Competitive intelligence is the systematic analysis of competitor activity to inform a company's own decisions. In e-commerce specifically, this means analysing what competing brands sell, at what price, with what promotion, with what review reception, on which platforms — and how those patterns change over time.
The discipline operates entirely on publicly observable signals. E-commerce makes this possible at unusual depth: marketplace listings, public review counts, sales rankings, public KOL collaborations, regulatory filings, and corporate announcements together form a much richer surface than was historically available outside of dedicated panels.
Competitive intelligence is one of the four pillars described in Market Intelligence. It draws on the same data streams as the other pillars but reads them through a competitor-centric lens.
What it covers#
Five operational areas are typically in scope:
Sales and share analysis. Volume, revenue, and category share at the brand and SKU level. Who is gaining share, and where is that share moving from?
Pricing and promotion. List pricing, average selling price after promotion, promotional cadence by season and channel. Pricing patterns reveal strategy: a competitor that holds price through a category-wide promotion is signalling something different than one that discounts aggressively.
Product portfolio. New product launches, line extensions, formulation changes, packaging changes. Often the leading indicator of a category move.
Marketing and KOL activity. Which KOLs and KOCs competitors partner with, what messaging they are running, when. The Social Listening surface is where most of this lives.
Reception and reputation. Review sentiment, social mentions, share of voice. Where consumer perception is shifting and why. This overlaps heavily with Consumer Insights.
Granularity matters#
The difference between useful competitive intelligence and a noise-heavy dashboard is granularity matched to the question.
- Category sizing needs aggregate data — total category sales, growth rate, share concentration
- Defending share against a specific rival needs SKU-level data — which specific products of theirs are gaining at our expense
- Pricing decisions need promotional-period data — what is the actual transacted price, not the listed price
- R&D prioritisation needs attribute-level data — which features and ingredients are showing up in winning launches
A platform that only goes to brand-level breaks down the moment you ask a SKU-level question. A platform that goes to SKU-level can roll up to brand or category trivially.
Reading competitive signals correctly#
Three common mistakes:
- Confusing share of voice with market share. Loud is not the same as winning. Share of voice is a leading indicator that sometimes precedes share movement and sometimes does not.
- Misinterpreting promotional volume. A competitor with 30% volume growth during a flagship promotional period is not necessarily winning — they may be discounting unsustainably.
- Ignoring methodology coverage. A dataset that omits a major platform you compete on creates blind spots that are systematic, not random.
Methodology transparency on what is included and excluded matters more here than in most other applications.
How it shows up in decisions#
- Brand managers use it to defend share, identify whitespace, and benchmark performance
- Product managers use it to spot feature gaps and time launches
- Pricing teams use it to set position and respond to competitor moves
- Channel managers use it to allocate resources to platforms where category is moving
- M&A teams use it to validate or invalidate target growth claims
Common questions#
What can't you legally or ethically infer about a competitor from public data?#
Three categories typically sit out of scope. Internal cost structure and unit economics — list price minus promotion gives you transacted price, but never margin. Pre-launch product specifications — observable once a product is live, never before. Private commercial-agreement details such as platform-specific listing fees, take-rates, or co-marketing terms. Any tool offering visibility into these from public data alone should be approached with scepticism — a request for the underlying methodology will usually clarify quickly whether the claim is supportable.
What's the most common misread of a competitor's e-commerce signal?#
Confusing promotional volume with sustainable growth. A competitor that posts a strong sales lift during a flagship promotional period is often discounting unsustainably — the volume is real but the margin and the brand-equity cost are not visible in the same data. The cleaner read comes from comparing list pricing versus average transacted price across the period and watching whether promotional intensity is creeping up over multiple cycles. Repeated heavy promotion usually signals a defensive position, not strength.
What can SKU-level data tell you that brand-level data can't?#
Brand-level rollups hide the within-brand winners and losers. A brand growing 8% YoY at the brand level might be growing 40% on its hero SKU and shrinking on its older lines — radically different strategic implications. SKU-level data also exposes which specific variants (size, format, packaging, flavour) are pulling category preference, which is the level decisions actually get made at in product and pricing meetings. Aggregate platforms can answer "are they growing"; SKU-level platforms can answer "where is the growth coming from and is it durable."
How is publicly-derived competitive intelligence different from purchased sell-through panels?#
Sell-through panels rely on participating brands or retailers submitting their own sales data; coverage is limited to who agrees to participate, and the data flows on the participants' timeline. Publicly-derived intelligence works from observable e-commerce listings, public review counts, and public social conversation — the brand being observed has no involvement, and coverage extends to anyone selling on the relevant platforms. Panels go deeper for participants; public-signal datasets go broader, especially for the mass and online-prestige tiers where panel coverage is thinnest.