Who your website competitors actually are, and how to check

I picked this subject out of my own data, and the reason is slightly humiliating. Over the last 90 days, fifty different searches asking some version of how do I find my website competitors put my site in front of somebody, 666 times in total. I checked which of my pages Google was showing them rather than guessing: all of it went to one URL, my website comparison tool, at a weighted average position of 65. The biggest single query in that group — "how to find competitors of a website", 83 impressions — averaged position 95. Total clicks across the lot: one.

So the question gets asked, my answer to it sits on page seven, and the tool alone was never the answer anyway: you cannot compare yourself against a competitor until you know which competitor to type in. This is the part that comes first. It costs nothing, it takes about half an hour, and the only tool you need is one you already own.

How do I find my real website competitors?

Take the list of searches your site already appears for out of Search Console, keep the ones that could realistically bring you work, then run each of those searches yourself from where your customers are. Write down who comes back. That is your competitor list — one list per search, not one list per business.

Notice what that method does not start with: the names you already have in your head. Most owners can reel off three or four rivals without thinking — the ones they meet on jobs, the ones whose vehicles they see. Some of those names show up in the search results as well. Some do not, and there is nothing mysterious about why: a company only appears on a search page if somebody has done the work of putting it there, and that work is separate from being good at the job.

Why "who is my competitor" is the wrong question

Because Google does not run one competition for your industry. It runs a separate one for every phrase somebody types, and the results are different for each. Two phrases that mean the same thing to a human are two different searches with two different sets of winners.

Here is what that looks like in real numbers rather than in theory. Mississauga is the busiest city in my own search data, and over those 90 days people typed it 99 different ways — 5,804 impressions across 99 distinct phrasings. Four of them, all asking for exactly the same thing:

I checked which URL Google served for each of those four, because "the same page" is the sort of thing that is easy to assume and easy to get wrong. It was the same page all four times: my Mississauga web design page, and nothing else. One page, one city, one service, four ways of asking — and seventeen places between the best result and the worst. Down at that depth it only moves me between page five and page six, but the mechanism is the point: the phrasing changed and nothing else did. Whoever sits above me on the first phrase is not necessarily the company sitting above me on the fourth.

That is why "who is my competitor" has no single answer, and why a report that hands you five competitor domains for your whole business is answering an easier question than the one you asked. The useful version is narrower: who is beating me on this one search, the one I actually want.

Step 1: take the list you already own

Search Console is free, it belongs to you, and it reports the searches your site was genuinely shown for. That last part matters more than it sounds. Everything else on the market is modelled from samples; this is a record. If your site is not verified there yet, do that first — of all the free things a business owner can set up in an afternoon, this is the one I would pick, and I have written the plain-English guide to reading Search Console for exactly this reason.

1. Open Performance → Search results and set the date range to the last three months. A shorter window gives you noise.
2. Click the Queries tab under the chart. This is the list. Turn on the Average position and Impressions checkboxes above the graph so both columns appear.
3. Sort by impressions, not by clicks. On a site that is not ranking yet, clicks are almost all zero and tell you nothing. Impressions tell you what Google already thinks you are about.
4. Export it (the button top right) so you can sort it in a spreadsheet without fighting the interface.

One caveat worth knowing before you go hunting for your own name in the results: Google states plainly that "Even if a query appears in your list, you might not see your site in results if you run the same query in Google Search." The position column is an average across days, devices and locations — not a promise about what is on your screen right now.

Step 2: throw most of the list away

A three-month export looks impressive and is mostly noise. Mine came back with 1,314 distinct queries and 20,338 impressions. Of those queries, 561 — nearly 43% — earned two impressions or fewer. They are real searches that a real person typed once, and there is no page you could write that would turn two impressions into work.

The concentration is what you should be looking at. Here is how the demand stacks up on my own site:

Now filter what is left by intent, not by volume. A query is worth checking if somebody typing it could plausibly hand you money this month. "Web design mississauga" qualifies. "How to make a comparison website" — a real query in my list — does not, because that person wants to build one, not hire one. Ten to fifteen queries is a workable shortlist. If you end up with forty, you have not finished cutting.

While you are in there, look at where the impressions actually sit. On my site, 10,619 impressions came from queries averaging between position 20 and 50 — visible to Google, invisible to humans. That band is where the useful work is, because those are searches you are already relevant for and simply not winning. If that describes your site too, the diagnosis in why your website is getting no traffic is the companion piece to this one.

Step 3: run the search the way your customer runs it

Now open a browser and search your shortlist, one query at a time, writing down the first ten results for each. The trap here is that you are not a neutral observer of your own market, and the biggest reason is location.

Google's help page is blunt about it: "Your location helps to show you nearby results, even if you didn't include a location in your search." It also lists where that location comes from — device GPS, your saved home and work addresses, previous activity, and your IP address, which is "roughly based on geography." So four separate signals are quietly editing your results before you read a single one.

