Quick answer: the most common way to show a result that doesn't exist is to take credit for people who already knew you. Traffic from queries containing your company name arrives regardless of any SEO work, but in a report it gets added to everything else and produces a satisfying upward curve. The second method is showing impressions instead of clicks. The third is reporting the number of links bought rather than what they achieved. All three can be checked in Search Console in twenty minutes, and this article explains how.
A thirty-page report full of rising charts, and the same number of enquiries as before. The situation is so common it has become a genre, and it's usually explained as the agency being dishonest. Sometimes it is. More often something else is going on: the metrics being reported genuinely did grow — they just have nothing to do with what you're paying for.
Here are the three substitutions that appear most often, and how to check each one yourself.
Branded traffic in an SEO report: taking credit for someone else's result
The main method, and the least visible.
Queries split into branded and non-branded. Branded ones contain your company name: the person already knows you, saw a card, heard from a friend, remembered a sign — and types the name to find your site. Non-branded ones describe a need: "dentist near me", "auto repair website cost", "employment lawyer Warsaw".
The distinction matters. Branded traffic arrives because of your advertising, your signage, word of mouth and years of work. It would arrive without any agency. Non-branded traffic is people who didn't know you existed — and that's what SEO is paid for.
In a report they get added together. You opened a second location, ran some outdoor advertising, got mentioned in a local group — branded queries rose, the overall chart went up, and in the report it reads as the agency's work. No data was falsified: the numbers are real, the cause is simply different.
This is especially visible in seasonal businesses and after any offline activity. A spike in May for a home renovation company is probably May, not optimisation.
How to check. Open Search Console, the performance report, add a query filter and exclude your company name in every spelling, including misspellings. What remains is your non-branded traffic. Compare its trend over recent months against the overall one. If the overall chart is rising while the non-branded line is flat, the SEO work isn't producing results.
Impressions instead of clicks: why rising visibility brings no traffic
The second method is simpler, and it appears even with honest contractors — out of a desire to show something.
An impression means your page appeared in the results for a query. Not that anyone saw it, and certainly not that anyone visited. A page on the eighth page of results accumulates impressions for months and receives no clicks at all.
Here are concrete figures so this doesn't stay abstract. A young site whose data I have access to accumulated 33,617 impressions and 46 clicks over the last twelve months. That's 0.14 percent. The cause isn't the titles or the copy: almost everything sits at position forty-five or below, where clicks essentially don't happen. Impressions rose steadily all year — and from that data it would be easy to assemble a report showing "visibility up fourfold", using real numbers and without falsifying anything.
The same family includes "number of keywords in the top 100" and "average position across all queries". The first means almost nothing: position one hundred and position one count equally. The second averages branded queries, where you rank first, with commercial ones, where you rank sixtieth, and produces a respectable-looking figure with nothing behind it.
How to check. In the same performance report, look at clicks rather than impressions, and at positions for specific commercial queries rather than the average. If your agency shows impressions and average position, ask how clicks on non-branded queries changed.
Bought links: why the number of placements isn't a result
The third substitution: a report in which the deliverable is the number of links placed.
Links do influence rankings, and external mentions are a genuine factor. The problem is twofold.
First: a link count is a report of effort, not of outcome. "We placed 40 links this month" is the same as "we painted 40 square metres" — you know what was done, not what came of it. The outcome is positions on target queries and non-branded traffic.
Second, and more important: links placed in bulk usually sit on sites created for the purpose of hosting links. Nobody reads them, and search engines learned to recognise them long ago. Mass link buying achieves nothing at best, and at worst becomes a problem in itself: link schemes carry penalties, and unwinding one takes longer and costs more than doing the work honestly would have.
One separate signal: if an agency won't show you where exactly it places links, that isn't commercial confidentiality. That's the answer.
How to check. Ask for the list of sites. Open five at random. The questions are simple: does this look like a site anyone reads? Is there anything on it other than articles with links? Would a real person arrive from that page? If all five look identical and none relates to your industry, you're paying for a report rather than a result.
