Semrush is an SEO suite: it tells you which words a site shows up for on Google, who links to it and what's wrong with its pages. Similarweb is a market intelligence tool: it estimates how many visits a site gets and which channels they come from, whether search, social, direct or advertising. They overlap on one point only, estimating competitors' traffic, and that's exactly where almost every comparison pits them against each other.
For most businesses the useful question is a different one: is the site you want to watch big enough for either of them to have data on it? Similarweb says in its own documentation that it's more confident in its estimates above 100,000 visits a month and shows nothing below 5,000. For a lot of small businesses and their competitors, that settles the choice before you even look at the price.
On 10 October 2026 we searched "semrush vs similarweb" on Google UK. The AI Overview at the top put Similarweb's entry price at "from ~$199/month", while Similarweb's own pricing page says $125. Below it came a Similarweb alternatives roundup, comparisons from BloggingX, Link Publishers and Shane Barker, a Reddit thread and YouTube videos. None of them gives prices in pounds in its search snippet. Here you'll find the comparison with prices converted to sterling, the thresholds below which the data disappears and how we use the two tools on real projects.
Semrush vs Similarweb: the differences in one table
The table puts side by side what each vendor says about itself on its own official pages, not what review sites write. Where a vendor's figures don't match from one page to another, we point it out.
| Item | Semrush | Similarweb |
|---|---|---|
| What it measures first | Google rankings, keywords, backlinks, a site's technical health | Estimated total visits and how they split by channel |
| Where its traffic data comes from | Panel of over 200 million users in more than 190 countries, processed with neural networks | Four sources: analytics of connected sites, a network of devices, public data, partners such as ISPs |
| Declared coverage | 808 million domains, 28.8 billion keywords, 142 geographic databases | 100 million websites and 4 million apps (data page); over 1 billion websites and 8 million apps (comparison page) |
| Backlinks | 43 trillion, crawler covering about 10 billion pages a day | Not the main product |
| Competitor traffic estimates | Only with the Traffic & Market toolkit, from $289 a month | Included from the first plan |
| Entry price | $139 a month (SEO plan) | "from $125/mo"; the rest of the price list isn't public |
| Visibility in AI answers | From the Semrush One plans (Starter, 50 prompts a day) | Gen AI Intel module in the Web Intelligence packages |
| Ownership | Adobe, since 28 April 2026 | Listed on the NYSE (SMWB), offices in New York and Israel |
The coverage row is the oddest. On the page where it describes its data, Similarweb talks about more than 100 million websites and 4 million apps. On the page where it compares itself with Semrush, the websites become more than a billion and the apps 8 million, against the 808 million domains it attributes to Semrush. They're two pages from the same vendor, checked on the same day. If you're choosing, don't use that row as an argument.
Where the numbers come from, and why they don't match
The two tools estimate traffic in different ways, and they say so themselves.
Semrush has two traffic figures that often get mixed up. The first is the one you see in the domain overview: according to Semrush's knowledge base it comes only from Google rankings, multiplying each keyword's search volume by an average click-through rate. The second sits in the Traffic & Market toolkit and is built on clickstream data: according to the page comparing it with Google Analytics, it comes from over 200 million real internet users and more than a hundred apps and browser extensions. The first figure counts organic search only, the second every channel.
Similarweb starts from total traffic instead. On its methodology page it lists four sources: data from sites that have connected their Google Analytics, a network of products installed on millions of devices, public data and partners such as connectivity providers. It says it collects 10 billion signals a day and has 200 data scientists. The organic share it shows is a slice of that total, not a calculation made keyword by keyword.
Put like that, the two figures measure different things even when they have the same name. If Semrush's domain overview gives a much lower number than Similarweb for the same competitor, it doesn't mean one of them is wrong: the first counts only estimated clicks from Google, the second also counts people who type the address or arrive from a newsletter. To see how much each channel weighs in a site's traffic, and why tools' numbers never match your own, see our guide on how to check a website's traffic.
The thresholds below which there's no data
This is the part comparisons deal with in one line, and for small and mid-sized sites it decides almost everything.
Similarweb is explicit. In its support article on data accuracy it says that if the latest snapshot is under 5,000 visits, it won't display any data. In the one on comparing with direct measurement it adds that above 100,000 monthly visits it is "more confident" in its estimates, and that even two direct measurement tools can differ by up to 30% on the same site. Below those thresholds the site's pages show no figures.
We checked it on our own domain. Similarweb's public page for visilay.com, checked on 10 October 2026, shows "No Data Available" for the global, country and category rankings, and no figure under total visits. It isn't a verdict on the site: it confirms that for a niche B2B site the model doesn't have enough signal.
