Gong Will Not Publish a Price. 114 of 264 Do Not

There is no price for Gong.

Not a high one, not an annual-only one, not one hidden behind a currency selector. This site read gong.io on 11 August 2026 and found no figure on the page. The record kept here holds null in every price field, and it is one of 114 records out of 264 in this directory that do.

The reason is structural rather than coy, and understanding it is worth more than any number a competitor's comparison page would hand you. Gong bills two separate things: a fixed platform fee, and a per-seat licence on top of it. Neither is published, and the ratio between them inverts completely between a team of eight and a team of fifty. There is no single figure that would be honest for both buyers.

So the useful question is not what Gong costs. It is what shape the bill has, and at what headcount that shape starts working.

What follows: the two components and why only one of them scales, the floor below which the arithmetic cannot close, what a Gong seat can be bracketed against using the rates this category does publish, the repositioning that added an unpriced agent layer, and the four things to get in writing before a quote means anything.

What does Gong cost?

Unknown, and per user is the wrong unit to ask in.

There is a platform fee and there is a seat rate, both unpublished, sold on an annual cycle. Ask for the two figures separately during the sales process, because a quote that merges them into one per-rep number hides exactly the variable that decides whether the product works at your size.

There is no free tier and no self-serve path. The earliest point at which you can learn what packages exist is after a demo, which puts Gong in the set of 96 tools here that publish no price and also have no checkout.

The one number this site will stand behind about Gong is not a price at all. It is roughly twenty reps.

Why no single number would be honest

A fixed fee behaves differently from a per-seat one, and Gong's bill contains both.

The seat rate scales linearly with your team. The platform fee does not scale at all: it is the same number whether ten people use the product or two hundred. Divide the total by headcount and you get a per-rep figure that falls steeply as you grow, which means any published per-rep price would be wrong for almost every buyer who read it.

Buying it is closer to hiring a builder than to buying software. The quote arrives after somebody has looked at your house, and it is specific to the house.

That is a defensible reason not to publish, and it is worth conceding before criticising the consequence. The consequence is still real: a tool you cannot price is a tool you cannot shortlist, and you are being asked to spend an hour on a discovery call to learn the one thing that decides whether Gong is even in scope.

The platform fee, not the seat rate, decides this

Model Gong as total contract divided by reps, never as the seat rate you are quoted.

The cruelty in that table is that the two variables move together in the wrong direction. At eight reps you are paying the most per head and feeding the analysis the least data to work with, because pattern detection needs patterns and eight people do not generate enough conversations to produce any that a manager listening to calls would not already have found.

Where the twenty-rep floor comes from

Two independent constraints land in the same place, which is why the number is worth carrying into a demo.

The first is the fee arithmetic above. A fixed cost spread across forty reps is a rounding error on the per-head figure and doubles the effective seat price spread across eight.

The second has nothing to do with money. Gong's output is pattern detection across a team: deals mentioning a specific objection close at half the rate, the competitor named in three conversations nobody logged, the forecast that will not land. Those findings require enough conversations for a pattern to be distinguishable from an anecdote. Below about twenty reps the analysis has too little to work with, and a manager who listens to calls reaches the same conclusions for free.

This is this site's reading and not a Gong statement. It is also the most useful thing to know before taking a demo, and it is not something the demo will tell you.

What a Gong seat can be bracketed against

The full sixteen-tool ladder, and where the published half of it stops

The seven-row bracket above is the short version. Widened to every recording and coaching tool in this directory, with both billing bases separated, the picture is starker: the published prices cluster inside a $25 band and then the table simply stops, because ten of the sixteen tools in the category publish nothing at all.

Two things fall out of that table and both are useful in a Gong conversation. The first is that the published half of the category has no seat rate above $34, which puts a ceiling under any bracket you build for the licence component. The second is that four of the six vendors who publish will also take a card, so you can hold a live comparator through the whole negotiation without ever entering another sales process. Gong, Clari Copilot and Chorus will not let you do that, and Chorus is the one you already pay for if you have ZoomInfo.

The annual column is the one to read carefully. Otter drops from $16.99 to $8.33, a 104.0% monthly premium [computed], which is the steepest gap in the category and a reminder that a seat rate quoted without its billing basis is not a rate. Grain and Trellus carry no annual figure in this directory at all, and they are marked as not held rather than estimated.

You cannot price Gong from outside. You can walk into the demo with an anchor, which is more than most buyers have, and the anchor is what the rest of this category charges for adjacent jobs.

Avoma is the closest published comparator for the job Gong actually does, and reading it carefully sets a floor. Its plan is $29 a seat monthly, its Conversation Intelligence add-on is a further $29, so the honest coaching price with a checkout is $58 a seat on the monthly basis [computed], or $48 if you take the $19 annual plan rate the vendor's own materials pair the add-on with.

A Gong quote materially below that would be surprising. A Gong quote several times above it is normal, and the multiple is the platform fee.

