How to evaluate revenue intelligence software

What revenue intelligence software covers

Revenue intelligence software sits between the CRM and the people who inspect it. It captures what happened with customers without asking reps to log it, turns that into deal and pipeline signals, and gives managers a single surface for weekly inspection and the forecast. The term covers four jobs that vendors bundle differently, which is why two quotes for the same label can differ by a factor of five.

Job one is activity capture: reading calendars, inboxes and dialers, creating contacts, and attaching activity to the right opportunity. Job two is conversation intelligence: recording, transcribing and analysing calls so coaching and deal evidence come from what the buyer said. Job three is deal and pipeline inspection: risk scoring, stage hygiene, week-over-week movement, and a commit a manager owns. Job four is relationship intelligence: who inside the account the team touches, and where the coverage gaps are.

No product is strongest at all four. Gong leads on conversations, Clari on forecast inspection, People.ai (renamed Backstory in April 2026) on activity and relationship data, and Scratchpad on making rep updates cheap. Write down which of the four is failing today before you take a single demo, because vendors will happily sell you the other three.

Pricing follows the same split. Three of the four jobs are quote-only at the serious end, while activity capture and hygiene tooling has published rates from $19 per user per month. The revenue intelligence software category lists every vendor we track with a fit line and a directory-level price. All figures checked September 2026.

Four jobs buyers conflate

The most common evaluation failure in this category is buying the wrong job well. Four symptoms map cleanly onto four purchases, and mixing them up costs a year.

If managers cannot tell which deals are real without a conversation, the job is inspection, and the products to compare are Clari, Aviso and Terret. If coaching is the gap, because new reps ramp slowly and nobody can say what the good reps do differently, the job is conversation intelligence and the comparison set is Gong, Chorus by ZoomInfo and Salesken. If the CRM is empty and reports are unusable, the job is activity capture, and Nektar, People.ai and Scratchpad are cheaper and faster than either of the above. If nobody knows which executive at an account the team has reached, the job is relationship intelligence, which People.ai and Ebsta sell.

The revenue intelligence definition is deliberately broad, so insist on the job-level comparison rather than the category-level one. A Gong quote and a Nektar quote are not alternatives; they solve different problems and only one of them is your problem this year.

There is also a sequencing argument. Activity capture is upstream of everything else, because inspection scores and conversation analytics both read the same opportunity records. Buying an inspection platform while activity data is thin produces a confident score on an incomplete book, which is the most expensive form of being wrong in this category. Our RevOps tech stack audit is the place to establish what you already capture before adding anything.

Activity capture is the foundation

Every score in this category is a function of captured activity, so evaluate capture first and features second. Five questions settle it.

What sources does it read? Calendar and inbox are table stakes; dialer, video conferencing, support tickets and chat are not. Does it create contacts it finds, or only match to existing ones? Contact creation is how buyer groups get discovered, and it is also how duplicates get created, so ask about matching rules in the same breath.

Whose activity is captured, and who needs a licence for it? This is the commercial question hiding inside a technical one. Nektar prices per user per month across both its packages and does not publish figures, and capture can reasonably extend beyond sales into success and leadership. People.ai charges per connected customer-facing person, with managers, finance, deal desk and ops using it without a seat, and pools an allowance of AI answers across the org. Adding CS or partner managers grows that bill faster than sales headcount does.

Can it backfill history? Forecast and relationship scores need a baseline, and Nektar publishes the one rate in this corner of the market: $8 per user for each month of past calendar and inbox data recovered, as of September 2026. Twelve months of history for 50 people is therefore $4,800 as a one-off, which is a real number to put in the business case rather than a surprise in the statement of work.

And what happens at renewal if usage grows? People.ai resizes the contract at renewal rather than shutting usage off or back-billing, which is fair but means heavy use by unlicensed viewers raises next year's price. Ask for a cap.

Conversation-led platforms

These products make the recorded call the unit of evidence. None publishes a list price, and the per-seat figure is only half the quote.

Gong is the one most teams benchmark against, with broad adoption across coaching, deal inspection and forecast support. Its model is per-user licences plus a platform fee charged on the number of users supported, with integrations included, and buyers pick a team-size band before a proposal arrives. Vendr reports a median contract of $55,333 a year across 1,141 purchases, ranging from $11,474 to $203,173. At the median a 50-seat team sits at $1,107 per seat per year and a 10-seat team at $5,533, because the platform fee sits on top of seats and does not scale down. Get that fee as its own line, ask whether it is re-rated when headcount falls, and ask what happens to recordings if you do not renew.

Chorus by ZoomInfo is the easier fit for teams already inside ZoomInfo. ZoomInfo publishes nothing, and Claap's July 2026 pricing research estimates a base of about $8,000 a year including three seats plus about $1,200 per additional seat, often on two-year terms, which would put a 10-seat team near $16,400 a year. Treat that as third-party research rather than a rate, and ask first whether Chorus features are already in your Copilot tier.

