Tecof • September 9, 2026
CTR, CPC, CPM, CPA: Digital Ad Metrics Guide
In Brief
CTR is the share of people who clicked after seeing your ad: clicks divided by impressions, times 100. CPC is the price of one click: spend divided by clicks. CPM is the price of a thousand impressions: spend divided by impressions, times 1000. CPA is what one sale or signup costs you: spend divided by conversions. All four measure different points on the same funnel, and reading any one of them alone will mislead you. Only one threshold decides whether a campaign makes money: whether your CPA sits below your net profit per order.
It is Monday morning and your Google Ads dashboard is open. One campaign row shows a 2.4 percent CTR, an average CPC of 3.50 TL, a CPM of 45 TL and a cost per conversion of 85 TL. Every number looks reasonable at first glance. The bank balance at the end of the month still says otherwise.
The trouble is usually not the numbers themselves but the order you read them in. CTR judges your creative. CPC judges competition and your quality score. CPM tells you how crowded and how contested your chosen audience is. CPA measures everything that happens after the click: the product page, the shipping fee, the checkout step, even your returns policy. A single metric glowing on the dashboard means nothing if the three behind it are broken.
The math below runs on one account: a mid-sized online store with a 30,000 TL monthly ad budget. Swap in your own margin and the logic holds.
CTR: Is Anyone Interested in Your Ad?
CTR is clicks per impression. A thousand people saw your ad, twenty-five clicked, so your CTR is 2.5 percent. The only thing it measures is whether the headline and the image landed with the audience you picked. It says nothing about sales, and anyone who claims otherwise is reading it wrong.
The formula and four quick calculations
CTR = (clicks / impressions) x 100
| Channel | Impressions | Clicks | CTR |
|---|---|---|---|
| Google Search | 1,000 | 25 | 2.5 percent |
| Instagram feed | 5,000 | 200 | 4 percent |
| Google Display | 10,000 | 50 | 0.5 percent |
| Email campaign | 1,000 sent | 150 | 15 percent |
That same 0.5 percent is perfectly normal on display and a disaster on search. Talking about CTR without naming the channel is pointless.
What counts as a normal range
- Google Search: above 3 percent is strong, 1.5 to 3 percent is average, below 1 percent needs rewriting
- Google Display Network: around 0.5 percent is fine, below 0.1 percent means the creative is dead
- Meta feed ads: 1 to 2 percent is healthy, swap the image below 0.3 percent
- Instagram Stories: expect it to sit under feed performance, 0.5 percent is enough
Industry averages put ecommerce and education near the top of search CTR, while finance and insurance run lower because competition crowds the auction. Your own history beats any published table. Take the last 90 days as your baseline and measure every new creative against it.
Why a high CTR can be a trap
An apparel brand ran the headline "Everything 99 TL". CTR jumped to 6 percent. Visitors landed, found only three products at that price, and left. Conversion rate stalled at 0.3 percent and CPA climbed to 400 TL. Record clicks, record waste. When the promise in the ad and the promise on the landing page do not match, a high CTR just buys you a more expensive disappointment. Comparing variations through a properly structured split test before scaling saves that money.
CPC: What a Click Really Costs
CPC = total spend / clicks. You spent 500 TL, you got 100 clicks, each click cost 5 TL. Simple enough. The hard part is deciding whether 5 TL is cheap or expensive.
Typical ranges by sector
| Sector | Google Ads CPC | Meta Ads CPC |
|---|---|---|
| Ecommerce | 2 - 8 TL | 1.5 - 5 TL |
| Fashion and accessories | 2 - 6 TL | 1.5 - 6 TL |
| Finance and insurance | 10 - 25 TL | 5 - 15 TL |
| B2B and SaaS | 8 - 20 TL | 4 - 12 TL |
| Food and restaurants | 2 - 6 TL | 1 - 4 TL |
A 20 TL click is normal in insurance, where one policy is worth thousands over its lifetime. The same 20 TL bankrupts a shop selling 150 TL phone cases. The useful question is not whether CPC is high, but where CPC sits relative to the profit one customer brings you.
Four moves that pull CPC down
- Build a negative keyword list. In retail, terms like "free", "second hand", "complaint" and "jobs" eat most of the wasted spend
- Raise your quality score. When the keyword in your ad copy also appears in the landing page headline, Google charges you less per click
- Turn off weak placements. Meta's Audience Network often delivers cheap clicks and almost no orders
- Split retargeting into its own campaign. Warm audiences routinely click at half the cost of cold traffic
When you decide how to divide spend across the two platforms, working out the Google Ads and Meta Ads budget split separately gives a sharper answer than staring at one dashboard.
CPM: The Price of a Thousand Impressions
CPM = (total spend / impressions) x 1000. You spent 100 TL for 5,000 impressions, so your CPM is 20 TL. This is not a performance metric. It is a price tag that tells you how contested your target audience currently is.
