Blog

Sales call coaching tools compared: what the eight of us charge

Eight tools for reviewing and coaching sales calls, compared on one thing we can all be held to: what each publishes about price. Three of the eight publish nothing at all. Every figure linked and dated.

Published September 20, 2026 · updated September 21, 2026 · 4 min read

ComparisonBuyingPricing

If you run a sales team between three and twenty-five reps, you have probably worked out that the category sells to someone else. The demos assume a revenue operations function. The pricing assumes a procurement cycle. And roughly half the vendors will not tell you what it costs until you have taken two calls and named a budget owner.

So here is the whole category on one page, compared on what each of us publishes about price. A feature grid would tell you less, because whoever builds the grid picks the rows.

What each one publishes, as of today

Every figure below was read on 20 September 2026 on the vendor’s own pricing page, and every one of those pages is linked. Where a vendor publishes nothing, the link goes to the page where the price would be, so you can confirm the absence yourself in about a minute. Nothing here describes what any of these products do. That is theirs to claim.

Published price per seat, per month. Read on each vendor’s own pricing page on 20 September 2026. The lighter extension runs to the tier where coaching or conversation intelligence starts. 4 of the 9 publish no figure at all, so there is nothing to plot.

Fireflies.ai $10 to $19
Fathom $15 to $25
Avoma $19
Salesforce $50
Close-Rate $145
Gong No price published
Chorus No price published
Salesloft No price published
Jiminny No price published

Published pricing for 9 sales call coaching and conversation intelligence tools, checked 20 September 2026.
Tool	What it publishes	In their words
Close-Rate	$145 / user / monthon this site	Published on this site. A coaching report per rep per week, with every quoted line checked against the recording.
Gong	Not publishedtheir pricing page ↗	The pricing page states the structure but no figure: licenses are priced per user, and there is a platform fee based on the number of users supported. The page is a form.
ChorusZoomInfo	Not publishedtheir pricing page ↗	The page is titled Pricing & Plans and contains no numbers. It offers custom plans.
Salesloft	Not publishedtheir pricing page ↗	The pricing page carries no dollar figure, and its title element is literally “Talk to Sales”. Absorbed Wingman and Clari Copilot; clari.com now redirects here.
SalesforceConversation Insights	$50 / user / monththeir pricing page ↗	Billed annually, on top of Sales Cloud. Worth knowing before you buy anything: if you already run Sales Cloud you may already have a version of this.
Avoma	From $19 / seat / monththeir pricing page ↗	Annual billing. Conversation intelligence is not in the entry plan; it is a $29 add-on. No seat minimum below the Enterprise tier.
Fathom	Free, or from $15 / monththeir pricing page ↗	Annual billing, two-seat minimum on paid tiers. Coaching metrics and AI scorecards start on the $25 Business tier. Acquired by Superhuman on 14 September 2026, so treat the packaging as liable to move.
Fireflies.ai	Free, or from $10 / seat / monththeir pricing page ↗	Annual billing. Conversation intelligence begins on the $19 Business tier. No published seat minimum.
Jiminny	No pricing pagetheir pricing page ↗	jiminny.com/pricing returns a 404. A tiers page names three tiers with no figures and routes to a demo booking.

Which sales coaching tools publish their pricing?

Gong, Chorus and Salesloft publish no figure of any kind, and Jiminny has no pricing page at all. That is a normal enterprise motion, and for a company with a procurement function it is barely an inconvenience.

For a smaller buyer it costs more than patience. A price behind a form is set after someone has assessed what you will pay, and you cannot compare anything without spending a week of calls to assemble the row you are reading here in ten seconds.

A useful thing to ask any vendor on this page, including us: what is the minimum, in seats and in term? Of the eight, Fathom publishes a two-seat minimum and Avoma publishes none below Enterprise. For the rest we could not find a published answer. Ask on the first call.

If Gong is the one you are actually weighing us against, there is a longer page on that specific comparison.

Why not just build this yourself?

You could get something working this week. Transcription is an API call. Ask a language model to critique a sales call and it will write paragraphs without being asked twice. If that were the product, it would not be worth paying anyone for, and we would not have built a company around it.

Three things are hard, and none of them is the pipeline.

The part you could build this week Record the callTranscribe itAsk a modelGet paragraphs

The part that makes it worth reading

Knowing what to say A model asked to critique a call returns advice that is fluent, plausible and useless: ask more discovery questions, build rapport earlier, confirm next steps. The rep already knows. Working out which version of a correction a rep will act on rather than nod at is a sales problem, and our framework comes out of arguing about it with the people who run sales teams.
Making the rep believe it One invented quote and they never read another report. So every quoted line is matched against the transcript before a report may print it, and a line that fails is downgraded to a paraphrase or dropped. That machinery is most of the engineering.
Knowing whether it worked We watch close rate across every team on the platform, so the framework is corrected by what moved a number rather than by whoever argued hardest in the room.

Step 3 feeds back into step 1. That loop is the part a tool you build for one team cannot run, because one team is one data point.

The pipeline in the middle is a commodity. The loop around it is the product.

What happens when the AI invents a quote?

We read all eight product pages on the same day to build the table above. Not one of them states what happens when the model attributes a quote to the wrong speaker, or invents one outright. Ours did not either, until we wrote it down.

Here is ours, running the rule the product runs. Edit the quote — change a word, expand a contraction, tidy the grammar — and watch what a report would be allowed to do with it.

The transcript

So walk me through how you handle renewals today. Right now it is mostly manual, we pull a report at the start of each month and the reps chase whoever is up. Okay, and who owns that list? Nobody owns it, it just sort of happens. That is the part I would want to fix first. We looked at two other tools last year and both of them died in procurement, so I am cautious about anything with a long implementation.

An illustrative excerpt written for this page, not a customer call.

The quote a report wants to print

Verified. This wording appears in the transcript, so the report may print it as a quote.

Try changing “Nobody” to “No one”. One word is enough.

That is an observation about their marketing. Any of them may handle it well. But it is the first thing a rep will test, so ask every vendor you talk to. If the answer describes a prompt, nothing happens. Ours is written down in full, including what we do when a claim fails.

Check it on your own calls

Import a handful of real calls and read the coaching report against the recordings. That is the only evaluation of any tool on this page that means anything, and it takes about fifteen minutes.

Capability claims about other products reflect their own public documentation, last checked 2026-09-20. Gong, Chorus, ZoomInfo, Salesloft, Salesforce, Avoma, Fathom, Fireflies.ai and Jiminny are trademarks of their respective owners and are not affiliated with Close-Rate. Corrections: hello@close-rate.com.