Salesken prices licences per recorded user plus a platform fee set by user count, and offers a choice between analysing 100% of calls and hourly analysis-based pricing, with a discount for a combined sales and QA plan. Ask for both rates in writing. Checked September 2026.

Forecast-led platforms

These are bought to change what managers see on a Tuesday. All quote per customer, all assume an annual term, and all carry an implementation line.

Clari is the structural benchmark for forecast inspection. Since the merger with Salesloft closed in December 2025 it sells through the combined company, with Clari Forecast continuing as a named product and Clari Copilot as the call-recording piece. Vendr describes annual subscriptions with per-user pricing for core modules, platform fees for enterprise features, and implementation services ranging from basic setup to multi-month deployments. We found no attributed contract median we would rely on, and public reviews describe it as expensive for smaller teams. Ask whether bundling with Salesloft's engagement or conversation products lowers the total, because that is the main lever the combined company has.

Aviso quotes on users and modules. Vendr reports a median contract of $73,662 a year, about $6,139 a month, which across a 50-user forecast is roughly $123 per user per month. Ask which modules are bundled and whether executives who only view the forecast are priced below full users. Terret, the product sold as BoostUp until its September 2025 rebrand, combines forecast workflow with revenue intelligence and added an AI agent layer at the rebrand that is a separate line to ask about.

Two more sit on the Salesforce side. Ebsta, acquired by Fullcast in August 2025, lists three quote-only plans (Conversation Intelligence, Relationship Intelligence and Revenue Intelligence) plus an Enterprise option, and if all you need is email and calendar sync to Salesforce, price lighter sync tools before buying the platform. InsightSquared has been part of Mediafly since January 2022, and the quote you get is likely a Revenue360 bundle rather than a standalone analytics licence, so ask whether the forecasting views can be bought without the enablement modules. Checked September 2026.

Lighter and hygiene-first picks

The cheapest useful purchase in this category is usually not an intelligence platform. It is whatever gets data into the CRM without rep typing, and two of these publish rates.

Scratchpad is the one to price first if managers spend their week chasing updates. Free covers three workspace views, 10 hours of call recording a month and 100 AI credits per user. Solo is $19 per user per month billed yearly with 400 credits, and Team is $49 with 800 credits pooled across the team and support for Gong. A 10-rep team on Team is $5,880 a year; on Solo it is $2,280, but credits are the real driver and heavy users will exhaust Solo mid-month. Its own worked example of five calls a day with four field updates each uses roughly 800 credits a month, the full Team allowance. It is Salesforce-centric, which decides eligibility for everyone else.

Nektar works the same end of the problem from the data side, with two quote-only per-user packages and the published $8 per user per month of backfilled history. Salesloft enters the category when teams want call insight and deal inspection inside their engagement platform, and since the merger its line-up includes Cadence, Conversation Intelligence, Deals, Rhythm, Clari Forecast, Inspect and Analytics; it publishes nothing and has not said how AI agent usage is billed, so ask whether agents are in the seat or metered.

The pattern is consistent across all three. Spend on capture and hygiene first, because every score the expensive products sell is computed from the records these tools fill in. Checked September 2026.

Revenue intelligence software at a glance

Fit, published price and the dimension that decides each purchase, as of September 2026.

Gong: recorded conversations as the evidence. Quote only; per-user licences plus a platform fee on users supported. Vendr median $55,333 a year. Decider: the platform fee as its own line.

Clari: forecast-led organisations wanting structured inspection. Quote only, sold through Salesloft. Decider: how much of the Salesloft bundle you must take.

Aviso: AI assistance on quarter predictability. Quote only; Vendr median $73,662 a year. Decider: whether view-only executives cost a full seat.

Chorus by ZoomInfo: teams already on a ZoomInfo contract. Quote only; Claap estimates about $8,000 a year for three seats plus roughly $1,200 per extra seat. Decider: whether your Copilot tier already includes it.

Terret, formerly BoostUp: RevOps teams wanting a Clari alternative. Quote only. Decider: how the AI agent layer is billed.

People.ai, now Backstory: activity and relationship data across sales, CS and channel. Quote only, per connected customer-facing person with a pooled answer allowance. Decider: who counts as a connected person.

Ebsta: Salesforce forecast teams wanting relationship scoring. Quote only across three plans. Decider: whether you need the platform or only calendar and inbox sync.

Salesken: coaching-led inside sales teams. Quote only, per recorded user plus a platform fee. Decider: whether every call needs analysing.

Scratchpad: Salesforce teams fixing rep update friction first. Free tier; Solo $19, Team $49 per user per month yearly. Decider: AI credit burn per rep.

Nektar: RevOps teams whose first problem is CRM data quality. Quote only; backfill at $8 per user per month of history. Decider: whose activity you license, not how many reps you have.

