What counts as a good click-through rate depends on who's asking.

Someone runs a Google ad for their dental practice, gets a 6% click-through rate, and feels good about it until a friend who sells real estate mentions hers is sitting at 9%. Same platform, same kind of campaign, same effort put into the ad copy. The dentist assumes the real estate agent just writes better ads. That's the wrong lesson, and it's worth being precise about why.

CTR is the easiest number in advertising to check and one of the easiest to misread, because the question "is this good?" only has an answer once you know which industry, which year, and which spot on the page you're talking about.

The average moves by industry, and by a lot.

Start with industry. WordStream, which tracks ad performance data across thousands of advertiser accounts, found in its 2024 benchmark report that the industries with the best average click-through rates that year were Arts and Entertainment at 13.04%, Sports and Recreation at 9.66%, and Real Estate at 9.20%. On the other end, attorneys, home improvement contractors, and dentists were all clustered down near 5 to 5.6%. That's not a small gap. It's the difference between a click rate that looks embarrassing in one field and looks completely normal in another.

So the dentist and the real estate agent from the opening example aren't actually comparable. A 6% CTR is a solid result for a dental practice. A 9% CTR is roughly average for someone selling houses. Neither number tells you which ad was better written; it mostly tells you what kind of thing people click on more readily, and how much competition exists for the same search terms. What you actually need is a way to judge the writing itself, separate from the market it's competing in, and it's the problem we're building Knowing to solve.

The average also moves by year, sometimes violently.

Here's where it gets less intuitive. Even within the same industry, the benchmark isn't stable from one year to the next. WordStream's numbers show Finance and Insurance click-through rates dropping 32.4% from 2023 to 2024, and dentists falling nearly 20% in that same one-year window, while Apparel, Fashion and Jewelry rose 112% and Career and Employment rose 81% over the same period.

Nobody in apparel advertising suddenly got twice as good at writing headlines in twelve months. What moved was almost certainly the mix of who's searching, how much competition bid up or down, and how the platform's algorithm distributed impressions that year. A related report makes the same point from a different angle: the overall average CTR across all industries climbed from 1.35% in 2015, when this tracking began, to 6.64% by 2024. That's not because ads got five times better over those nine years. Search behavior changed, ad formats changed, and the platforms themselves changed what an "impression" even counts as.

Where your ad sits on the page changes the number more than the ad does.

This is the part most benchmark charts leave out entirely: position. Whether your ad shows up first, third, or on the second page matters enormously, independent of whether the ad itself is any good.

A study by researchers Ye Chen and Tak W. Yan on sponsored search found that click-through data "has been confounded by positional bias, that is, users tend to click more on ads shown in higher positions than lower ones, regardless of the ad relevance." In plain terms: people click the top thing because it's on top, not necessarily because it's the best thing. The same pattern shows up outside of paid ads, too. Neil Patel's data on organic search results (the non-ad listings) makes it concrete: the top three organic positions get click-through rates of 39.8%, 18.7%, and 10.2% respectively, a drop of roughly two-thirds between the first and third spot. Similar positional effects appear in paid search.

That single fact quietly undoes a lot of DIY ad testing. If you run two versions of an ad and one clicks better, you don't automatically know whether the copy was actually more compelling, or whether it simply got shown in a higher slot more often. A small gap in your dashboard can be noise, not signal. A broader survey of research on sponsored search, by Xinfei Wang, catalogs how much of the click-through modeling literature exists specifically to try to strip position out of the picture and isolate what the ad itself contributed. If separating those two things takes researchers a whole body of statistical modeling, a business owner eyeballing two numbers should be cautious about drawing hard conclusions from a small gap.

A high number can still be a bad result.

Even a well-adjusted, industry-correct, position-corrected click-through rate answers a narrower question than owners think it does. It tells you people clicked. It says nothing about what happened next.

The benchmark report itself notes this directly: "high CTR doesn't always mean high performance," and recommends reading it alongside conversion rate and cost per click, not alone. Attorneys (grouped under legal services in the same benchmark data) sit at the bottom of the CTR rankings but pay the highest average cost per click of any industry tracked, at $9.87 per click. That's not a contradiction. Legal cases are high-value, so even a low click rate on expensive clicks can still turn a profit if just a few of those clicks become clients. It's a reminder that a low click rate in a competitive, high-value field can still be an efficient result, and a high click rate in a cheap, low-stakes field can still be a wasteful one.

What to actually compare your number against.

Look at the industry average once, to see what's normal in your field before you launch. Then set it aside. The more useful comparison is your own history: your own account, your own past ads, the same platform, roughly the same time of year, roughly the same position in the auction. That comparison holds most of the moving parts steady, which an industry-wide average never can.

If this month's click-through rate beats last month's, on the same kind of ad, aimed at the same kind of person, that's a real signal. If it's four points below the number a stranger in a different industry posted on a different platform in a different year, that's not evidence of anything except that you looked at the wrong comparison.