Gong renamed itself and priced none of the new part

The product now presents as the Gong Revenue AI OS, assembled from a Revenue Graph for context, a Revenue Harness for execution, and applications on top, described by the vendor as where humans, agents and tools work together.

Gong's own words for the Harness are worth quoting rather than paraphrasing: the execution layer that governs how AI agents plan, act, and deliver results. Plan and act are the vendor's verbs. A governance layer is only worth building if the thing underneath it does something consequential, so the existence of a Harness is itself the disclosure.

None of it carries a price. The conversation-intelligence product this site reviewed is now the bottom layer of something sold as an operating system, and a buyer can price the agents exactly as well as they could price the seats, which is not at all. What those agents are permitted to touch is also not published on any page reachable without a sales call.

Search for this and you get a musical instrument

This site audited the search results for a set of vendor queries and recorded which ones return the wrong product. Gong pricing returned a result about the percussion instrument, headed How much does a Gong cost?, on one of the result pages checked.

That is a small thing and it points at a larger one. A vendor that published a rate would own this query outright. Because Gong publishes nothing, the space fills with competitors' comparison pages, aggregators working from stale figures, and an orchestral supplier. Every number you find for Gong in that environment was written by somebody with a reason to want it to look a particular way.

What you are actually paying for

Deal inspection, and it is a genuinely different product from call transcription.

Plenty of tools transcribe. What Gong sells is the reading of conversations against deal outcomes: the deal with no economic buyer on any call, the one that has gone quiet, the forecast that will not land. Forecast accuracy is the reason enterprises buy it, more than the coaching, and the coaching is what gets demonstrated.

It does not create pipeline. Everything on this site that touches prospecting sits upstream of Gong, and Gong analyses whatever those tools produced. Point it at an empty funnel and it will tell you, accurately and expensively, that your funnel is empty.

What recording obliges you to do, at any price

The consent question attaches to you rather than to the vendor, and it is a legal question before it is a product one. Several US states and most of Europe require the consent, knowledge or informing of every party to a recorded conversation.

This site audited sixteen notetakers on what they do before anyone configures them. Gong's row: the auto-join default is not stated publicly, the bot appears as a named participant so the other side can see it, whether the recording notice is on out of the box is not stated publicly, and the setup wording reads Turn On, which implies off. Read 21 August 2026 and spot-checked on the 25th.

One column Gong does well. An administrator can enforce use of consent page, which is a control rather than a training item, and only three vendors of the sixteen can force a policy against a rep who would rather not comply. The full sixteen-row audit is in what every notetaker does by default.

The four things to get in writing before a quote means anything

• The platform fee, as its own line, for your named headcount. Not blended into a per-rep figure. This is the number that decides the purchase and it is the one most likely to arrive bundled.
• The per-seat licence, separately, and what happens to it at renewal if your team shrinks. Annual cycles rarely let seats come down mid-term, and a team that over-hires seats in month two carries them.
• Which layer of the Revenue AI OS the quote covers, and what the agents in the Harness are permitted to act on. The vendor's own verbs are plan and act, and none of it is priced publicly.
• Whether the consent enforcement setting is available on the tier you are being quoted, or only above it. Otter puts the equivalent behind an Enterprise contract, and finding that out at renewal is expensive.

What breaks in month three

Two things, and neither is a product fault.

Nobody reviews the recordings. Recording is cheap and reviewing is not. Teams that do not name an owner and a weekly slot before launch generate thousands of unwatched hours and then conclude the tool failed. That failure is indistinguishable from the tool being wrong, and it costs the same.

Rep resistance, which is a management problem the software cannot solve. Being recorded and scored is a real change to the job. Framing it as coaching rather than surveillance is work somebody has to do before the first scorecard lands, not after.

When Gong is the wrong purchase

When you have fewer than about twenty reps. The platform fee dominates, the analysis is data-starved, and a manager who listens to calls does the same job. No feature comparison changes that arithmetic.

When what you wanted was call notes. Notetakers do that from $0, and the published seat rates in the table above run from $16.99 to $29. Buying deal inspection to get a summary in Slack is an expensive way to solve a solved problem.

And when the problem is not enough pipeline rather than unclear pipeline. Gong sits downstream of prospecting. It will diagnose the gap and it will not fill it, and the diagnosis is not cheap.

So what do you do with a page that has no price?

Count your reps first, and let that decide whether to book the call at all. Under twenty, the arithmetic above says the product cannot pay back and the demo will not change it. Over fifty, book it, and negotiate the platform fee separately from the seat rate.

Between twenty and fifty the honest answer turns on one thing, and it is not a feature. It is whether anybody will run weekly call review. If yes, Gong pays back. If no, you are about to buy the most expensive unwatched video library in your stack.

Whatever you decide, keep one published comparator on the shortlist even if you will never buy it. With every tool in an evaluation gated, the first quote you receive becomes your only anchor, and you cannot obtain an independent one once negotiations have started.

What this page does not know