What a Lead from Instagram, Telegram and WhatsApp Really Costs: Calculating It with End-to-End Analytics

In our experience at Brutal Marketing, on most projects that drive traffic into messengers, cost per lead and cost per customer point in opposite directions. The reason is simple: the ad account sees an inquiry, but the money comes from a closed deal. Between the inquiry and the deal sit qualification, the sales rep's work, and the sales cycle — and the ad account sees none of it.

Below is a step-by-step calculation on real project numbers. What a lead costs in each messenger, how to calculate CAC by channel including sales team salaries, which fields you need in your CRM for these numbers to come together at all, and what to do with the result. Copy the table and plug in your own data.

Meta Ads counts "results." A result is whatever event you selected when setting up the campaign: a click on the Message button, a conversation started, a first message sent. None of these means you now have a lead.

The gap is technical. The ad account records an action on its own side and has no idea what happened next. Someone wrote "how much?" and vanished — that's a result. Someone messaged from two devices — two results. A competitor asked for your price list — also a result. A spam bot promoting crypto — result again.

Here is what the gap looked like on a custom kitchen manufacturer project (average order value $8,200–9,600) where we built out end-to-end analytics:

Cost per "result" — $23. Cost per inquiry — $30. Cost per qualified lead — $107. Cost per deal on ad spend alone — $750. Four different numbers for one channel, each of them true in its own context. The trouble starts when budget decisions get made on the first one.

Before any calculation, put the definition in writing. We ask every client one question: what is the minimum action a person has to take before you're willing to pay money for them?

On most projects the working definition looks like this: a unique contact who asked about your service and left a way to reach them. Not "viewed," not "followed," not "reacted to a story."

From there, the definition has to live as a status in the CRM, not in the marketer's head. As long as everything counts as a lead, your CPL will be beautiful and meaningless at the same time.

One more condition, without which the math never comes together: the source lands in the record automatically. We run into the same situation constantly — the field is filled in manually by a rep, and a month later 60% of deals carry the source "word of mouth" or "other."

This isn't sabotage. For a rep juggling 30 conversations a day, picking the first value in the dropdown is faster than working out where the person came from. The fix is technical, not motivational: the integration fills the field, and editing is locked.

A mistake we see regularly: a company sets up end-to-end analytics "for ads" and assumes messengers will follow along. They don't. Each channel loses the source in its own way, and each hole needs a different tool to close it.

On Instagram, an inquiry reaches you through click-to-Direct from an ad, through Direct from the organic profile, through comments under reels, and through the link in your bio. Out of those four paths, the ad account confidently sees one.

The biggest hole is reels and stories that generate inquiries a week after the impression. Someone watches a video, follows you, and writes ten days later. In the CRM that shows up as "Instagram Direct" with no detail. Formally, organic. In reality, the same ad budget with a delayed effect.

The second hole is comments. A customer writes "price?" under a post, the rep replies in Direct, and the inquiry enters the system as an incoming message with no tag at all. On content-heavy projects that loses 10–15% of leads.

What we do: separate click-to-Direct from everything else using ice-breakers and automated first messages carrying different tags, and route the bio link through properly structured UTM tags to a landing page with a messenger widget. For comments, we add an automated reply that moves the conversation into Direct — it applies a tag too. This doesn't close 100% of the gaps, but it lifts Instagram attribution from a typical 40% to 80–85%.

Telegram is technically the most convenient channel for analytics and, in practice, the worst configured. The reason is always the same: companies use a single t.me/companyname link for every campaign at once.

A deep link with the ?start= parameter solves it completely: t.me/your_bot?start=ig_reels_kitchens_july. The bot receives that tag in the very first message and passes it to the CRM along with the contact. Attribution is 100%, with no guesswork and no "so how did you hear about us?"

The same mechanism works for placements in third-party channels, QR codes on printed materials, and broadcasts. Every placement gets its own tag, and a week later you know exactly which one produced inquiries and which one produced only views.

The limitation: the parameter works with a bot, not with a rep's personal account. If your rep runs conversations from their own number, you lose the tag, the message history, and the contact itself the day they leave. We move those projects to a bot-plus-CRM setup first — it adds more data than any other configuration change.

Click-to-WhatsApp from Meta Ads passes a campaign identifier into the webhook — provided you run on the WhatsApp Business API rather than the app on a rep's phone. In the app you see a phone number and nothing else.

