A claim inventory is defined as a list of the distinct assertions a set of competing stores make, separated from the sentences carrying them and scored for specificity and uniqueness. That inventory is how to benchmark your messaging against competitors, because reading competitor copy side by side tells you almost nothing: the claims stay buried inside well-written sentences and the crowding is invisible. Take the same three surfaces from each of five competitors, write every assertion as its own line, mark whether it names something measurable, and tally how many stores make it. The output is a map with two useful features: the places where everyone is shouting the same thing, and the specific true claim nobody is making.
- Extract claims, not copy. One marketing sentence routinely carries three claims.
- Score specificity: a number with a unit, a named material, a stated time or a cited source.
- Tally crowding. A claim everyone makes cannot differentiate you however well written.
- Include one competitor positioned differently. They usually hold the uncrowded claim.
- Repeat twice a year and keep the old inventory. The drift between passes is the finding.
To benchmark your messaging against competitors usefully, stop comparing copy and start comparing claims. Omniconvert has measured commercial messaging and page performance across 70,000+ experiments and 13 years in eCommerce, and the finding that transfers most often is that stores in a category converge on the same four or five assertions without any of them intending to. Nexus by Omniconvert is the AI for eCommerce growth engine that unifies commerce data, prioritises experiments by True Profit and generates campaigns and creative you approve before they go live, and the approval step is the relevant one here: a generated headline is only as good as the claim underneath it, which is a decision a person has to make. Last updated: October 2026.
This article covers the language axis. For the whole set of axes, see the DTC competitor research stack, and for the commercial axis, analyse a competitor's pricing and offers.
Why you cannot benchmark your messaging by comparing copy
The problem with a side-by-side read is that it invites a judgement about quality rather than about position.
Put five home pages next to each other and a team will rank them on tone, design and polish. Those conversations feel productive and they end in a brief to write better copy, which is the one output that cannot be evaluated. Nobody can say afterwards whether the new copy was better.
Extraction changes the unit of analysis. Once every assertion is a line in a list, the structure of the category becomes arithmetic. Four of five competitors claim fast delivery. All five claim quality materials. One claims a specific tested durability figure. Two make no claim about returns at all. That is a map, and a map supports a decision.
It also exposes your own position honestly. Most teams discover their differentiating claim is one four competitors also make, which is uncomfortable and more useful than any amount of tone comparison.
The three surfaces to collect, and why only three
Scope discipline is what makes this a two-day task rather than an abandoned one.
The home page headline block. The claim a store leads with, chosen deliberately and usually argued over internally. This is their stated position.
One comparable product page. The claims they make when a shopper is close to deciding, which are frequently different and more specific than the home page ones. Pick a product that exists across the set.
One paid advertisement. The claim they are paying to put in front of cold audiences, which is the strongest signal of what they believe works. Meta's Ad Library is the public record for this and is the fastest route to a competitor's current creative claims.
Three surfaces across five competitors is fifteen documents, which yields perhaps a hundred and fifty claims and reduces to thirty or forty distinct ones. That is a readable inventory. Collecting category pages, email sequences and social posts as well triples the work and rarely changes the conclusion.
Scoring specificity and crowding
Both scores are deliberately crude, because a finer scale drifts as the person applying it changes their mind.
Specificity has three levels. Measurable, meaning a number with a unit, a named material or a stated time. Sourced, meaning an external citation or a testable reference. Assertive, meaning an adjective only the brand applies. Most claims in most categories land in the third level.
Crowding is just a count across your five competitors plus yourself. No weighting, no judgement about how prominently each one makes it.
Then the four quadrants. A specific uncrowded claim is the prize and is worth testing as a headline immediately. A specific crowded claim is a cost of entry you must match rather than lead with. A vague uncrowded claim is usually empty rather than available, and worth checking whether there is a measurable version underneath it. A vague crowded claim is category noise, and the discipline is to stop spending headline space on it.
| Quadrant | What it looks like | What it means | Action |
|---|---|---|---|
| Specific, uncrowded | A measured fact nobody else states | Your available position | Test it as the headline now |
| Specific, crowded | A figure everyone publishes | Cost of entry in the category | Match it, do not lead with it |
| Vague, uncrowded | An adjective nobody uses yet | Usually empty, not available | Find the measurable version first |
| Vague, crowded | Premium, trusted, award-winning | Category noise | Stop spending headline space on it |
| Absent from the category | A claim no store makes at all | A gap or an irrelevance | Check it matters to buyers first |
| Yours alone and unverifiable | A claim only you make, with no proof | A liability under scrutiny | Substantiate it or retire it |
The last row is worth dwelling on, because it is the one this exercise most often surfaces about your own store. A unique claim you cannot substantiate is weaker than a crowded claim you can.
