-year margin covers CAC.
0.8 × ACV ≥ $6,000 → ACV ≥ $7,500
$7,500 is a floor, not the answer. Larger accounts bring more approvers and longer cycles, so deals per rep fall as the contract grows. Assume a rep closes 20 deals a year at that size rather than 30: rep cost rises to $7,500 a deal, CAC to $8,500, and the contract needed to $8,500 ÷ 0.8 ≈ $10,600. The threshold is where first-year margin catches a CAC that rises with it, about $10,000 a year for this company. Above it a sales motion pays inside the year; below it the rep is a cost the contract cannot carry. What changes inside a large deal, the pilot, the annual contract and the expansion, is the subject of what the deal depends on.
Hybrids: a rep where the product has already qualified the account
Between a $1,200 plan and a $10,000 contract sits the case most companies are in, and the answer is usually neither pure motion. In a hybrid every account starts self-serve, and a rep calls only the accounts whose usage says they are ready for more.
For the invoicing app the signal is a free workspace that has invited five or more people. The rep offers those accounts a Business plan: 15 people, $3,600 a year, billed annually. Because the prospect is already using the product, the calls are short; assume a rep calls 200 such accounts a year and closes half, 100 deals, which puts rep time at $150,000 ÷ 100 = $1,500 a deal.
Test the rep against what the account would have done without the call. A signalled workspace left alone buys the Team plan at $1,200. The call turns that into $3,600, so the rep adds 0.8 × ($3,600 − $1,200) = $1,920 of first-year margin for $1,500 of rep time. The rep pays back inside the contract year, on accounts the self-serve funnel had already paid to acquire.
Two conditions hold the hybrid up. The signal has to be frequent enough to fill a rep's year: 200 calls a year is about 17 signalled workspaces a month, and below that the founder makes the calls. And the close rate on called accounts has to be measured, not assumed, because the rep's cost per deal moves with it. Whether to price per seat or per use for those larger accounts is a separate decision, worked in pricing an AI feature.
What AI changed, and what it did not
Research on an account, call notes and follow-up drafts now take minutes, so a rep can close more deals a year. Assume 40 instead of 30: rep cost falls to $3,750 a deal, CAC to $4,750, and the contract a sales motion needs to $4,750 ÷ 0.8 ≈ $5,900. The line moves down; it does not disappear. An in-product assistant that answers setup questions pushes the other way, letting self-serve close accounts that used to need a call. Both are changes to the inputs of the same inequality.
In practice
To the founder or the board, before the first sales hire: "A rep selling the $1,200 Team plan costs $6,000 a deal against $960 of first-year margin, and pays back in year 18. A rep calling workspaces with five or more people, and selling the $3,600 Business plan, adds $1,920 of first-year margin for $1,500 of time. We hire the first rep when signalled workspaces reach about 17 a month."
To the rep, once hired: their number is first-year margin added per deal against their cost per deal, measured against what the account would have bought alone.
The mistake that costs money is hiring sales capacity before the contract can carry it. A rep at $150,000 closing $1,200 contracts costs $180,000 with media in the first year and brings in $28,800 of margin.
Check yourself
1. A network of agencies wants a 60-person contract at $14,400 a year, with procurement and a security review. A rep closes about 12 deals of that size a year. Does a sales motion meet the 12-month target?
CAC is $150,000 ÷ 12 + $1,000 = $13,500. First-year margin is 0.8 × $14,400 = $11,520, so no: it pays back in year 2 if the contract renews. To meet the target the contract has to reach $13,500 ÷ 0.8 ≈ $16,900. Quote that, or accept the second-year payback for a few such deals with the cash approved for it.
2. Three months into the hybrid, the rep's close rate on called workspaces is 30%, not 50%. Is the rep still worth it?
200 calls at 30% is 60 deals, so rep time is $2,500 a deal against $1,920 of added first-year margin. It misses the year; with 85% renewal the second year adds $1,632, and it pays back in year 2. Before cutting the rep, tighten the signal, for example ten invited people instead of five, and measure whether the close rate rises on fewer, better calls.