VoC
VoC

From messy export to board-ready deck — in minutes, while you watch

Drop in your real customer-feedback exports, exactly as they come out of Trustpilot, Bazaarvoice, your NPS tool or the app stores — VoC auto-detects the columns, sources, markets and languages, and hands you a branded, interactive dashboard plus a downloadable board deck. No column-mapping, no setup project, no two analyst-days.

Last updated 20 June 2026

Most "feedback analytics" tools make you do the boring part first: name your columns, tag your sources, reconcile your markets, then wait. VoC flips it. You drop the raw, ugly files in — and the branded dashboard builds itself in front of you. This page shows exactly how, with the real artefacts, so you can answer the only two questions that matter before you trial: "will it handle my exports?" and "how much setup is this?" (Spoiler: yes, and almost none.)

14-day free trial · no credit card · per-seat. Your files are parsed in your browser, not uploaded.

The three stages, end to end

That's the whole flow. No integration to wire up, no schema to define, no implementation call. You bring files; you leave with a dashboard and a deck.

1 · Drop in the raw exports

Export from Trustpilot, Bazaarvoice, your NPS survey, Google reviews, the App Store and Play Store — whatever you've got — and drag the files straight in. CSV or XLSX, one file or a dozen, one market or fifteen. You don't clean, rename, or pre-sort anything.

Here's the kind of file we mean — a genuine Bazaarvoice ratings-and-reviews export, with the header sprawl nobody enjoys:

Review IDSubmission TimeRating RangeLocaleReview TextNet Promoter (raw)
BV-99204312026-04-03T11:42:09Z4 of 5de_DE"Lieferung war schnell, aber die Verpackung war beschädigt…"
BV-992051803/04/2026 14:085/5fr_FR"Service client impeccable, je recommande !"
BV-99210072026-04-042it_IT"App si blocca al checkout, molto frustrante."
BV-99211442026-04-05T09:15Z★★★★★nl_NL"Snel en netjes, top winkel."

Mixed date formats, four ways of writing a rating, five languages, and a half-empty NPS column from a different survey — all in one file. This is exactly what VoC is built to eat.

2 · Zero-config auto-detection maps everything

The moment a file lands, VoC reads the shape of it. It works out which column is the rating, which is the comment, which is the date, which is the market, and which source the file even came from — then normalises the ratings, parses the dates, and groups it all by market and language. You watch it happen; you don't configure it.

  • Source recognition — Trustpilot vs Bazaarvoice vs NPS vs app-store formats are detected from their fingerprints, not from you tagging them.
  • Column mapping — rating, comment, date, country and language columns are found and aligned across files that name them differently.
  • Markets & languages — DE / FR / IT / PL / NL / DA and more are split out so each market reads on its own and rolls up to a group view.
  • Format logic — review-star scales and NPS 0–10 scales are handled correctly, including the net-base maths (promoters minus detractors on the right denominator).
  • Dedup — re-exports and overlapping date ranges don't double-count; the same review loaded twice stays one review.

3 · Your branded, interactive dashboard — and the deck

Out comes a Voice-of-the-Customer dashboard in your brand: your logo, your colour, your name on the cover. It's interactive — drill from the group number into a market, into a source, into the comments behind a score — and it writes the so-whats for you, in plain language. One click exports the whole thing as a board-ready deck: an exec summary, a slide per source, and the headline movements, every number deck-exact.

This is the "aha": the first time you see your own dashboard, built from your own messy files, you stop wondering whether it'll work for you.

What auto-detection handles — so you don't

The column-mapping wizard is the part of every other tool that quietly costs you a day. Here's the work VoC absorbs versus the work that lands on you.

The messy reality in your exportsWhat VoC does automaticallyWhat you do
Different sources, different layouts (Trustpilot, Bazaarvoice, NPS, Google, app stores)Recognises each source from its structure and maps itDrop the file in
Columns named differently in every fileDetects rating / comment / date / market columns and aligns themNothing
Comments in DE, FR, IT, PL, NL, DA…Keeps languages and markets separate, rolls them into a group viewNothing
Reviews on a 1–5 star scale and NPS on 0–10 in the same setApplies the right scale and the correct net-base survey maths to eachNothing
Re-exports and overlapping monthsDe-duplicates so nothing is counted twiceNothing
Inconsistent dates (ISO, DD/MM/YYYY, partial)Parses them into one timelineNothing

Net effect: the setup step that's normally a config project becomes a drag-and-drop. If a column ever can't be read, VoC tells you in plain language rather than failing silently — but for standard exports from these sources, you won't touch a mapping screen.

What happens to your text — stated plainly

This is the part most tools are vague about, so we'll be exact.

  • Your raw files are parsed in your browser. The CSV/XLSX you drop in is read and analysed locally, on your machine. The files themselves are never uploaded to our servers.
  • The AI never sees raw comments or personal data. The plain-language so-whats are written by an AI step — but it only ever receives aggregated statistics plus comments that have been redacted first: emails, links, IBANs and long number runs (phone, order and customer numbers) are stripped out before anything is sent.
  • Nothing is used to train a model. The redacted, aggregated input is used to generate your narrative and nothing else. Your customers' words are not added to anyone's training data.
In one line: raw files stay in your browser; only redacted aggregates leave it, only when you ask for the AI so-what, and the AI never sees a raw comment or a piece of PII. This is the same boundary an EU/GDPR data team would ask us to draw — so we drew it by default.

Want the full detail — sub-processors, retention, the DPIA? See Security & data protection.

Raw rows in, the so-what out

The point of all the parsing is this: turning a wall of rows into the one sentence a leadership team can act on. Same data, two views.

What's in the export (raw rows)What the engine writes
237 IT reviews mentioning "checkout / app crash / blocca", avg rating 2.4, down from 3.6 last month; concentrated in the iOS app Italy's score fell a full point this month, and it's the iOS checkout crash driving it — 237 reviews name it. Fix it and IT recovers.
NPS DE +41 (promoters 58%, detractors 17%); top promoter theme "Lieferung schnell"; top detractor theme "Verpackung beschädigt" German shoppers love the speed and dock you on damaged packaging — your detractors are almost entirely a fulfilment problem, not a product one.

What you get

  • Executive summary — the three or four so-whats, the headline movements, in plain language.
  • Per-source slides — Trustpilot, Bazaarvoice, NPS, app stores: one clean view each, every number deck-exact.
  • Interactive dashboard — drill from group → market → source → the actual comments.
  • Board-ready deck — download the whole thing as slides, branded, ready to present.

Time to first dashboard

Minutes of clock time — not two analyst-days.

The build runs while you watch; there's a timer on screen so you can see it. In our pilot with a leading European omni-channel retailer, eight monthly exports — tens of thousands of responses — become one branded performance in about three minutes. The honest version: it's parsing, mapping and charting fast, not magic. You'll see exactly what it did and every number it stands on.

Set the expectation correctly: the work that used to be a two-day analyst job is now the length of a coffee. The judgement is still yours — VoC just removes the grind before it.

See it on your own data

You can read about it, or you can drop a real export in and watch your branded dashboard appear. The second one is more convincing.

14-day free trial · no credit card · per-seat · EU/GDPR-aware · files parsed in your browser.

Start free — no credit cardSee a sample dashboard