They hear your name. We hear the market.
Listening tools start from a keyword you register. Verbatim starts from the bigger market you sell into: the comment threads, the Reddit threads and what’s said to camera around your kind of product, whether or not anyone types your name, and where your rivals stand in it. Here is what happens, in three parts.
Every example on this page comes from a made-up hiking-backpack market: invented brands, quotes and numbers, there to show what Verbatim does. None of it is a client’s data.
Gather. Where it listens.
Every week Verbatim goes out to four platforms and comes back with what people said around your category: the videos, the threads underneath them, and what was said out loud.
- Videos found by search, and the comment thread under each
- The platform’s caption track when there is one, transcribed otherwise
- Reels found by search, and the comments under each
- The audio transcribed, so a spoken review counts
- Videos found by search, and the comment thread under each
- The caption track, read alongside the comments
- Posts and comments in the communities where your category is discussed
- Communities found per category, not hand-picked
Three lists seed every search.
Your brand, your competitors, and the category. Every video that comes back is sorted by who it’s actually about, and a video that names a lookalike brand is checked before it’s counted as yours.
Competitor and category terms are search terms only. Attribution comes from the names, confirmed one video at a time.
Finding the communities.
For Reddit, likely communities are proposed, then probed with a sample of real posts. One only goes live if it keeps talking about your category, and it drops out again when it goes quiet.
Everything goes through the filter.
Before a comment is analysed it passes three checks: is the video about your category, is the brand it names really the one it seems to be, and is the comment a real person actually saying something. Off-topic videos go, lookalike brands go, spam is set aside. Short comments stay. “too small” is an opinion.
Anyone who asks to be removed is taken out within seven days and never gathered again.
Analyse. How it reads.
The analysis runs in passes, each one building on the last, and every pass keeps the quotes it stands on. Four stages turn thousands of comments into a model of your market you can act on.
Read each video with its thread.
What kind of video it is, how the audience received it, and everything people volunteer in the comments: pain points, questions, buying intent, objections, praise, the moment someone switched. The transcript is read too, so a spoken claim counts.
Each insight keeps its exact quote. A quote that doesn’t match the source word for word is dropped, and the insight goes with it.
Find the themes, and keep them.
Insights are grouped by meaning within each bucket: yours, each competitor’s, the rest of the category. A theme gets its own place on the page once it’s in 10 or more of the month’s videos; below that it’s listed by its count alone.
Themes keep their identity month to month, so this month’s count can be set against a later one. Until there are months that compare fairly, a theme says how big it is and nothing about which way it’s heading.
Compare you to the field.
With the themes in place, the buckets are read against each other: how many videos each brand comes up in, and for each rival you track, what people do under its videos, the questions asked there, and where its talk stands out from the category.
Mood is read video by video, and shown as a share of videos, never as a score.
Analyse the market.
From all of it: the kinds of person in the conversation, the market insights, and ranked recommendations grounded in the quotes retrieved for them. A separate check compares what you say in your own videos with what the audience hears back.
Everything is said in calibrated words. A count of videos, with what it’s out of, and on each conclusion a plain label: strong evidence or early signal. No scores, no magnitudes the data can’t back.
Deliver. What you see.
The model lives in the app, one page for each question you bring to it. Whatever someone else in the company needs goes to them as a report.
The app.
Every page is one view of the same model, and its counts open to the videos and comments behind them.
This week.
Each weekly update reads the new videos and comments into your market, and This week says what that update brought: what came in, what was heard for the first time, what stood between buyers and a yes, and the comments worth a reply.
A week is read as part of its month, never as a movement on its own.
Reports for your team.
The same market, written for the people who act on it, so the ones who never open the app still hear what the market said. Every number in a report is the number in the app, and every quote traces to the comment it came from.
Two different tools.
- Starts from your name.
- Reads what’s said when you’re mentioned.
- Tells you when you come up.
- Returns a dashboard for you to make sense of.
- Starts from the market you sell into.
- Reads the threads underneath, where people say what they really think.
- Tells you what the market you sell into wants, and where your rivals stand in it.
- Returns a market already read: what it talks about, the brands in it, what to do, and an analyst you can question.
If you need to know the moment you’re mentioned, choose a listening tool. If you need to understand your market, choose Verbatim.
Stop listening for your name.
Verbatim runs every week on a real brand’s market today. A few more brands join this quarter.