Pricing models and quote drivers

Three models dominate. Per recorded or connected person is the first: Gong, Salesken and People.ai all meter on who is captured rather than who logs in, which means the licence count grows when you extend capture to success or channel teams. Per user plus platform fee is the second, and the platform fee is the line buyers forget. Gong charges it on users supported, so it does not scale down for a small team, and Salesken sets it by user count. Quote-only with modules is the third, covering Clari, Aviso, Terret, Ebsta, InsightSquared, Chorus and Salesloft.

Only two published rates exist in the whole category: Scratchpad at $19 and $49 per user per month billed yearly, and Nektar's $8 per user per month of backfilled history. Everything else is negotiated, so third-party benchmarks are the only external anchor. The Vendr medians cited on the product pages here are $55,333 a year for Gong across 1,141 purchases, $73,662 for Aviso and $33,500 for ZoomInfo. Those are reported contract values rather than rates, and Gong's range from $11,474 to $203,173 shows why a midpoint is weak guidance alone.

Five things drive your number: licence count, the platform fee, which modules are in scope, the term and seat minimum, and the renewal uplift. Get all five in writing, plus two answers most buyers skip: what happens to your recordings and transcripts if you do not renew, and whether the platform fee is re-rated if headcount falls. A call archive you cannot export is a switching cost, not an asset. Our revenue intelligence RFP template holds these questions in the order a procurement process needs them, and how we check vendor pricing explains why a quote-only vendor gets no price table here.

Recording consent and rep trust

A revenue intelligence rollout fails on adoption more often than on features, and the failure is predictable. Reps who believe the tool exists to monitor them start managing what it sees, which degrades the data the whole purchase depends on.

Settle consent before procurement finishes. In two-party-consent jurisdictions every participant has to agree before a call is recorded, which means an announcement at the top of every call and a documented process for external attendees who decline. Your own counsel sets that policy, not the vendor's help centre. The question has teeth: the consolidated class action against Otter.ai in the Northern District of California turns on recording and training on conversations without the consent of participants who hold no account with the vendor, and on 13 August 2026 the court let federal wiretap, California privacy and Illinois biometric claims proceed. Liability is undecided, and the point for a buyer is that consent handling is a diligence item rather than a checkbox.

Then separate coaching from inspection. If the only time a rep hears about a transcript is in a performance conversation, the platform becomes a threat and call volume on the platform quietly drops. Publish what managers can see, say it once at rollout, and give reps something in return on day one: pre-filled next steps, a draft follow-up, a searchable record of what the buyer said.

Finally, decide what the tool writes without approval. The safe default is that it proposes CRM field updates for a human to accept, with close date, amount and stage staying human-owned. A model that nudges those three is editing the number leadership commits to.

Scoring revenue intelligence software

Run the evaluation as a scored comparison on one sheet, not a sequence of demos. Six criteria, weighted before the first call.

Capture coverage: which sources, whose activity, contact creation and matching rules, and whether history can be backfilled. Inspection depth: snapshot history, manager-owned forecast categories, and accuracy tracked by manager. Auditability of any risk score: what inputs drive it and whether a manager can see them. Licence model: who needs a seat and whether mixed tiers are allowed. Exit terms: export rights on recordings and enriched records. And services: implementation, data cleanup and partner hours.

In the demo, ignore the executive dashboard and make the vendor run a manager's Tuesday live: pull the deals with past close dates, show the last buyer-side activity and its date, change a close date, record a dated action, and put last week's snapshot beside this week's. Then run a paid pilot on one segment with a written success metric, because free trials in this category are scoped to hide behaviour at volume. Score vendors with our sales software evaluation scorecard so the criteria stay fixed instead of drifting toward whichever vendor demoed most recently.

Do not buy while stage definitions are unsettled, below roughly fifteen reps, or when the real problem is that nobody runs the weekly inspection. Read how to run a pipeline review first in that case, and if the only job you need is the number, our guide to sales forecasting software is the narrower comparison. The honest summary of how to evaluate revenue intelligence software: pick the one of four jobs that is failing, buy capture before scoring, and insist on seeing the platform fee, the seat definition and the exit terms in writing.

Buyer checklist

• Name which of the four jobs is failing today (inspection, conversation intelligence, activity capture, relationship data) before any demo.
• Evaluate activity capture first: sources read, whose activity is licensed, contact creation, matching rules and history backfill.
• Ask who needs a paid seat, whether view-only executives cost less, and whether mixed tiers are allowed in one org.
• Insist on the platform fee as its own line, and ask whether it is re-rated when headcount falls.
• Get export rights on recordings, transcripts and enriched records in writing before signing.
• Ask what inputs drive any AI risk score and whether a manager can see them, and keep the commit a human call.
• Set the recording consent policy with your own counsel, including a process for external attendees who decline.
• Separate coaching from inspection, and publish at rollout what managers can see.
• Cap the renewal uplift, and get the implementation and data cleanup estimate as separate lines.
• Run a paid pilot on one segment with a written success metric rather than a scoped free trial.

Related

Revenue Intelligence guides