Independent studies point the same way. There are few of them, and none says which countries its sites were in:
- SparkToro (Rand Fishkin, 2022) compared the estimates with Google Analytics data from 641 sites, over twelve months between June 2020 and June 2021. The correlation with real users was 0.790 for Semrush, 0.720 for Datos, 0.659 for Similarweb and 0.504 for Ahrefs, which however only estimates organic traffic. Similarweb got the figure within ±30% about 63% of the time on larger sites, but on sites under 5,000 monthly visitors it was the worst of the group. Semrush tended to overestimate. The sample was gathered through SparkToro's own social channels and newsletter, and the sites' countries aren't stated.
- Shane Barker published a test on 6,000 small and mid-sized business sites taken from Crunchbase (article last updated in April 2024): Semrush returned data for 41% of the sites, Similarweb for 20%. The test doesn't give the raw data or the sites' countries, so read it as an order of magnitude.
Taken together, these numbers say neither tool is reliable on small sites, but Similarweb more often gives no answer at all, while Semrush answers more often, even when it's wrong. The second behaviour isn't necessarily better. A missing number doesn't lead to bad decisions; an inflated one does.
What they cost in pounds, for the same work
The price comparisons you find online usually put Semrush's base plan next to Similarweb's. They aren't the same thing: Semrush's SEO plan doesn't include estimates of competitors' total traffic, which is Similarweb's main product. To get a comparable figure on Semrush you need two subscriptions.
| Configuration | Per month in dollars | Per month in pounds | What you get |
|---|---|---|---|
| Semrush SEO | $139 | £105.10 | Keywords, rankings, backlinks, audit; organic traffic estimated from rankings |
| Semrush Traffic & Market Pro | $289 | £218.51 | Estimated visits across all channels, top pages, market; 1 user, 3,000 requests a day |
| Semrush SEO + Traffic & Market | $428 | £323.61 | The closest equivalent to what Similarweb does, plus all the SEO side |
| Similarweb, declared entry price | from $125 | from £94.51 | Traffic and channel estimates; limits only visible in the checkout process |
| Semrush extra user (SEO plan) | $45 | £34.02 | Each additional person on the SEO plan |
Similarweb's entry price is 29% of the Semrush configuration that most resembles it. But it's the only figure Similarweb publishes openly: the pricing page talks about Web Intelligence plans "from $125/mo" and sends businesses to the sales team. The packages page mentions up to 37 months of history and coverage of more than 100 countries, without saying which limits apply to which plan. Semrush, by contrast, publishes every plan with its limits, and the full breakdown in pounds, with VAT, users and refunds, is in our analysis of Semrush pricing.
For a UK business there's another sum to do. Both vendors price in dollars, so the real cost moves with the exchange rate, and Semrush adds UK VAT at 20% if you buy without a VAT number (Semrush knowledge base). And if the work is only about rankings on Google UK, there are cheaper options: we compare them in our piece on Semrush alternatives, while the comparison with another American competitor is in Semrush vs Moz.
The figures that get passed around and don't add up
Reading the comparisons that rank on Google for this search, and the AI Overview above them, we found three different entry prices for Similarweb and at least three different sizes for its keyword database.
| Source | Similarweb entry price given | Keywords in Similarweb's database |
|---|---|---|
| Traffic Think Tank, March 2026 | $149 a month | 1 billion |
| Exploding Topics, January 2026 | $199 a month | not given |
| CrawlWP, February 2026 | $199 a month or $1,500 a year | 5 billion |
| Spike, July 2026 | $199+ a month | 6 billion |
| Google UK AI Overview, 10 October 2026 | "from ~$199/month" | "approx. 5B-7B" |
| Similarweb, pricing page | from $125 a month | not given |
We have no way of saying which is right, because Similarweb's full price list isn't public, and that's why none of these figures appears anywhere else in this article. The problem isn't only Similarweb's: in the same comparisons Semrush's base plan appears at $139.95, the old Pro price, while the official page says $139. And among the sources the AI Overview cites is a comparison page published by Similarweb itself. Check the price on the official page on the day you buy.
What we use, and for which question
In a digital pre-assessment we prepared for an Italian insurance company, the two tools did different jobs in the same document. Semrush, together with Sistrix and Majestic, covered the SEO picture: ranking keywords, estimated organic traffic, authority and competitors' backlinks. Similarweb was used for one thing only: working out how much of competitors' traffic came from paid search and social, which no SEO tool shows. The competitors were national and international insurers: the case where Similarweb has enough signal to say something.
On an industrial project, like the one in our Macropix case study (an LED screen manufacturer in Italy), the choice is the opposite: competitors are B2B companies with little traffic, Similarweb often has no data, and what you need is to know who holds the top positions for a precise set of keywords. There, the work is done by rank tracking and share of visibility, not traffic estimates.