Checking the gap is real before you move into it
This is the step that separates a positioning decision from a clever observation.
A category can be silent about something for a good reason: nobody buying in it cares. Finding that no competitor mentions a product property is therefore two hypotheses, not one, and the cheaper of them to test is whether your own customers mention it unprompted.
The sources for that check are already in your business. Review text, survey responses, support tickets and the questions your sales or service team answers repeatedly. If a property appears in customer language and in nobody's marketing, that is a genuine gap. If it appears in neither, the silence was correct.
Nielsen's work on consumer trust in peer recommendation is a useful external reference for why a claim corroborated in customer language outperforms one asserted only by the brand, and it points at the practical move: lead with the claim your customers already make about you.
Then test it rather than adopting it. Omniconvert Explore is the CRO platform for running the new headline against the current one with A/B and multivariate testing, on-site surveys and recordings, and Explore averages a 23.2% conversion uplift across 70,000+ experiments. A positioning change validated against a control is the only kind you can defend in six months.
What a growth team should do this week
- Pick five competitors. The two you are compared with, two growing, one clearly positioned differently. Write down why each is in the set.
- Collect three surfaces each. Home page headline block, one comparable product page, one current advertisement. Fifteen documents, one morning.
- Extract and score. Every assertion on its own line, with a specificity level and a crowding count. Expect thirty to forty distinct claims.
- Check the gap against customer language. Reviews, surveys, support themes. A gap your customers never mention is not a gap.
- Get your free benchmark score. Ecommerce Benchmark scores your store across six dimensions, including AI Visibility and Agentic Commerce, against real competitors in your category and country, which shows whether messaging is the weakest of the six.
Where review text, survey responses and support themes sit in separate systems and have to be assembled before any of this can be checked, that unification is what Nexus addresses, with every proposed campaign staying something a person approves before it goes live.
FAQ: benchmarking messaging against competitors
How do you benchmark messaging against competitors?
Extract the claims rather than comparing the copy. Take the same three surfaces from each competitor, write out every distinct assertion as its own line, then score each one for specificity and count how many competitors make it. The output is a map of where the category crowds and where it is empty, which is information a side-by-side reading of competitor copy never produces because the claims stay buried inside well-written sentences.
How many competitors should you include?
Five is the working number: the two shoppers compare you with most often, two that are growing, and one that is clearly positioned differently. Below four the crowding counts are meaningless, and above seven the extraction becomes a project nobody finishes. The odd one out matters more than people expect, because it is usually the store making the claim the rest of the category has not thought to make.
What makes a claim specific rather than vague?
A number with a unit, a named material, a stated time, or a cited source. Fast delivery is vague and dispatched within four hours is specific. Premium materials is vague and a named fabric with a weight is specific. The test is whether a shopper could be disappointed by the claim: if nothing could falsify it, it carries no information, and a whole category of such claims is why so many stores read identically.
Should you copy a competitor claim that seems to work?
Only if it is true of you and not already crowded, which is a narrow window. Adopting a claim four competitors already make adds you to a group rather than distinguishing you, and shoppers discount a claim in proportion to how often they have just read it. Copying also forfeits the real prize of this exercise, which is the empty space: the specific true thing about your product that nobody in the category currently says.
How often should this be repeated?
Twice a year, and keep the previous inventory. Messaging moves slowly compared with price, so a quarterly repeat mostly records noise. What the second pass gives you is drift: which claims the category has adopted since the last reading, which have been abandoned, and whether the gap you found has been filled. That comparison is the output, which is why the first inventory is worth filing carefully even if it changes nothing immediately.
The bottom line
Messaging comparisons fail because they compare the wrong thing. Copy is packaging, and ranking five competitors on tone produces a brief to write better sentences, which nobody can evaluate afterwards. Claims are the content, and once they are extracted into a list with a specificity level and a crowding count, the category stops being a matter of taste and becomes a map. The map almost always says the same two things: that your leading claim is one several competitors also make, and that there is a specific, true, verifiable thing about your product that nobody is saying. Check that second thing against your own customers' language before you build on it, because a silence in the category is not automatically an opening. Then test it against your current headline rather than simply adopting it.