The vendors in this category do not do the same job, and the label hides it. Five of the eight start with speech recognition; three never touch audio. This list is organised by what each one actually listens to.
The short answer
“Conversation intelligence” covers two different products under one name. For sales calls, Gong. For contact-centre calls and agent QA, CallMiner, Observe.AI, Cresta or Level AI. For written conversations — support tickets, chats and reviews — Chattermill, Enterpret or Indellia, which do no speech analysis at all. Decide which conversations you mean before you compare anything, because a speech vendor and a text vendor are not substitutes for each other.
A numbered “best” list implies one winner, and in this category that would be misleading. Gong is not better or worse than Chattermill; it listens to a different thing for a different buyer. So the order below groups by input — sales calls, then contact-centre calls, then written conversations — and each entry says who it is for.
Facts come from each vendor’s own site, read in Q4 2026. Where a vendor does not publish something, this page says so rather than estimating. Indellia appears eighth because it covers the narrowest slice of the category, not because the list is a ranking.
The default for revenue teams. Gong records, transcribes and analyses sales conversations to coach reps and forecast deals. It is the most searched name in this whole category — roughly 74,000 US searches a month — and if your question is about pipeline rather than product, this is the category leader and the rest of this list is irrelevant to you.
Best for: Sales and revenue teams. Not a customer-feedback tool.
One of the oldest speech-analytics vendors in the market, aimed at contact centres with large call volumes. Deep acoustic and transcript analysis, compliance monitoring, agent scoring. If your conversations happen on the phone and volume is high, this is a serious incumbent choice.
Best for: Enterprise contact centres with phone-first volume.
Speech analytics plus agent coaching and quality assurance, built around the contact-centre workflow rather than around analytics for its own sake. Positions on automating QA across every conversation rather than sampling a few.
Best for: Support organisations that score agent performance.
Real-time rather than retrospective — Cresta aims at coaching the agent during the conversation, not reviewing it afterwards. A genuinely different product shape from the analytics-first vendors, and a different buying case.
Best for: Contact centres wanting in-the-moment agent assistance.
Quality assurance and analytics across support conversations, with scoring against a rubric as the central workflow. Sits in the same neighbourhood as Observe.AI and evaluagent.
Best for: Support QA teams moving off manual sampling.
Text rather than speech, and the broadest review-source list in this category: Bazaarvoice, Trustpilot, G2, Capterra, Yelp, Google Reviews, Reviews.io, Yotpo, Feefo, Tripadvisor, ConsumerAffairs and more, alongside Zendesk, Intercom and the app stores. Describes itself as an AI-native Voice of Customer platform “built for enterprises and the agentic era,” with Lyra Agent as the headline and an MCP server for assistant access. Publishes SOC 2 Type II and ISO 27001.
Best for: Enterprise CX teams with feedback across many text channels.
Also text-first, and the one vendor here that weights what it hears by commercial consequence — “every request is weighted by the accounts and revenue behind it.” Repositioned recently to “customer context for AI-native product teams,” with Enterpret Agent and an MCP server listed in the Claude Connector Directory. Publishes SOC 2 Type II, ISO 27001, ISO 27701 and ISO 42001.
Best for: Product-led B2B SaaS with a feedback-operations function.
Ours, and placed honestly: Indellia has no speech analytics. There is no ASR, no call recording and no agent scoring, so for anything phone-shaped the five vendors above are the right answer. What Indellia does is the written half — retail product reviews from Amazon, Walmart, Best Buy, Costco, Lowe’s, Target and Bazaarvoice-powered retailer pages, plus support tickets and returns — with every record resolved by the SKU Agent to one physical product across ASIN, Item ID, Model# and UPC. $495/mo SME and $1,995/mo Mid-Market, published, with an MCP server for Claude, ChatGPT and Cursor.
Best for: Consumer brands selling physical products through retail.
1. Which conversations do you actually have? If most of your customer contact is on the phone, the text-first vendors cannot help you and no amount of feature comparison changes that. If your customers never call you but leave thousands of reviews, the speech vendors are the wrong shelf entirely.
2. What is the unit of analysis? Ask every vendor to name it. The call, the ticket, the account, the theme, the location, the product. This single answer predicts whether the platform will be able to tell you what you want to know, and it is rarely on the marketing page.
3. Will the same question return the same number next month? If a figure is going into a leadership deck, it has to reproduce. Ask how, and ask what happens when the taxonomy changes underneath it.
The text-first comparison, including where Chattermill’s coverage is broader.
Account-weighted feedback versus SKU-level feedback.
The wider VoC field, compared for consumer brands.
Verified 2026-11-10. Every claim about another vendor on this page comes from that vendor’s own site:
www.enterpret.com/pricing — pricing; returns 404 — no pricing page at the top level.Indellia’s own rates are on the pricing page. We do not quote unnamed “industry benchmarks” for a competitor’s price; where a figure comes from a third party it is named in the sentence that uses it. If something here is out of date, it is wrong — tell us and we will re-read it.
Software that analyses the conversations a company has with its customers — calls, chats, tickets, reviews — and turns them into themes, scores and alerts. The term started in sales-call analysis and has widened since, which is why the vendors in this category do genuinely different jobs. The useful question is never “which platform is best” but “which conversations do you mean.”
Originally yes, and for Gong, CallMiner, Observe.AI, Cresta and Level AI it still does — those products start with speech recognition. The text-first group — Chattermill, Enterpret, Indellia — analyses written conversations instead: tickets, chats, reviews. A vendor from one group is usually a poor substitute for one from the other, whatever the category label says.
Of the platforms here, Chattermill reads the widest set of review sites and Indellia reads retail marketplace listings directly with SKU-level resolution. The speech vendors do not ingest product reviews at all. If product reviews are your main source, start with those two and ask each for its source list in writing.
Partly, and we would rather be precise than claim the category. Indellia analyses written customer conversations — support tickets and product reviews — so on the text half of this category it competes directly. It does no speech analytics whatsoever: no call transcription, no acoustic analysis, no agent scoring. If a vendor tells you it does everything in this list, ask which of these eight products it has replaced for a real customer.
Only one of the eight publishes a price. Indellia publishes $495 per month SME and $1,995 per month Mid-Market, flat and unmetered. The other seven are quote-only as of Q4 2026, and the paid-search data gives a hint why: the most expensive term we priced in this category was “ai for customer experience” at $188.94 a click against 590 US searches a month, which implies contract values that support an enterprise sales motion.
Three questions settle most of it. Which exact sources do you ingest, named, in writing? What is the unit of analysis — the call, the ticket, the account, the location, or the product? And what does a repeat of the same question next month return, to the digit? The third one catches more vendors than the first two.
Indellia's AI agents answer with citations from real customer feedback across Amazon, Walmart, Best Buy, and 20+ retail channels.
Indellia reads retail product reviews and support tickets, resolved to the SKU. Start the free trial, or run one ASIN through the analyzer first.