What is CTR?

An ad serves 250,000 impressions and receives 4,500 clicks. CTR is (4,500 ÷ 250,000) × 100 = 1.8%. Doubling that to 3.6% halves the cost per click at an unchanged CPM — which is why creative work shows up in the click price even when the auction never moved.

Click-through rate divides clicks by impressions. It answers one narrow question — of the people the ad reached, what share acted on it — and it answers nothing else. It says nothing about who they were, what they did next, or whether the click was worth its price.

That narrowness is the metric's strength. CTR is the cleanest available read on whether an ad earned attention, uncontaminated by the landing page, the checkout, the offer or the attribution window. When a creative test moves CTR, the creative did it.

The complication is what counts as a click. Meta reports an all-clicks figure that includes reactions, comments, shares and taps on the page name, alongside a link-clicks figure that counts only outbound clicks. Databox's Meta data shows the gap: a 1.18% link click-through rate against 1.90% counting every placement interaction. Both are labelled CTR. Two people comparing Facebook CTRs are frequently comparing different metrics, and the difference is larger than most of the performance gaps they are arguing about.

The published figures cluster by surface, not by advertiser skill.

On Google search, WordStream's tenth-edition study — 13,474 US campaigns between April 2025 and March 2026 — reports an average of 6.64% across all industries. Databox's median across roughly 4,800 contributing companies is 4.71%, and the gap between those two figures is instructive on its own: a mean sits above a median in a right-skewed distribution, which is what advertising performance always is. A handful of very high performers pull the average up past what most accounts experience.

On feeds the numbers fall sharply. Facebook traffic campaigns average 1.71% and leads campaigns 2.59% in WordStream's 2025 set. LinkedIn's median in Databox's cross-platform set is 0.85%. TikTok's is 0.33%. The Google Display Network sits at 0.46% in Store Growers' round-up.

The range from 6.64% to 0.33% is not a range of advertising quality. It is a range of intent. A search ad is served to somebody who typed the problem into a box moments earlier; a feed ad interrupts someone doing something else; a display ad appears beside an article about something unrelated. Holding a TikTok campaign to a search benchmark is a category error, and it is one that gets budgets moved for no reason.

Within a single surface, the benchmark is worth rather more — but even there, an account's own trend line is a better instrument than an industry average, because it holds the industry, the offer and the audience constant.

Click-through rate is one of the two inputs to the click price:

Cost per click = (CPM ÷ 1,000) ÷ CTR

An US$8 CPM at 1% produces an US$0.80 click. The same US$8 CPM at 2% produces a US$0.40 click. The auction did not move; the creative halved the cost.

There is a second, less visible effect. Most ad auctions rank bids by expected value rather than by bid alone — the platform multiplies what you offered by how likely it thinks the impression is to produce the outcome you asked for. Google's quality score is the explicit version of this, and Meta's ranking works on the same principle without publishing a number. An ad the model expects to be clicked wins auctions it did not outbid, and an ad the model has learned to distrust pays a surcharge that appears in the report as a higher CPM.

So click-through rate moves the click price twice: once by arithmetic and once by auction ranking. This is the strongest argument for spending effort on creative rather than on bid tuning.

A click is a cost. Every one of them is paid for, and a click from somebody who was never going to convert is money spent on nothing plus a metric that looks better afterwards.

The reliable ways to buy worthless clicks are well known. A headline implying a discount that the landing page does not offer. A claim about free shipping, or free anything, that turns out to be conditional. Broad match keywords that attract adjacent intent — people searching for a definition when you sell a product, or for a competitor when you sell your own. Creative built around a hook with no relationship to the offer.

Each of these raises CTR and lowers conversion rate together. The net effect on cost per acquisition is usually negative, and the account looks healthier in the one column most people check first.

The test is simple: when CTR moves, look at what conversion rate did over the same window. A rise in both is genuinely better targeting or genuinely better creative. A rise in one and a fall in the other is a promise the page did not keep.

Click-through rate falls with repetition. The same person seeing the same ad for the fourth time is less likely to click it than the first time, and the decline is steep enough to be the leading indicator most accounts have for creative fatigue — it moves before cost per conversion does.

The practical consequence is that CTR should be read against frequency rather than against last month. A campaign whose CTR fell while frequency doubled has a fatigue problem with an obvious fix. A campaign whose CTR fell at constant frequency has a competitive or seasonal problem, and new creative may not touch it.

Small audiences accelerate the decay. A tightly targeted campaign with real budget exhausts its audience quickly and starts repeating, so the fatigue curve arrives in days rather than weeks. Widening the audience buys time as effectively as new creative does, and it is faster.

On search, the largest single lever is relevance to the query. An ad whose headline contains the words the person typed outperforms a better-written ad that does not, consistently and by a wide margin. Tight ad groups exist for this reason and for no other. After that: extensions and assets, which expand the ad's footprint on the results page and give more surface to click; and negative keywords, which raise the rate by removing impressions that were never going to click.

On feeds, the first moment carries most of the decision. The opening clause of the copy, the first frame of the video, the thumbnail — these are judged in the time it takes to scroll past, and testing them beats testing the body. Aspect ratio matters more than most advertisers expect: an ad built for a square feed and served into a vertical placement is competing at a real disadvantage.

Everywhere, test one thing at a time and give each test enough impressions to separate signal from noise. A creative test on two thousand impressions has confidence intervals wide enough to swallow the entire effect being measured, and calling that test is how accounts end up optimising toward randomness.

CTR and CPM are the two halves of the click price, and neither is readable alone. CPM says what a thousand impressions cost; CTR says how many of them earned a click. Cost per click is CPM divided by a thousand, divided by CTR — so an advertiser can pay a premium for impressions and still buy cheaper clicks than a rival on cheap inventory, purely on the strength of the creative.

• Match the ad to the query or the audience rather than to the brand. The largest single lever on search CTR is whether the headline contains the words the person typed.
• Refresh creative before frequency climbs. Click-through rate decays as the same people see the same ad again, and the decay is usually visible before the cost per conversion moves.
• Test the first three words, not the whole ad. Feed placements are judged in the time it takes to scroll past, and the opening clause carries most of the decision.
• Watch what a rising CTR does downstream. A more clickable ad that draws people who bounce has raised one metric and lowered the account's conversion rate; a click is a cost, not a result.