Hence the table we use to decide.
| Work question | Right tool | Why |
|---|---|---|
| Which keywords does a competitor rank for? | Semrush | It's built on SERPs; Similarweb shows keywords as a share of traffic |
| Who links to competitors' sites? | Semrush | Similarweb has no comparable backlink index |
| How much total traffic does a large competitor get, and from which channels? | Similarweb | It starts from total traffic; on Semrush you need Traffic & Market |
| How much does a competitor invest in advertising compared with organic traffic? | Similarweb, then Semrush for Google ads | Channel mix is Similarweb's strong point |
| How is my own site doing? | Neither | Google Analytics and Search Console measure, the tools estimate |
| How is a niche B2B market moving? | Semrush, with rank tracking | Below 5,000 visits Similarweb shows no data |
| How often does my brand appear in ChatGPT or Gemini answers? | Semrush One or a dedicated tool | Compare it with the specialist tools |
For the last row there are tools built for the job, compared in our roundup of AI visibility tools. For your own site, the real data is in Google Search Console and Google Analytics, which cost nothing.
The test to run during the free trial
Semrush offers a 7-day trial for Traffic & Market too, and Similarweb a free trial on its self-service plans. Before paying, use those days to measure the error on sites whose real numbers you know.
- Pick three sites you have Analytics access to: yours, a client's, a partner's. Ideally of different sizes.
- Take the real visits for the last three months from Google Analytics and the organic clicks from Search Console.
- Note the total visits estimated by Similarweb and by Semrush Traffic & Market, and the organic traffic in Semrush's domain overview.
- Work out the percentage gap between each estimate and the real figure. Similarweb itself considers a difference of up to 30% normal.
- Repeat on the competitors you actually care about and check how many have a figure. If half of them show no data, you have your answer.
It's the same check we run before putting a traffic estimate into an SEO audit: when the gap on known sites is too wide, the report shows tracked rankings instead of the estimate.
How often they're searched in the UK
According to Google Ads volumes for the UK pulled through DataForSEO on 10 October 2026 (average for September 2025 to August 2026), "semrush" gets 27,100 searches a month and "similarweb" 18,100, the same as "ahrefs" and four times "se ranking" (4,400). "semrush vs similarweb" gets 50. Similarweb's figure swings a lot: from 27,100 searches in March 2026 down to 9,900 in August 2026. In the UK, then, Similarweb is a mainstream name, not a niche one. But the searches around it suggest how people use it: "similarweb extension" (390) and "similarweb free" (170) together are searched more than twice as often as "similarweb pricing" (210). Our reading is that a good share of that interest is quick checks with the free pages and the browser extension, rather than a daily working tool.
Frequently asked questions
It depends on the size of the site. In SparkToro's 2022 study Semrush had the highest correlation with real data (0.790 against 0.659), while Similarweb more often got the figure within 30% on larger sites. Below 5,000 monthly visits Similarweb was the least accurate, and it often shows no data at all.
Yes. It's the Traffic & Market toolkit, which estimates visits and channels for any site using clickstream data. It costs $289 a month, about £219 before VAT, and is bought separately from the SEO plans.
Partly. The public pages on similarweb.com show a sample of data for each site, without the detail, and there's a browser extension. For full analysis you need a paid plan, which starts at $125 a month according to the official pricing page.
Almost always Semrush, or a lighter SEO tool. A small business's competitors often have little traffic, below the thresholds at which Similarweb shows reliable data, and the useful work is on keywords and links, which are Semrush's strong point.
Yes, if you study large competitors and need to know how much advertising, social and direct traffic weigh. It's the split we use ourselves on large competitors: Semrush for the SEO work, Similarweb for channel analysis. For a single site or a niche market it's paying twice.
Because they count different things. Semrush's domain overview estimates organic traffic only, starting from Google rankings; Similarweb and Traffic & Market estimate all visits, starting from user panels. Neither sees a site's Analytics data unless the owner connects it.
The number both of them miss
Neither tool sees searches that end without a click. When Google answers with an AI Overview, or when someone asks ChatGPT instead of searching, a competitor can gain visibility without a single extra visit showing up in Similarweb's or Semrush's estimates. The Google UK results page for "semrush vs similarweb" itself opens with an AI Overview that answers the question before any link. As zero-click searches carry more weight, comparing two models of traffic estimation risks becoming an exercise on an ever smaller share of demand. The metric that holds up better is presence: in how many answers, organic or generated, your name appears compared with your competitors'. It's the work we do in our SEO for AI service and in our SEO consultancy, where access to the suites is included and the tools are chosen to fit the question, not the other way round.
