Skip to content

Why don’t brands trust their market data?

The question I hear most in a first meeting is a simple one: where does this data come from?

It’s a fair question. It’s also a revealing one. Nobody asks it because they think we’ve made the numbers up; they ask it because they’ve spent years being handed numbers they had no way of checking.

The report that arrives finished

Think about how a legacy category report lands. Someone commissions it, a provider disappears for four weeks and then a PDF arrives. It’s polished. It’s confident. It’s also describing a market that has already moved.

Then it goes into a range review, which is where the trouble starts. The buyer asks why a competitor is missing. The category manager points out that a price looks wrong. Someone asks how the market share figure was worked out; the honest answer is that nobody in the room knows.

The report isn’t necessarily wrong. The problem is that nobody in that room helped build it, so nobody can defend it. Every figure has to be taken on trust, with no visibility of how the numbers were generated.

I don’t think that’s a data problem. It’s an ownership problem: people trust what they can check and they defend what they helped shape.

Start with their competitive set, not ours

Given that people trust what they helped shape, the first thing we do with a new customer is hand them the set-up of their data. In Competitor Intelligence the recommended route is called ‘Built around you’. It does what it says: the customer defines its own competitive set.

They give Rove a product description, keywords, a price range and any filters that matter. They approve, reject or reorder the products Rove suggests; they pin their own product at the top.

It takes longer than the quick option that Rove also offers: up to 15 minutes rather than about two. We lead with it anyway, because the effort a brand puts into set-up is part of the value. A competitive set someone built themselves is one they’ll stand behind in a meeting.

Rove even shows its working on what it leaves out. When the search finishes, the brand sees how many products were checked and left out, grouped by reason: a requirement that differed, an accessory or part rather than the product itself, a price outside the range. If a competitor they expected isn’t there, they can see why instead of wondering.

Every change should be checkable

The second thing is verification. In both Competitor Intelligence and Competitor Monitoring every change Rove reports is checked against ground-truth sources and timestamped. When a competitor drops its price, the alert says what changed and when; if a customer doesn’t believe it, they can open the retailer’s own page and look. That holds across multiple Amazon marketplaces and 64 other retailers, down to loyalty prices such as Tesco Clubcard sitting next to the standard price.

We’d much rather a prospect checked than took our word for it. That’s why I run demos on the prospect’s own category, with products they already know. When the prices on screen match what they can see on the retailer’s site, the conversation changes: they stop asking where the data comes from and start asking what to do about it.

Say what’s estimated and where the data stops

Most of what Rove shows is observed, at SKU level: rank, price, rating, reviews, stock and Amazon’s own “bought in past month” figure. One figure that is estimated is each product’s share of its tracked set’s sales. It’s always labelled “Estimated”, it always comes with a likely range and the platform explains how it’s worked out. A thin shelf gets a wider band and a warning rather than false precision.

The same applies to what an Amazon shelf is worth. Rove gives a range of monthly customers and revenue, from floor to ceiling, never one tidy number; it also shows each listing and the arithmetic behind it.

Where the evidence doesn’t support an estimate, there’s no figure at all. If a retailer doesn’t publish a number, Rove doesn’t invent one.

To some, that might sound like a weakness in a sales conversation. In my experience it’s the opposite; when a single number arrives with no range and no method, buyers have learned to assume it might be modelled. Showing exactly where the observed data ends and the estimate begins is what makes the rest believable.

Recommendations you can trace back

The same rule applies to what Rove recommends. The actions a tracker suggests are decided by fixed rules, using the shelf’s own numbers: a rank fall while priced above the shelf median, a rival running a deal where the brand has none, a rival that’s just overtaken them, a drop in rating. Every action names the product and shows the evidence behind it.

What this means for how I sell

It means I don’t sell a report. I don’t arrive with a finished deck about someone else’s category and ask a buying team to believe it.

Instead, the demo runs on their category; the trial runs on a competitive set they built; the return on investment is written up from their own numbers within the month. Buyers purchase on evidence from their own category, so that’s where the evidence has to come from.

It also means I welcome the awkward questions. Where does this come from? Why is that competitor missing? Is that figure measured or estimated?

Each one has an answer in the product.

A report handed down asks for trust. Data you helped build has already earned it.

Get the insight you need to move faster

See how Rove helps your team understand competitors, markets and customers, with less effort and more clarity.

Get started free