Season doubles the number
| Period | Meta feed CPM | What is happening |
|---|---|---|
| February to April | 15 - 25 TL | Low competition, ideal for testing |
| June to August | 20 - 35 TL | Holiday season, attention scattered |
| November sales week | 45 - 80 TL | Everyone bidding on the same audience |
| Pre new year | 40 - 70 TL | Gift searches peak |
Tripled CPM during November is not your fault. Planning is the only defence. Warm your audience lists in October, then sell to a ready audience in November. Running brand awareness to a cold audience during the most expensive week of the year means buying the cheapest job at the highest price.
Keeping CPM under control
- Set a frequency cap. Showing the same person an ad more than three times a week inflates CPM without adding clicks
- Do not over-narrow the audience. With 20,000 people to choose from, the algorithm runs out of options and impressions get pricier
- Separate placements. Stories usually runs 20 to 30 percent above feed, and mixing them in one ad set can push budget toward the expensive side
- Add video creative. A video that stops the scroll in three seconds earns more qualified impressions from the same budget
CPA: The Metric That Sets Your Profit Line
CPA = total spend / conversions. You spent 1,000 TL, made 20 sales, so each sale cost 50 TL. The other three metrics hold your attention. This one pays your bills.
Work out your maximum CPA
One rule covers it: CPA must stay below your net profit per order. Net profit is what remains after product cost, shipping, payment commission and your returns allowance come out of the sale price.
| Sale price | Net profit | Maximum CPA | Actual CPA | Result |
|---|---|---|---|---|
| 200 TL | 80 TL | 80 TL | 50 TL | 30 TL profit per order |
| 100 TL | 20 TL | 20 TL | 35 TL | 15 TL loss per order |
| 1,200 TL | 420 TL | 420 TL | 260 TL | Ready to scale |
The second row is the one dashboards hide best. A 35 TL CPA reads as respectable against any benchmark while the store loses money on every order. Marketplaces add another line to the math: Trendyol and Hepsiburada take a category-based commission that eats into the same margin, so your maximum CPA there is lower than on your own site. Selling one product through both channels means running two separate CPA targets.
Most CPA fixes live outside the ad platform
- Show the shipping fee on the product page, not at the cart. Surprise delivery costs are the most common reason people abandon
- Use a familiar checkout such as iyzico or PayTR, and put the instalment option directly under the price
- Drop forced registration. Guest checkout lifts mobile conversion noticeably
- Print the delivery window on the product page. Yurtici, Aras or MNG, it barely matters, seeing a date speeds up the decision
Getting more orders out of the same ads almost always comes from lifting your conversion rate rather than from adjusting bids.
How the Four Metrics Feed Each Other
These numbers are not independent. One formula ties them together: CPA = CPC / conversion rate. You lower your cost per sale either by making clicks cheaper or by making the site better. There is no third route.
| CPC | Conversion rate | Resulting CPA |
|---|---|---|
| 8 TL | 1 percent | 800 TL |
| 8 TL | 3 percent | 267 TL |
| 4 TL | 1 percent | 400 TL |
| 4 TL | 3 percent | 133 TL |
Look at the spread. Halving the click price halves CPA. Moving conversion rate from 1 to 3 percent cuts it to a third. Work on the site beats work on the bid strategy nearly every time.
Three diagnoses, three different owners
- Low CTR with high CPC: the creative is the problem. Rewrite the headline and change the image, leave targeting alone
- High CTR with high CPA: the landing page is the problem. The ad sets an expectation the site fails to meet
- Low CTR with low CPA: a small but accurate audience. Raise budget in steps and do not touch the creative
To see the profit side in a single figure, put ROAS next to CPA. Together they separate a genuinely cheap sale from a cheap sale with a tiny basket. Knowing how return on ad spend is calculated lets you read one campaign from two angles.
Setting Up the Right Columns
The default dashboard view is useless. Google Ads opens on impressions and clicks and tells you nothing about profit. Ten minutes of setup fixes that for good.
Google Ads column setup
- Open the Campaigns screen from the left menu in Google Ads
- Click the columns icon at the top right of the table, then Modify columns
- Under Performance, add Clicks, Impr., CTR and Avg. CPC
- Under Conversions, add Conversions, Cost / conv. and Conv. rate
- Under Competitive metrics, add Impr. share, which shows whether budget or bid is holding you back
- Save the layout so you are not rebuilding it on every login
Meta Ads Manager setup
- Pick a campaign in Meta Ads and open the Columns menu at the top right
- In Customize Columns, tick Impressions, Reach, Frequency and CPM
- Add Cost per result and Purchase ROAS, and keep the two side by side
- Open Breakdown by Placement to see the CPM gap between feed and Stories
- Check the age breakdown weekly to see which bracket your budget is draining into
The two platforms will never agree. Meta counts a conversion up to seven days after a view, while Google Analytics leans on last click. A gap of 15 to 30 percent is expected. Pick one source of truth and report everything through it.
