Why Meta's Lead Count Doesn't Match Your Telegram Sales

What Meta is counting

Meta reports the event you chose to optimise for, based on what it can observe. If that event is a landing-page view or a link click, a lead is anyone who tapped the ad. If it is a form submission, a lead is anyone who submitted the form, including duplicates and automated traffic. None of these counts indicates whether the person opened Telegram, stayed in the group, replied to a message or paid. Meta reports accurately up to the limit of what it can see.

Where the drop-off occurs

Click to Telegram: some people who click never open the app, because of the device they are on, a slow page or a change of mind.
Telegram to join: some people open the invite and do not join.
Join to conversation: members who never engage are members, not leads.
Conversation to deposit: typically the longest step with the largest drop-off, and the only one that produces revenue.
Attribution window: if the typical time from join to deposit is two weeks and the attribution window is seven days, the deposit exists but Meta does not credit it.

Replace the lead count with a funnel

Clicks, recorded through tracking links so that each has an id.
Bot starts or invite opens.
Joins, matched deterministically to clicks.
Messages to the admin or bot.
Deposits, with value, tied to the same click.

Read as conversion rates between stages, this table shows where the 300 leads became 6 sales and which stage needs attention. It also provides two more meaningful metrics to report in place of leads: cost per join and cost per deposit.

How to fix it

Change the optimisation event from link clicks to a Telegram outcome sent through the Conversions API: a join mapped to CompleteRegistration or, once volume allows, a deposit mapped to Purchase.
Give each campaign its own tracking link so the funnel can be split by source.
Report deposits as they happen, with value, so bidding shifts towards paying customers.
Reconcile Meta's figures with your own funnel each week. Some divergence is normal; increasing divergence is a warning sign.
Report joins, deposits and cost per deposit to stakeholders instead of Meta's lead column.

Example

Illustrative example: a trading community optimises for link clicks and reports 300 leads a week. After switching to joins sent through the Conversions API, Meta reports 140 registrations, which matches the group's actual growth. Once deposits are also reported as Purchase events, the campaign receives the signal it needs to optimise for people who resemble depositors rather than people who only click. The team then reports cost per deposit in place of the lead count.

Practices that widen the gap

Treating the Ads Manager figure as authoritative for a funnel it cannot observe.
Selecting ad sets by cost per click when lower-cost clicks convert less well.
Not sending conversions back, so the same targeting issues persist from week to week.
Sharing one invite link across campaigns, which removes the ability to compare them.

How Tellmetric reduces the gap

Tellmetric records every stage from click to deposit against a single click id, shows the funnel by campaign with drop-off between stages, and sends joins, messages and deposits to Meta as standard events with deduplication and hashed identifiers. The match-quality panel shows how much of your data Meta can use. Some difference between Ads Manager and your own figures remains, but it becomes explainable and typically narrows as the platform optimises on actual outcomes.

Frequently asked questions

Will Meta's reported conversions match my dashboard exactly?

No. Meta applies its own attribution windows and modelling. Meta will usually show somewhat fewer conversions than you record; a large gap usually indicates weak match quality.

Is a join a good optimisation event?

It is a good starting point because of its volume. Move to deposits once you see roughly fifty a week per ad set.

What about bot and click-farm traffic?

Tellmetric filters out events with bot user agents or no matchable signal before delivery, so they do not influence the platform's optimisation.

Why Meta's lead count and your Telegram sales differ