What Is a Good CTR? Start With the Rate Your CPM Requires

A media buyer opens Meta on Monday, sees a 1.1 percent click-through rate on the prospecting campaign and starts writing new hooks. On the next tab, a Google Display line is running at 0.28 percent and nobody has mentioned it in months. The display number is nearly four times lower and it is the healthier of the two, because those two rates are being asked to clear completely different prices.

CTR on its own grades nothing. It is one term in the arithmetic that turns a price for impressions into a price for customers, and the only rate that counts as good is the one high enough to carry the CPM you are actually paying. The published ranges are below, because the question deserves them. The division that answers it comes after.

What the reported ranges actually cover

Click-through rate is clicks divided by impressions, and the spread across formats is wider than in almost any other advertising metric. A single percentage point separates a strong display campaign from a mediocre search campaign in absolute terms, but they sit two orders of magnitude apart in what they mean.

Read that as a map of neighbourhoods rather than a scoreboard. Every row is an aggregate across thousands of accounts, industries and seasons, and the variation inside one row usually beats the gap between two rows. There is also a definitional trap sitting in the Meta lines: the platform reports both a CTR for all clicks and a CTR for outbound link clicks, and the first is routinely two to three times the second because it counts profile taps, image expands and comment opens. Accounts get declared healthy on the wrong one every week.

CTR is a divisor, not a grade

The reason a 0.28 percent display rate can be fine and a 1.10 percent Meta rate can be a problem is that CTR never appears alone in any decision. It sits between the price you pay for impressions and the price you pay for clicks.

Cost per click = CPM / (CTR x 1000)

Break-even CTR = CPM / (contribution per order x conversion rate x 1000)

Take the store this site keeps returning to. Average order value $86, contribution margin 42 percent once goods, shipping, packaging, payment processing and a returns provision are removed, so $36.12 of real money per order. It runs Meta prospecting at a $14 CPM with a 1.1 percent CTR and a 2.5 percent site conversion rate.

A thousand impressions costs $14 and produces 11 clicks, so each click costs $1.27. At a 2.5 percent conversion rate it takes 40 clicks to make an order, which puts cost per acquisition at $50.91 against $36.12 of contribution. Every order that campaign delivers destroys $14.79. Nothing in the CTR column signals that, which is the whole problem with reading it as a grade.

The CTR you need is a division

Run the chain backwards. At a $14 CPM, $36.12 of contribution and a 2.5 percent conversion rate, the campaign needs 14 divided by 903, or a 1.55 percent click-through rate, to come out level. It is running 1.10 percent. The gap is not a rounding problem: the rate has to rise by 41 percent before the campaign stops losing money, and that is at today's CPM, which nobody has promised to hold.

The shape of that chart explains most arguments about whether a CTR is good. A retargeting audience at a $40 CPM needs 4.43 percent to break even on the same funnel, which is above the top of the reported retargeting range. Warm-audience CTR looks impressive precisely because the impressions cost four times as much, and the two facts cancel. The display line from the opening is the same story inverted. It buys impressions at $3, where break-even sits at 0.33 percent, so at 0.28 percent it is 16 percent short of the rate it needs. The Meta campaign at 1.10 percent against a target of 1.55 is 29 percent short. The line that looked four times worse has half the problem.

Find your CPM down the left, your conversion rate across the top, and the cell is the rate you have to hit. Anything above it is good in the only sense that pays. The CTR calculator does this both ways round, and the CPM calculator covers the price side.

The trade that every CTR fix makes

Here is where chasing the number stops being harmless. Cost per acquisition depends on CTR and conversion rate as a product, not as two independent scores, so anything that lifts one while dragging the other has to be judged on the multiplication.

The baseline campaign multiplies 1.10 by 2.50 and gets 2.75. New creative with a sharper hook takes CTR to 1.54 percent, a 40 percent lift, and any dashboard would call that a win. If the sharper hook also pulls in browsers who were never going to buy and conversion slips to 1.75 percent, cost per acquisition lands at $51.95, slightly worse than the $50.91 it started at. The lift was real and it bought nothing.

The tolerance is easy to hold in your head. A 40 percent gain in CTR pays for itself only while conversion falls by less than 28.6 percent, because one divided by 1.4 is 0.714. Double the CTR and you can afford to halve conversion, no more. Any creative test that reports one of those two numbers without the other is reporting half a result, which is why the conversion rate calculator belongs open in the next tab, and why what counts as a good conversion rate has the same answer shape as this page.

The third bar is what a genuine improvement looks like: better targeting or a clearer offer lifts CTR to the same 1.54 percent while conversion holds, and cost per acquisition drops to $36.36. That is a rounding error away from the $36.12 the order actually earns, so even the good version of this fix only reaches the line. A campaign 41 percent short on CTR is rarely rescued by creative alone.

Why your CTR moved when nothing changed

Most week-on-week CTR movement is not creative fatigue, whatever the account manager says. Placement mix is the usual culprit: an algorithm shifting spend from feed to Audience Network or Reels will halve a blended CTR while every individual placement holds steady, and the fix is a segmented report rather than a new ad. Frequency is the second, and it cuts in both directions, since early repetition often lifts CTR through recognition before fatigue pulls it down past four or five exposures.

Audience width is the third and the least discussed. Narrowing an audience raises CTR almost mechanically and raises CPM at the same time, which is the trade the first chart prices. Broadening does the reverse. Judging a targeting change on CTR alone will always favour narrowing, and narrowing is exactly the move that pushes you into the expensive right-hand bars. On social placements the same logic applies to engagement rate, which moves with CTR often enough that people treat them as one signal and then optimise against a metric with no price attached to it at all.

What to do with the number this week

Pull three figures for one campaign: the CPM you paid last month, the conversion rate of the traffic it sent, and the contribution per order from your own cost sheet rather than the gross margin from the accounts. Divide the CPM by the contribution times the conversion rate times a thousand. That single number is your CTR target, and it is worth more than any benchmark table because it belongs to your prices.

Then compare it against what the campaign delivered. If the delivered rate clears the target, the CTR is good and no further work is needed there. If it does not, decide before touching a single ad whether the shortfall is small enough for creative to close, because a gap over about 30 percent usually means the CPM is wrong for the audience or the contribution is too thin to buy that audience at all. The contribution margin calculator settles the second question, and what a thousand impressions should cost you settles the first.

Related reading: the CTR formula and how CPM becomes CPC.