Self-serve checkout is the transparency forcing function
Price transparency is set by whether a self-serve purchase path exists, not by whether a sales team exists. Sales-led companies publish prices 69% of the time — level with the corpus base — but remove the self-serve path and publication collapses to 2 of 77 companies, about 3%. A checkout cannot take money without a number on the page.
What's happening — and why
What's happening: the presence of a sales team tells you almost nothing about whether a vendor publishes prices. 334 corpus companies carry a sales-led motion and 69% of them publish — statistically indistinguishable from the 72% corpus base. What predicts publication is whether a buyer can complete a purchase alone.
Among companies whose only motion is sales-led, just 2 of 77 publish a price. Among PLG plus self-serve companies, 39 of 40 do. That is roughly a 33x swing on a variable that has nothing to do with company size, vertical or enterprise-ness.
Why: the mechanism is mechanical rather than cultural, which is what makes it closer to a definition than a discovery. A checkout cannot take money without a number on the page. A trial can be gated behind a form and still be called self-serve; an activation funnel that must convert without a human cannot hide the price.
What keeps it empirical is the exception set — the small number of companies that publish a rate card they refuse to sell self-serve. That set has been stable across corpus expansions, and it just lost a member: Regie.ai relaunched with a self-serve Free plan and $49/mo Pro on 2026-08-25, resolving its own exception in the direction the trend predicts.
How it works
Evidence over time
13 supporting · 6 counter — hover or tap a point for detail, click to jump to the row.
Evidence
| Company | Date | What happened |
|---|---|---|
| Harvey | May 2026 | Gated enterprise per-seat pricing, sales-led quote only. sales_motion is [sales-led] with no self-serve or PLG tag and no free tier — one of the 70 companies in the sales-led-ONLY-and-opaque cell. |
| Glean | Jul 2026 | Gated hybrid — per-user seats plus pooled FlexCredits, or seats plus per-token Model Hub usage. Three pricing-model tags and two live meters, none of them priced on the page; no self-serve path, no free tier. |
| Gong | Jun 2026 | Per-user licences plus a platform fee scaled to supported users, quote-only. A horizontal-SaaS incumbent in a segment where 91% of companies bill seats — and no seat price is published, because there is nowhere to buy one. |
| Ada | Jun 2026 | Outcome-based per-resolution billing, sales-quoted, no public price (third-party data puts the rate near $1.00-$3.50 per resolution). sales-led only; no sign-up flow exists to price against. |
| Sierra | Jul 2026 | Outcome-based pay-per-resolution plus volume pricing for routine interactions, sales-only/custom. Tagged outcome-based, pure-usage and hybrid simultaneously with no published number on any of the three. |
| Mercor | Jul 2026 | Sales-quoted marketplace: the buyer-side take-rate is undisclosed while the supply-side hourly expert pay is published. Opacity is one-sided — the price the platform charges has no self-serve path, so it has no published number. |
| Skydio | Jun 2026 | No first-party rate card at all: X10/X10D hardware plus per-drone software subscriptions (Skydio Care, Remote Ops, Cloud) are quoted via sales or GSA. Reseller and government-schedule prices exist publicly; Skydio publishes none directly. |
| Orb | Jun 2026 | Gated sales-led custom pricing on billings plus events usage, with a platform fee on higher tiers — a usage-based billing vendor with no self-serve tier and therefore no published rate. |
| m3ter | Jul 2026 | Gated hybrid: a core platform fee bundling usage-data and bill-calculation allowances, plus add-ons, support and implementation. Metering-infrastructure vendor, no sign-up flow, no published price. |
| ZoomInfo | Jul 2026 | Seat plus credit-pool pricing, quote-only and gated, with commits — a NASDAQ-listed company with a fully documented three-part meter and no self-serve purchase path, so no number reaches the page. |
| Lorikeet | Jul 2026 | WRONG-SIDE CASE. Publishes Start $1,500/mo and Scale $4,000/mo on annual per-resolution credit pools, yet sales_motion is [sales-led] only: the 2024 tiers had a pay-by-card 'Get started' checkout and by May 2025 every tier routed to 'Book a demo'. The price outlived the checkout — direct proof the two fields are coded independently. |
| Regie.ai | Jun 2026 | WRONG-SIDE CASE. Publishes AI SEP at $180/user/mo and Force Multiplier Rep at $499/user/mo with bundled AI/enrichment credits, on annual contracts with no self-serve checkout — a public rate card sold entirely through a rep. |
| Tabnine | Jun 2026 | WRONG-SIDE CASE. Publishes Code Assistant Platform $39/user/mo and Agentic Platform $59/user/mo, plus a capacity-tiered headless add-on, but both are annual subscriptions quoted via sales and the free tier was retired in 2024 — published price, no purchase path. |
Counterexamples
- Automation Anywhere · Jul 2026 — Tagged sales_motion [self-serve, sales-led] and has_free_tier true, yet price_transparency sales-only — because the self-serve half IS the free Community Edition and every paid Automation 360 edition is quoted by bots, seats and cloud consumption. A product-line split, not a vendor that sells paid plans self-serve without publishing a price.
- Sana AI · Jun 2026 — Same split, and the file states it outright: Sana Agents has a public rate card (Free / $30 per user/mo Team / custom Enterprise) while Sana Learn is quote-only with a 300-seat minimum. price_transparency is single-valued, so the enterprise line sets the tag and the self-serve line is invisible to it.
- Unbabel · Jun 2026 — Core LangOps translation is sales-led and quote-only; its Widn.ai self-serve product runs free to $90/mo. Two motions, two transparency states, one field — the tag records the enterprise surface.
