The Agency Audit Questions We Ask When Reviewing Competitor Work

Competitor audits have a bad reputation, and largely deserve it. Most are a deck of screenshots, a table of numbers nobody can verify, and a recommendation to do roughly what the competitor is doing but slightly more so.

We run them differently, mostly because we have had to defend the findings in rooms where somebody knew the market better than we did. Here is what we ask, and — more usefully — what we refuse to claim.

First: we do not use industry benchmarks

The traditional centrepiece of a competitor audit is a comparison table. Your bounce rate against the industry average. Your session duration against the benchmark. Your conversion rate against what firms like yours supposedly achieve.

Almost none of those figures survive scrutiny. They come from analytics vendors summarising the accounts on their own platform, with an undisclosed sample, an undefined sector and no published methodology — and in any case you cannot see a competitor’s bounce rate, so the comparison is between one real number and one invented one. We stopped using them because we could not answer the obvious question about where the benchmark came from. If we cannot tell you where a number came from, it does not go in the report.

What we use instead are the two standards that are actually published, defined and measurable from outside. Google’s Core Web Vitals guidance sets out what “good” means — a Largest Contentful Paint within 2.5 seconds, an Interaction to Next Paint of 200 milliseconds or less, and a Cumulative Layout Shift of 0.1 or less, all at the 75th percentile of real page loads. And WCAG sets out what accessible means. Both apply equally to you and to your competitor, both can be measured on a site you do not own, and neither is anybody’s marketing.

The six questions we actually ask

1. What does the home page say in the first ten seconds?

We open the site cold, read only what is visible without scrolling, and write down what a stranger would now believe about the business. Not what it looks like — what it claims.

This is the highest-value question because it is where most sites lose people. Writing in September 2011 on an analysis of more than 205,000 pages, Nielsen Norman Group concluded that users often leave within 10-20 seconds and that to hold attention beyond that you must communicate your value proposition within 10 seconds. If a competitor is spending that window on a mood statement, that is an opening. If they are using it to say something concrete that our client cannot match, that is a problem worth naming early.

2. Is it fast, measured properly?

On a mid-range phone, on mobile data, not on the office connection. We check both sites the same way, on the same day, on the same pages.

Speed is the one difference here with money attached to it. In Milliseconds Make Millions, research commissioned by Google and run by 55 and Deloitte across 37 brands, a 0.1 second improvement in mobile speed moved 3.2% more retail shoppers from listing to product page and 9.1% more from product page to basket, with retail spend up 9.2% and travel bookings up 10%. Retail and travel, on mobile — we do not stretch it further than that. But if a competitor clears the thresholds and our client does not, we know what to recommend first and roughly what it is worth.

3. Who cannot use it?

We navigate both sites with the keyboard alone, check colour contrast, and run an automated accessibility scan on the key pages.

This one usually finds problems on both sides. The WebAIM Million survey for 2026 found detected WCAG 2 failures on 95.9% of the top one million home pages, averaging 56.1 errors per page — and automated tools only catch the failures software can see. So it is rarely a differentiator. It is, however, a cheap and uncontested advantage: a market where nobody has done the work is a market where doing it costs little and is noticed by the people it affects.

4. What are they claiming, and what backs it up?

We list every substantive claim on the site — accreditations, guarantees, coverage areas, response times, named clients — and mark which ones are evidenced on the page and which are asserted. Then we do the same for our client.

This is the part that produces the most useful findings, and it needs no tools at all. A competitor promising a two-hour response with nothing to support it is beatable by anyone willing to publish the actual figure.

5. Where do they ask for the business, and how hard?

How many steps to an enquiry, how many fields, what they demand before they give anything, and whether there is any route for someone who is not ready to talk yet. Sites that only offer “book a call” are leaving the entire early-stage audience unattended.

6. What are they publishing, and does it answer real questions?

Not how often they blog. Whether the pages they have written answer the questions a buyer actually types — pricing, comparisons, whether they cover a particular area, what goes wrong. Most competitor content libraries look substantial and answer almost nothing.

Those six do not all produce the same kind of answer, and the report says which is which. Two of them are measurements you could repeat yourself tomorrow and get the same result. The other four are judgements — defensible ones, made the same way on both sites, but judgements, and an agency that presents all six in the same confident tone is flattering you.

What we will not claim to know

This is the section most competitor audits leave out, and it is the reason ours are shorter.

• Their conversion rate — invisible from outside, at any price
• Their traffic — third-party estimates are modelled, not measured, and can be out by a wide margin
• Their revenue, or how much of it comes from the website at all
• What they spend on acquisition, and whether the site is doing the work or the paid search is
• Whether their apparently better site is actually performing, or simply newer

Any agency putting those on a slide is estimating and presenting the estimate as a finding. We would rather hand over five things we are certain of than fifteen we are not, because the recommendation that follows has to survive contact with a client who knows their market.

Turning it into a decision

An audit that ends in a list of things the competitor does better is not much use — it produces a plan to become a slightly worse version of them. There is a decent argument against that instinct on the evidence, too.

Tuch and colleagues, in the International Journal of Human-Computer Studies in 2012, found across 119 sites that low visual complexity and high prototypicality — looking like what people expect — produced the most positive first impressions, measurable at 17 milliseconds. Copying a competitor’s layout will not hurt you on first impressions. It will simply make you interchangeable at the moment a buyer is comparing three suppliers and looking for a reason to pick one. Match them on the mechanics; separate on the substance.

Done properly, a competitor audit is less about the competitor than it sounds. It is a way of finding out what your own site is failing to say, using someone else’s as the contrast.