Business & Growth , , , , ,

Product-Led Growth vs Sales-Led: How to Choose

A team meeting around a table in an office

The choice between product-led and sales-led growth is mostly settled by your average contract value before you ever get a vote. Below roughly $3,000 a year per customer, a quota-carrying rep cannot pay for themselves. Above about $30,000, self-serve signup alone will not get a procurement team to sign. The interesting decisions live in the gap, and in what happens when a company tries to move from one motion to the other.

What the two motions actually are

Product-led growth means the product acquires, converts and expands its own users: free tier or free trial, self-serve signup, credit card checkout, usage that grows without a human in the loop. Sales-led means a human identifies the buyer, runs a process, negotiates and closes, and the product is proven in a pilot rather than in a signup flow.

The distinction that matters operationally is not marketing versus sales. It is who bears the cost of the first ten minutes of a prospect’s experience. In PLG the product does; in sales-led a person does, at somewhere between $80 and $250 an hour of fully loaded cost. That single fact drives every threshold below.

The ACV bands, with numbers

Christoph Janz’s 2014 essay Five ways to build a $100 million business remains the cleanest framing: to reach $100m in revenue you need 1,000 customers at $100,000 a year (“elephants”), 10,000 at $10,000 (“deer”), 100,000 at $1,000 (“rabbits”), a million at $100 (“mice”) or ten million at $10 (“flies”). Janz paired each tier with a motion: field sales for elephants, inside sales and channel for deer, inbound marketing for rabbits, virality for mice, advertising for flies.

Kyle Poyar retested the framework in June 2026 against 1,043 SaaS and AI companies that had reached $10k MRR by May 2023 and were still operating three years later. The distribution at $10k MRR: 53% were selling rabbit-sized deals ($30 to $299 per month), 34% mice (under $30), 12% deer ($300 to $2,999) and 1% elephants ($3,000 and up).

The outcome data is where it gets pointed. Companies that stayed focused on deer grew 22% year over year on average, against 5% for elephant hunters, 4% for rabbit hunters and 2% for mouse hunters. Annualized gross revenue retention at $10k MRR ran 24.6% for mice, 31.6% for rabbits, 70.5% for deer and 100% for elephants. And 70% of companies were still selling to the same customer type three years on, with rabbit-to-deer the most common migration at 13% of the dataset.

Annual contract value Viable motion What has to be true Typical failure
Under $360 Pure self-serve, viral or ad-supported Product spreads through use; support is documentation, not people Retention collapse; ChartMogul found 23% median gross revenue retention below $50/month
$360 to $3,600 Self-serve with inbound marketing Cost per signup low enough that paid acquisition works without a rep Paid channels saturate and CAC payback stretches past two years
$3,600 to $36,000 Self-serve plus sales assist, or inside sales Product-qualified leads exist and are routed fast Reps chase every signup instead of the 5% that convert
$36,000 to $250,000 Inside sales, land-and-expand from a team-level entry point Multi-seat expansion inside the account Free tier cannibalises paid seats
Above $250,000 Field sales, sales engineers, security review, procurement Budget owner is an executive with an annual planning cycle Founder-led deals that no rep can replicate

Freemium conversion: verified benchmarks

OpenView’s 2022 Product Benchmarks report, based on a survey of more than 450 practitioners, put the median free-to-paid conversion rate at 5% for freemium products and 17% for free trials. Fifty-five percent of respondents identified as product-led, up from 45% in 2019, and 61% of the Cloud 100 had a PLG strategy. Standout PLG companies above $30m ARR contacted only 14% of signups with a human, against 33% at typical freemium companies.

A wider distribution comes from Lenny’s Newsletter’s free-to-paid benchmark study, which pooled data from Pendo, OpenView and Growth Unhinged across more than 1,000 B2B products. Its thresholds:

  • Freemium, pure self-serve: good is 3% to 5%, great is 6% to 8%. Examples cited include Canva, Trello and Typeform.
  • Freemium with sales assist: good is 5% to 7%, great is 10% to 15%. Airtable, GitLab and HubSpot are cited.
  • Free trial: good is 8% to 12%, great is 15% to 25%. Shopify, Google Workspace and Intercom are cited.
  • Developer-focused products converted at roughly half the median of non-developer products.

Note the shape of the distributions rather than the medians. In the freemium group, 20% of products converted below 2.5%; in the free-trial group, only 7% did. Freemium has a much fatter left tail, which is another way of saying that a free tier is a bet you can lose.

Hybrid is the default now, not the exception

The productive question in 2026 is not which motion but where the handoff sits. The sales-assist pattern that OpenView measured, where a small fraction of self-serve signups get routed to a human on a product-qualified signal, is what produces the 10% to 15% conversion rates in Lenny’s “great” band for freemium.

Atlassian is the reference case for the top of that funnel. It built its business with no outbound sales force at all, and in fiscal 2026, ended 30 June 2026, reported $6,572m of revenue, up 26%, with cloud revenue up 28%, subscription ARR of $6,606m and remaining performance obligations of $4,817m, up 44%. That RPO growth is the signature of long, negotiated enterprise contracts layered on top of a self-serve base, and the company’s own commentary describes enterprises “deepening their commitment to our open platform.”

HubSpot shows the same arc at a different price point. Poyar notes it launched at roughly $250 per month and now averages close to $1,000 per customer per month. Its 2025 results put average subscription revenue per customer at $11,683 in Q4 across 288,706 customers and $3.13bn of revenue. A rabbit hunter that became a deer hunter, over roughly fifteen years.

Pricing is drifting the same way. Poyar’s State of B2B Monetization report, surveying more than 230 B2B software and AI companies in April and May 2026, found hybrid pricing (a subscription base plus usage) had become the most common model at 37%, up from 25% twelve months earlier, while per-seat pricing persisted mainly at companies above $150m ARR.

What breaks when you switch motions

  1. Going up from self-serve: your free tier becomes a competitor to your own sales team. Anything a buyer can get free at team scale is a discount your rep has to argue against. Companies that survive this typically move the entry point rather than removing it: keep the free tier for individuals, gate everything an administrator needs (SSO, audit logs, permissions, data residency) behind the paid plan.
  2. Going up from self-serve, part two: your metrics stop meaning what they meant. Signup-to-paid conversion falls as sales cycles lengthen, which looks like funnel decay and is actually mix shift. Split the reporting the day you hire the first rep, not six months later.
  3. Coming down from sales-led: the product is not built for self-serve because implementation, configuration and data migration were always done by humans. Onboarding is the whole project, and it usually takes longer than building the checkout flow.
  4. Coming down, part two: support volume scales with users, not revenue. A $200-a-year plan cannot absorb a single 20-minute support call per customer per year.
  5. Either direction: compensation. Reps paid on new ARR will not nurture a self-serve pipeline, and marketers measured on MQLs will not optimise for product-qualified signals. The comp plan changes before the motion does, or the motion does not change.

How to decide this week

Compute your fully loaded cost per closed deal and divide it by ACV. If a rep’s total cost per closed customer exceeds about a third of first-year ACV, a sales-led motion will not clear payback in a reasonable window and the product has to do the work. Then check gross revenue retention against the price band you are actually in: if you are selling below $50 a month, the Poyar and ChartMogul data both say retention, not conversion, is your binding constraint, and no change of motion will fix it.

The uncomfortable finding in the 2026 data is that the mid-market band nobody romanticises, deals of a few hundred to a few thousand dollars a month, grew four to eleven times faster than the tiers on either side of it. If your ACV is drifting up from consumer pricing, that drift is probably the most valuable thing happening at your company.

Sources

Tag

Related Posts

Post Comment

You May Have Missed