- Clockwise · Jun 2026 — A dead-surface artefact rather than a live exception: the product was switched off 2026-03-27 after Salesforce acqui-hired the team, so the live site has no prices. Its historical self-serve ladder (Free / Teams $6.75 / Business $11.50 per user/mo) is reconstructable only from Wayback.
- Lago · Jul 2026 — The cleanest live bound on the claim: open-core usage-based billing with a genuine self-serve on-ramp (free self-hosted, has_free_tier true) and sales_motion [self-serve, sales-led], yet the one paid package — Lago Premium — carries no published price, so transparency is gated. A free self-serve path forces a $0, not a rate card.
- Maxio · Jun 2026 — Publishes Grow at $599/mo for up to $100k in monthly billings and a free Build developer sandbox, then makes Scale quote-only above $100k billings — self-serve plus sales-led, coded gated. The checkout forces a number only for the band it can transact.
Trivia
-
The corpus grew by 27 companies between reviews and this correlation did not move a single percentage point: sales-led-with-no-self-serve-or-PLG is opaque 75/78 (96%) at 380 against 70/73 (96%) at 353, and the reverse cut is 56/60 (93%) against 51/55 (93%). The three companies that publish inside the sales-led-alone cell are also the same three. Stability at that resolution is the signature of a mechanical relationship rather than a market fashion.
-
One published sub-claim died in this window. The 2026-07-30 review found that sales-led + self-serve without PLG published at 33 of 50 (66%), BELOW the 71% corpus base, and told buyers "a self-serve path alone is not enough." At 380 the cell reads 41 of 58 (71%) against a 72% base — level. All eight companies added to the cell publish, a 100% marginal rate that erased the deficit. The gap between "has a checkout" and "is product-led" closed on this measure.
-
Exactly 3 companies in 380 publish a price they will not let you buy: Lorikeet (Start $1,500/mo, Scale $4,000/mo, verified 2026-07-21), Regie.ai ($180 and $499 per user/mo, 2026-06-05) and Tabnine ($39 and $59 per user/mo, 2026-06-09). Lorikeet's own page history records the move that created the mismatch: its 2024 tiers carried a pay-by-card "Get started" click-through, and by May 2025 every tier routed to "Book a demo" — the checkout was removed and the numbers were left behind.
-
Two of the ten opaque vendors sampled here sell usage-based billing and metering infrastructure and will not publish their own price: Orb (gated, verified 2026-06-03) and m3ter (gated, 2026-07-21). Both sit in the 70-company sales-led-ONLY cohort, and both sell the tooling that makes a published usage rate card operable for someone else.
-
Skydio (verified 2026-06-14) shows opacity is a property of the vendor's own surface rather than of the market: reseller and GSA government-schedule prices for its X10 drones and per-drone software subscriptions do exist in public, and Skydio publishes none of them directly. The number is knowable — just not from the company that sets it.
For buyers
Use the checkout as your leading indicator. If a vendor sells self-serve, a real price exists and you can anchor on it before you ever talk to sales. If a vendor has no self-serve path at all, expect no published price and budget for a quote cycle — that is 96% of the sales-led-only cohort. The corollary is the useful part at renewal: a vendor that removes its self-serve on-ramp is very likely to remove its published price in the same change, as Artisan did when it deleted its free tier, credit estimator, self-serve checkout and rate card together.
For vendors
If you want a published price to be credible, ship a checkout behind it — buyers read an unbuyable rate card as an opening position, which is roughly what it is. Conversely, if you are moving up-market and dropping self-serve, understand that you are also making a transparency decision whether or not you intend to, and that it is highly visible. The rarest and most interesting posture is the exception set: publishing a rate card you will not transact on. It is defensible as an anchoring strategy but it is unstable, and the corpus's examples keep resolving in one direction or the other.
Outlook — what to watch
This claim has now survived a 27-company corpus expansion and a zero-intake window with both rates essentially unmoved, so expect it to keep holding — it is closer to a structural definition than a market trend. The thing worth watching is the exception set, which is shrinking: Regie.ai left it on 2026-08-25 by adding a self-serve Free plan and $49/mo Pro. If the remaining exceptions also resolve, the relationship becomes near-deterministic and stops being interesting as a finding, which would be its own result.
Bottom line
Whether a vendor publishes a price is decided by whether you can buy without talking to anyone. Sales-led companies publish at the corpus average; sales-led-only companies publish 3% of the time. The checkout is the forcing function.
FAQ
Does having a sales team mean a company hides its pricing?
No — that is the finding. 334 corpus companies carry a sales-led motion and 69% publish a price, level with the 72% corpus base. The presence of a sales team carries essentially no information about transparency. What matters is whether a self-serve purchase path also exists.
Why does self-serve force a published price?
Because a checkout cannot take money without a number on the page. A trial can be gated behind a form and still be called self-serve, but an activation funnel that must convert without a human cannot hide the price. That is why this sits closer to a definition than a discovery.
What are the exceptions?
A small, stable set of companies that publish a rate card they will not sell self-serve — historically Lorikeet, Regie.ai and Tabnine. The set survived a 27-company corpus expansion unchanged, which is what keeps the claim empirical rather than tautological. Regie.ai left the set on 2026-08-25 by relaunching with a self-serve Free plan and $49/mo Pro.
What should I watch for at renewal?
The removal of the self-serve on-ramp. Companies that lose PLG tend to lose the public price in the same change — Artisan removed its free tier, credit estimator, self-serve checkout and rate card together, and LanceDB folded its self-serve usage-based tier into a sales-led annual-commit Enterprise so no paid option is reachable without a contract.