What that means in practice:

• Use a private window, and sign out of your Google account. That takes your account history out of the equation. It does not make you anonymous to Google's location estimate — your IP address is still there, and Google lists it as one of the four sources it uses.
• Check from where your customers are, not where your desk is. If you serve a city you do not live in, your desk results are not the results your customers get. Being physically in that area, even once, is worth more than any browser trick.
• Do not chase a perfect simulation. You cannot fully reproduce a stranger's result page and you do not need to. You are collecting names, not measuring a rank to one decimal place.
• Search on your phone too. The layout is not the same one you get on a desktop: the map block takes far more of the screen, and the first ordinary result can sit below the fold entirely.

Write the names down in a plain table: one row per query, the domains that appeared, in order. After ten queries you will start seeing the same two or three names repeatedly. Those are your actual competitors online. Any name that appears on one query and nowhere else is a competitor for that one search, which is useful to know and quite different from being a rival to your business.

Step 4: the map pack is a separate competition

On any local search there are two competitions on one screen, and they have different winners. The block of three businesses with a map is the local pack, and it is ranked by different rules than the blue links underneath it. If you write both into one column you end up with a list you cannot act on, because the fix for one has almost nothing in common with the fix for the other.

Google names the three factors: relevance ("how well a Business Profile matches what someone is searching for"), distance ("how far each business is from the customer who's searching") and prominence ("how well-known a business is"). And, in case anyone tries to sell you otherwise, the same page says: "There's no way to request or pay for a better local ranking on Google."

So keep a second column in your table for the three businesses in the map block. Distance is doing real work there, which is why a competitor two streets from your customer can outrank a better-known company across town — and why that column changes as you move. If the map pack is where your customers are looking, that is a different job from ranking a web page, and it is the one covered in the guide to the Google 3-pack and handled as ongoing work in Google Maps management.

What to compare, once you have the names

You now have a short list of real competitors per search. The temptation is to open their site and admire the photography for twenty minutes. Compare these four things instead, in this order, because each one is something you can act on this week.

1. Does their page answer the search? Click the result and read the page that Google actually chose. If somebody searched for a service in a city and landed on a dedicated page about that service in that city, while your equivalent is one line on a homepage, you have found your answer in ninety seconds.
2. Read the title Google shows. Not their logo — the blue line in the results. It tells you what that page is claiming to be about. Compare it with yours on the same search and see which one reads like it was written for the person typing.
3. Count their reviews and photos. On the map results, the review count and the date of the newest photos are visible without clicking anything. Of everything on this list it is the cheapest gap to close — asking customers for reviews costs nothing but nerve — and it takes months rather than days to show up.
4. Open it on your own phone. Time how long it takes to become readable, then try to find the phone number. A site that is beautiful on a desktop and hostile on a phone is beatable, and you will know within ten seconds.

For the mechanical half of that — speed, mobile handling, page titles, whether basics are missing — you can put both addresses into my free website comparison tool and have it checked side by side rather than doing it by eye. It will not tell you who your competitors are. That was this whole article. It will tell you where you stand against the ones you found.

What you cannot find out about a competitor, at any price

You cannot see another company's traffic, their conversion rate, their ad spend or their revenue. What you can buy is a model of the first one, built from sampled data and a click-through-rate curve. The tools themselves say so in their documentation, which is worth reading before you quote one of their numbers at a meeting.

Ahrefs, one of the better tools in this market, says it on its own help page: "Ahrefs only estimate organic and paid traffic," and the numbers are "not collected directly, and some discrepancy is normal." It goes further — "There's no such thing as 'accurate' search volume or traffic." I quote that not as a criticism; it is a straight answer from a company that could easily have avoided giving one. It means the number is good for comparing two sites against each other and poor as a figure to plan a budget against, because a plan built on an estimate inherits its error bars.

The same caution applies to the scored audit report that lands in your inbox after somebody scans your site. A number out of 100 compresses dozens of unrelated checks into one figure, so it tells you almost nothing about which one to fix. If you want to know whether the money you are spending is working, the checks are in is my SEO working, and every one of them reads your own data rather than a guess about somebody else's.

What to do with the list on Monday morning

Do one thing with the table this week, then put it away. Pick one query — the one where the money is closest and you are not hopelessly far behind, usually something already sitting in that position 20-to-50 band — and give it one page that is unambiguously about that search. Not a mention on a homepage. A page whose title, first paragraph and content all say the same thing the searcher typed.

Then leave it alone for a couple of months and re-run the same search. Positions move slowly and there is no version of this where you check daily and learn anything. When you repeat the exercise, keep the old table: the interesting signal is not where you sit today, it is which names have disappeared and which new ones arrived. That is how you find out whether the market is getting harder or whether somebody just published something better than you.

And if the honest answer after all this is that your page cannot win because the site around it is slow, thin or ten years old, then no amount of competitor research fixes that. Rebuilding the thing is a different conversation, and one I am happy to have — see what local SEO actually involves first, because it is usually cheaper than a rebuild.

Frequently asked questions

Sources, all read in full on 3 September 2026: Google — Location and Google Search results · Google — Improve your local ranking on Google · Google Search Console — Performance report · Ahrefs — checking organic traffic of sites you do not own. Search figures are from my own Search Console property for leadlls.com, 3 June to 31 August 2026.