Five more ways to show growth that isn't there
A short list of things that appear less often, but do appear.
Reporting on selected keywords. You're shown ten phrases where positions improved and not the other two hundred. Often these are long-tail terms you were already close on.
A convenient comparison period. Comparing against a weak month rather than the same period last year. A seasonal business can be shown as growing indefinitely this way.
An in-house reporting dashboard. The agency shows data from its own panel rather than from Search Console and Analytics. It can edit those figures, and you have no way to verify them. The normal arrangement is access to your own Google accounts and reports drawn from them.
All traffic counted as an SEO result. The report includes visits from ads, social media and direct entries. Technically that's traffic; it has nothing to do with SEO.
Irrelevant and bot traffic. A sharp rise in visits from countries you don't serve, with no change in the number of enquiries.
What agencies show in SEO reports and what to ask in return
- Growth in total organic traffic — May be entirely branded — How did non-branded traffic change?
- Growth in impressions — Visibility without clicks, possibly at position 60 — How did clicks and commercial-query positions change?
- Keywords in the top 100 — Almost nothing — How many are in the top 10, and which ones?
- Average position across all queries — Averages branded with commercial — Positions for ten target queries, separately
- Number of links placed — A record of effort, not outcome — The list of sites, and what changed in positions
- Data from an in-house dashboard — Unverifiable — A report from my own Search Console and Analytics
- Overall traffic growth — May include ads and direct visits — A breakdown by source
What an honest SEO report looks like
So that the list above doesn't read as "everyone is cheating", here's what a normal picture looks like.
Clicks on non-branded queries — the headline number. It answers whether people who didn't know you are arriving.
Positions for a specific list of target queries, agreed in advance. Not an average, not the top 100 — the actual phrases you're paying for.
Enquiries and calls, where they can be attributed to a source. It's the only metric a business genuinely cares about, and a good contractor works toward it even when attribution is difficult.
What was done in the period. Specifically: which pages were rewritten, which technical fixes went in, what was published. SEO work is routine and repetitive, and there's no reason not to show it.
What didn't work. The fastest way to recognise an honest contractor. In SEO some hypotheses fail, and a contractor whose every month goes according to plan either isn't testing hypotheses or isn't telling you the results.
What to ask an SEO contractor before starting
Four questions that save months.
How do you separate branded from non-branded traffic in reports? An answer along the lines of "why would we" or "we look at the total" closes the subject.
Which specific queries will we be growing on? The list should be agreed before work starts; otherwise the report will contain whichever ones happened to improve.
Do you buy links, and if so, where? The answer matters more than the fact itself: a contractor willing to show the sites and explain the reasoning is a different proposition from one who deflects.
What will you show me after three months if nothing has grown? A test of honesty. That does happen, particularly on a young domain, and the question is whether the contractor admits it or finds a metric that went up.
How I report on my work and why I don't buy links
I combine development and SEO, and I report from your Google accounts rather than my own dashboard. The report contains: clicks on non-branded queries separated from branded ones, positions for an agreed list of phrases, what was done in the period, and which hypotheses didn't work out.
I don't buy links. External mentions matter, but they're earned through publishing, profiles in relevant directories, and work somebody is willing to write about — not through packages of placements on sites built to host links.
And the example above — 33,617 impressions against 46 clicks — is there precisely because that kind of data is the easiest to build a flattering report from. That's the test: showing the inconvenient number next to the convenient one rather than choosing between them.
Where to start checking an SEO report in twenty minutes
Reduced to one principle: look at clicks on non-branded queries. Everything else is derivative, and nearly every substitution works by showing you something else instead.
A practical step for today: open Search Console, the performance report for the last six months, add a query filter and exclude your company name. Look at what's left of the chart. Twenty minutes is enough to understand what you're paying for.
If you'd like to establish what's actually happening with your site, write to me and we'll go through the data together.
