How to measure Telegram Ads performance: CTR, CPL, ROAS and more
In this guide
To judge Telegram Ads performance, compare spend with the outcome the campaign was built for. A channel may track cost per subscriber; a service may need cost per qualified enquiry; a shop may need cost per new buyer. Impressions and clicks help locate funnel problems, but they do not prove commercial return on their own.
The formulas below work in a manual report. Use comparable time periods, one currency and the same definition of included costs.
What data should you collect?
Get spend and delivery metrics from the ad account, target actions from your site or bot analytics, and lead quality, new customers and paid orders from your CRM.
Do not substitute a similar-looking event for a missing one. A bot start is not an enquiry; a registration is not a purchase. Define the event first and label it accurately.
If you cannot reliably attribute a sale to ads, state that limit. A company’s entire monthly revenue cannot be assigned to Telegram Ads by default.
Telegram Ads metrics and formulas
| Metric | Formula | What it tells you |
|---|---|---|
| CPM | Spend ÷ impressions × 1,000 | Cost per thousand impressions |
| CTR | Clicks ÷ impressions × 100% | Share of impressions that produced a click |
| CPC | Spend ÷ clicks | Cost per click |
| Cost per follower | Spend ÷ new tracked subscribers | Cost per subscriber under your tracking method |
| CPL | Spend ÷ enquiries | Cost per lead |
| CPA | Spend ÷ defined target actions | Cost per specified action |
| Conversion rate | Target actions ÷ entries to that funnel stage × 100% | Conversion of a specific stage |
“CPA” does not say which action was counted. Write “CPA for registration” or “CPA for booking”. Likewise, a click-to-lead conversion rate answers a different question from a lead-to-purchase rate.
If the denominator is zero, the unit cost is undefined. Spending €52 with no leads does not produce a €0 CPL; that would falsely suggest free enquiries.
Worked example: from impressions to customers
Assume a test spent about €125 and delivered 100,000 impressions, 800 clicks and 40 enquiries. Sixteen enquiries met the business criteria and six new customers paid a total of about €625. These are teaching figures, not our agency’s results.
| Metric | Calculation | Result |
|---|---|---|
| CPM | €125 ÷ 100,000 × 1,000 | €1.25 |
| CTR | 800 ÷ 100,000 × 100% | 0.8% |
| CPC | €125 ÷ 800 | About €0.16 |
| CPL, all enquiries | €125 ÷ 40 | About €3.13 |
| Cost per qualified enquiry | €125 ÷ 16 | About €7.81 |
| Click-to-enquiry conversion | 40 ÷ 800 × 100% | 5% |
| Ad spend per new customer | €125 ÷ 6 | About €20.83 |
The final row is not necessarily full customer acquisition cost. Full CAC would include any other acquisition costs you decide belong there, such as staff or agency work. This table includes media spend only.
Did the advertising pay back?
In the example, revenue divided by ad spend is €625 ÷ €125 = 5, or 500% ROAS. That means €5 of attributed revenue for each euro of ad spend. It does not mean €5 of profit.
To assess profit, include product costs. Suppose roughly €187 remains from revenue after product and variable costs. Subtract the €125 ad spend and about €62 remains before other costs. High ROAS does not automatically mean healthy profit.
Do not mix actual sales with hoped-for repeat purchases. If the economics work only when people buy again, put that forecast and its assumptions in a separate line.
Find the weak point in the funnel
Many impressions and few clicks may signal a weak offer or a poor context. Compare messages on similar audiences before concluding from one CTR.
Clicks with few enquiries call for a check of message-to-page fit, clarity and forms. Submit a test enquiry on a phone. Spending more on a broken form only increases waste.
Enough enquiries but few sales points towards lead quality, response speed, pricing or offer terms. With proper access and privacy safeguards, inspect a sample of conversations before rewriting every ad.
Compare ads without misleading averages
Use the same time window and result definition. A subscription campaign is not “better” than a purchase campaign merely because a subscriber costs less than a buyer.
For overall CPL, divide total spend by total leads. Do not take the simple average of each ad’s CPL when their volumes differ. If one ad spends about €10.40 for one lead and another spends about €93.70 for 90 leads, combined CPL is roughly €1.14—not the roughly €52 average of the two individual CPL figures.
Allow for a sales delay. If buyers typically decide over two weeks, judging yesterday’s clicks solely by today’s orders is premature. Set an observation window and preserve the link between acquisition date and later outcome.
What should a business owner see in the report?
Start with a table by ad or audience group showing spend, target actions, qualified enquiries, new customers and attributed revenue. Add technical metrics to diagnose changes, not just to fill a dashboard.
Record the decision and why. “Pause audience A because none of 12 enquiries fits the product” is more useful than a list of percentages going up and down.
Want to connect Telegram Ads spend to leads and sales? Message our team with spend data, the target-action definition and anonymised lead outcomes. We can help make the measurement meaningful.
Source
The numerical examples above are illustrative; the formulas and calculations are shown so you can reproduce them with your own data.