Second quirk: the attribution window for conversations is narrower than for standard conversions. A customer clicks today and writes three days later, and the link is gone. In niches with a long decision cycle, that loses up to a third of inquiries.

The workaround is a tag inside the pre-filled message text: "Hi! I'm interested in a kitchen [WA-GOOGLE-07]." It looks technical, so we place it at the end of the line or pass it through a ref parameter in the link. The rep never touches it; the CRM parses it automatically into the source field.

A separate note on Google Ads: WhatsApp often receives traffic from search rather than Meta. There, the only thing that works is a landing page with a messenger button and a UTM tag injected into the message text by script. Without it, all your paid search collapses into a single source called "WhatsApp."

Before you calculate anything, agree on terms. Half the arguments between owners and marketers happen because they mean different things by "cost per lead" — and both are right.

CPL (Cost Per Lead) — ad spend divided by unique inquiries. Shows how cheaply a channel brings people in. Says nothing about their quality.

CPQL (Cost Per Qualified Lead) — ad spend divided by qualified leads. Qualified means matching your criteria: a real need, a budget, a timeline, and the authority to decide. This requires written lead qualification criteria that every rep applies the same way, otherwise the number drifts with whoever is on shift.

CAC (Customer Acquisition Cost) — all acquisition costs divided by new customers. The operative word is "all": ads, sales team salaries, CRM and telephony subscriptions, agency fees. CAC is what you compare against LTV, and it's the number almost nobody calculates per channel.

The formula:

Channel CAC = (channel ad spend + allocated sales team costs) / deals from that channel

Allocate sales team costs in proportion to qualified leads — that's where reps actually spend their time, not on spam. It's the simplest method that works, and for most companies it's enough.

A more precise option is allocation by actual time logged in the CRM: the system counts calls, messages and meetings per deal. You need it when channels produce leads of fundamentally different complexity — an inbound search request that closes in two touches versus a cold lead from paid social that takes twelve.

The crudest option is allocation by total inquiries. It skews the picture in favour of channels full of spam, because a rep spends 30 seconds on an irrelevant message while the system counts it the same as everything else. Use it only if you haven't set up qualification yet.

The benchmark we work from: LTV / CAC ≥ 3. Below 3, the channel loses money once fixed costs are included. Above 8, you're under-investing and handing market share to competitors.

Calculate LTV on gross profit, not revenue. A customer with a $10,000 order and a 20% margin brings the company $2,000 — that's the number that belongs in the numerator. We covered how to connect CRM, LTV, CAC and ROI into one reporting system separately.

The data below comes from the custom kitchen manufacturer. Three channels, one quarter, a CRM with end-to-end analytics and tracking built the way described above. Numbers are rounded; ratios are intact.

At this stage the picture is unambiguous: Instagram delivers the cheapest inquiries, Telegram the most expensive, a 1.7x difference. Nine companies out of ten stop here and make their budget call.

The gap between channels has almost disappeared. Instagram brings in three times more people, but two-thirds of them are "what's your price per metre?" with no project, no budget and no intention of buying any time soon. Telegram brings fewer, but half of them make it to a site measurement.

The explanation is simple and not especially flattering to marketing: Instagram is a channel of impulsive reactions to attractive images; Telegram is a channel for someone already reading industry channels and comparing manufacturers. Those are different stages of readiness, not "good" and "bad" traffic.

Second surprise: average order value differs too. A Telegram customer orders $1,400 more than an Instagram customer — more often taking the full set with cabinetry and appliances, and negotiating less at contract stage.

Sales team costs for the quarter: payroll for three reps and a head of sales — $19,500; CRM, telephony and tools — $1,500; agency fees for advertising — $4,500. Total $25,500. Divided by 323 qualified leads, that's $79 per qualified lead.

The ranking has flipped entirely. The same Instagram that delivered a $30 lead delivers a customer at $1,303 — nearly twice the cost of Telegram with its "expensive" $50 leads.

Gross margin on this project is 35%. LTV is calculated over 24 months, including repeat orders (wardrobes, closet systems) and deals from referrals.

Three conclusions that only appear at this level of detail:

1. Instagram is running at the edge of unprofitability. LTV/CAC of 2.5 against a benchmark of 3. The channel produces the highest revenue and the smallest contribution to profit — the classic trap for companies that watch turnover.
2. Telegram is underfunded by a factor of two to three. LTV/CAC of 6.6 means you can keep raising budget until the ratio drops to 3–3.5.
3. WhatsApp is the steady one. Sufficient volume with healthy economics. Leave it alone.

