Discussion starter: an SEO measurement question, posted with an answer from Niraj Raut to open the thread. If you have dealt with this on a site, reply with what you saw, especially where it differs.
- SEO Analytics
When third-party clickstream estimates like Similarweb disagree with a competitor’s visible ranking trends, which do you trust for competitive analysis?
Disclosure: Niraj Raut, who posted this answer, runs the SEO consultancy linked at the end of it.
1 reply
-
adminAdmin
I trust neither on its own, because they measure different things. Clickstream tools such as Similarweb estimate total visits to a site from panel and partner data, modelled up to the whole population. Ranking and visibility tools estimate organic presence from a tracked keyword set multiplied by assumed click-through rates. When they disagree, it is often because both are right about what they measure: a competitor can gain rankings while losing visits, or gain visits from channels that have nothing to do with search. The practical rule is to pick the tool that measures the question you are asking, trust direction more than magnitude, and triangulate with a third source before acting.
Why they disagree
Reason Effect on clickstream Effect on ranking tools Brand demand rises from advertising or PR Visits rise, including direct and branded search Little change on non-brand keywords AI Overviews or other features reduce clicks on stable rankings Organic visits fall Rankings unchanged, visibility estimates may stay flat Competitor grows in queries your keyword set does not track Visits rise No change on your tracked set Paid, social, email or app growth Visits rise No change Ranking tool methodology changes, such as the loss of the &num=100parameter in September 2025No change Step change in tracked positions or visibility Panel bias: device, geography, demographic Distorted, especially for small or niche sites Not affected Match the tool to the question
- “Is the competitor winning organic visibility on the queries we compete for?” Use ranking and visibility data on a keyword set you define, ideally the same set you track for yourself. Clickstream cannot answer this reliably.
- “Is the competitor’s overall business growing?” Clickstream gives a directional view of total traffic, but it should be checked against other signals.
- “Where does their traffic come from?” Clickstream channel splits are useful for large sites and unreliable for small ones.
- “Did they get hit by an update?” Visibility tools show this faster and more clearly, provided the drop is not a tool methodology change.
How I triangulate
- Brand demand. Google Trends for the competitor’s brand name against yours shows whether interest in the brand is changing, independent of rankings.
- Your own Search Console data. For queries where you both appear, your impressions tell you whether demand is changing. If your impressions for shared queries are stable and your position moved down, the competitor probably gained. If your impressions fell too, demand or SERP layout changed.
- Search results themselves. For a sample of important queries, look at what ranks and what features appear. A tool can report a position without telling you that an AI Overview now sits above it.
- Business signals. Hiring patterns, product launches, pricing changes, funding announcements and app store rankings can explain traffic movement that search data cannot.
- Vendor change logs. Check whether either tool announced a methodology change around the date of the divergence.
Watch for this:
Clickstream estimates for small and mid-size sites can be wrong by a wide margin, and the error is not consistent across competitors. Comparing two small competitors’ estimated visits is comparing two noisy models. Treat those numbers as ranges, and do not build a forecast or a budget case on them.
A worked interpretation
Hypothetical example:
A competitor’s visibility score on your tracked keyword set has risen steadily for six months, while Similarweb shows their visits flat. Google Trends shows their brand interest flat too. Your own Search Console data shows your impressions on shared queries stable and your average position slightly down on the queries where they gained. The most likely reading: they are winning rankings on your shared queries, but those queries now lose clicks to AI features and shopping units, so the rankings are not producing more visits. The response is not to panic about their traffic, but to defend the rankings on the queries that still produce clicks and conversions for you.
A competitive monitoring setup that avoids the argument
Most disagreements between tools happen because the question was never pinned down. A setup that avoids them has three layers. First, a shared keyword set chosen by business value, not by what the tool suggests, grouped into the product lines or topics you report on. Track your own and competitors’ positions on that set, and weight them by estimated click-through at each position so a gain at position two counts for more than a gain at position fifteen. That weighted share of voice is the organic competitive metric. Second, a quarterly view of clickstream for context on total traffic and channel mix, read as direction only and only for competitors large enough for the panel to be meaningful. Third, a short monthly review of the actual search results for your twenty most valuable queries, noting new features, new competitors and page types that are gaining. When a stakeholder brings a Similarweb chart that contradicts the visibility trend, you already have the three sources needed to explain it.
Signs that one source is simply wrong
- A step change on a single day across many sites in the same tool usually means a methodology change, not a market event.
- Clickstream showing large traffic swings for a small site with no matching change in rankings, brand interest or business news is likely panel noise.
- Visibility gains on keywords that do not matter to the business can look like a threat when they are irrelevant. Check which keywords drive the change.
- If you can compare either tool against your own analytics for your own site, do it. The size and direction of its error on your site is the best available guide to how much to trust it on competitors.
Rankings tools for questions about search, clickstream for directional questions about the whole business, your own data and search results to arbitrate, and never a decision based on a single third-party estimate. When they disagree, the disagreement is usually the insight: it tells you something about the gap between visibility and visits that matters for your own strategy too.
Need help with this on your own site?
Niraj Raut works with businesses in Nepal, Australia, the UK, Europe and the US on technical SEO, ecommerce SEO, local SEO and AI search.
Add a reply Cancel reply
Replies are open to approved members. Log in or apply to join.