Splitting a 30,000 TL Budget in the Turkish Market
Numbers stay abstract until they sit in a plan. Take a home textiles store with a 30,000 TL monthly budget, an average basket of 450 TL and a 35 percent net margin. That puts the maximum CPA at 157 TL.
| Line item | Budget | Target metric | Expected outcome |
|---|---|---|---|
| Google brand search | 3,000 TL | CPC under 2 TL | 1,500 clicks, cheapest orders |
| Google Shopping | 12,000 TL | CPA under 130 TL | Around 90 orders |
| Meta retargeting | 6,000 TL | CPA under 80 TL | Around 70 orders |
| Meta cold audience | 7,000 TL | CPM under 30 TL | Audience growth, 25 orders |
| Creative testing | 2,000 TL | CTR above 1.5 percent | Four new variations a month |
Local rules change the math
Build your retargeting lists with explicit consent under KVKK, the Turkish data protection law. If you win customers back through email and SMS, sending commercial messages to numbers not registered in the IYS system carries a penalty, and a consented list converts better anyway. On the marketplace side, Trendyol and Hepsiburada both sell their own ad inventory. When you run marketplace ads and Google Shopping for the same product, measure the CPA of each channel separately instead of blending them into one average.
For brands caught between their own store and the marketplaces, the usual order is volume on the marketplace, margin on your own site. When speed and checkout improve on the infrastructure built for ecommerce brands, the same ad budget produces more orders, because the denominator in the CPA formula gets bigger.
Four Common Mistakes and a Weekly Routine
Habits that burn budget
- Locking onto one metric. Accounts that force CPC down to 2 TL by buying irrelevant traffic watch CPA climb every time
- Deciding on three days of data. A campaign needs at least 50 conversions or two weeks before it says anything reliable
- Putting likes and impressions in the report. Those numbers do not invoice anyone
- Measuring without auditing conversion tracking. A tag that fires twice makes CPA look half what it is
Forty minutes a week
- Monday: compare the last 14 days of CPA against your maximum CPA and pause every ad set above the line
- Wednesday: review creative CTR, switch off the two weakest ads and launch two new variations
- Friday: open the search terms report and push irrelevant queries into the negative list
Start tomorrow morning with one task: calculate your net profit per order, write that figure down as your maximum CPA, and pause every ad set sitting above it. Move the freed budget into the sets below the line, then run the same check again next week.
Frequently Asked Questions
What is the core difference between CTR, CPC, CPM and CPA?
CTR is a ratio and it measures interest. CPC and CPM are unit prices, one for a click and one for a thousand impressions. CPA is an outcome cost, the total you paid for a single sale or signup. Together they track four separate points from the top of the funnel to the bottom.
Which metric should I prioritise?
It depends on the campaign objective. Awareness campaigns live on CPM and reach. Traffic and consideration campaigns are judged on CTR and CPC. Sales campaigns answer only to CPA and ROAS, with everything else serving as diagnostic detail.
What is a good CTR?
It varies by channel. Above 3 percent is strong on Google Search, while 0.5 percent already counts as healthy on display. Meta feed ads sit comfortably between 1 and 2 percent. Your own 90-day average is a more honest reference point than any published benchmark table.
My CPC is high, should I pause the campaign?
High CPC on its own is not a reason to stop. CPA makes that decision. Even at 20 TL per click, a campaign is profitable if the cost per sale stays under your net margin. Check the conversion rate first, then touch the bid strategy.
Should I choose a CPM or a CPC campaign?
Launching a new brand or a new product line usually justifies an impression-focused setup for the first few weeks, since it reaches a wider audience faster. With a sales target, choose a conversion-focused setup so the platform steers bids toward likely buyers. Do not mix the two objectives inside one ad set.
How do I set my maximum CPA?
Subtract product cost, shipping, payment commission and your returns allowance from the sale price. What remains is your net profit per order, and that is your CPA ceiling. You can deliberately exceed it for a period to win new customers, but do it knowingly and put an end date on it.
When should I switch on target CPA bidding?
The campaign needs at least 30 to 50 conversions in the past 30 days. Below that volume the algorithm has nothing to learn from and costs swing wildly. Confirm your conversion tracking counts each order once, because a double-firing tag corrupts the target from day one.
Why do Google and Meta numbers disagree?
The platforms use different attribution models. Meta credits itself for conversions up to seven days after a view, while Google Analytics generally works on last click. A 15 to 30 percent gap is the expected outcome. Choose one source as your reference and read every report through it.
Can I measure these metrics meaningfully on a small budget?
CTR and CPC become readable within a few days even at 300 TL per day. CPA needs more volume, so leave the campaign alone until at least 30 conversions have accumulated. On a small budget, running fewer campaigns produces answers faster than splitting the same money across five ad sets.
Can I use the same CPA target for marketplaces and my own site?
You should not. Marketplace commission reduces your net profit, so the CPA ceiling there sits lower than on your own store. Set two thresholds and keep the reporting separate. Otherwise the profitable channel quietly covers the losses of the other one and you never see the problem.