The decision after the calculation looked like this. Instagram stayed on — we cut the budget from $18,000 to $12,000, dropped broad audiences on reels, and kept retargeting and click-to-Direct on warm segments. Telegram went from $6,000 to $13,000, with placements added in renovation and interior channels. WhatsApp got another $1,500.

In parallel we rewrote the automated first message on Instagram: instead of "Hi! How can we help?", three buttons asking qualification questions about renovation stage and timeline. The goal was to filter out "just browsing" before a rep spent 20 minutes on them.

Inquiry volume dropped — and that's a healthy outcome, not a problem. Reps stopped burning time on conversations that went nowhere, and end-to-end conversion to deal rose from 6.0% to 8.4%. Revenue gained a third on a budget increase of under 8%.

Worth stressing: this picture doesn't assemble without allocating sales team costs across channels. Counting ad spend alone, Instagram would have looked like the cheapest way to acquire a customer ($750 versus $400 on Telegram) — and the budget would have moved in exactly the wrong direction.

The main mistake when reading someone else's case study is transplanting the conclusions into your own business. Telegram won in kitchens not because Telegram is better, but because that's where an audience with a high order value and a long comparison cycle happens to sit. Here's how it plays out elsewhere.

Online English school, $1,200 per course. Instagram delivers a lead at $10, Telegram at $16. CAC: $195 versus $148. The decisive difference showed up in repeat purchases: 22% of Instagram students renew, versus 41% from Telegram. Counting second and third purchases, Telegram's LTV/CAC came out twice as strong, even though the two channels looked nearly identical on first sale.

Outsourced bookkeeping, $900 per month, average client lifetime 14 months. Here Telegram barely produces leads and the main channel is WhatsApp via Google Ads. CPL: Instagram $26, WhatsApp $42. But 71% of Instagram inquiries were individuals and micro-businesses outside the target segment. CAC: $1,840 versus $690 against an LTV of $4,410. Instagram doesn't pay back at all in this niche and stayed on purely as a content channel for visibility.

What both cases share: the channel with the more expensive lead turned out cheaper per customer. That isn't a law of nature — it's a consequence of cheap traffic being less qualified almost by definition. But you have to verify it on your own numbers, which is exactly what end-to-end analytics is for, and what someone else's case study can't do for you.

Real customer journeys are rarely linear. Saw a reel → followed → read a post in a Telegram channel a week later → googled the brand → wrote on WhatsApp. One deal, three channels. Who gets credit?

In our kitchen case, last-click gave Instagram 24 deals. First-touch gave it 37 out of 58. In other words, Instagram creates the demand that Telegram and WhatsApp later close. On a first-touch basis, its CAC comes out at $845 rather than $1,303.

That's precisely why we rebuilt the channel instead of switching it off. Killing Instagram would have collapsed inquiry volume across all three channels — and it would have looked like "Telegram suddenly stopped working."

Practical rule: read the report in at least two models at once, first touch and last click. If the numbers agree, your attribution is simple and you don't need to complicate it. If they diverge by 1.5x or more, you have a multi-channel journey, and decisions based on a single model will be wrong.

Now the technical side — what has to be configured for the table above to build in two clicks instead of three days of manual work in a spreadsheet.

One UTM scheme for every channel and campaign, written down in a document the media buyer, the agency and the owner can all access. The most common cause of scattered data isn't missing analytics — it's instagram, Instagram, ig and instsitting in the same report as four separate sources.

The rule that saves months: lowercase only, latin characters only, underscores instead of spaces. Audit the source list in your CRM monthly — new spellings appear every time a new contractor joins the project.

For messengers, add the specifics: ?start= on Telegram, a ref parameter or a tag in the pre-filled text for WhatsApp, and separate landing pages per campaign for Instagram.

All three channels go into the CRM through official integrations, not apps on reps' phones. That buys you three things at once: message history stored in the deal record, source recorded automatically, and the contact staying with the company when someone resigns.

We covered messenger integration with a CRM in detail — the technical specifics of the WhatsApp Business API, the constraints of the Instagram Messaging API, and choosing between a bot and a personal account on Telegram.

Three fields without which the calculation is impossible: source (auto-filled, locked from editing), qualification status from a fixed list of values, and loss reason chosen from a list rather than typed as free text.

Deduplicate by phone number and by username, separately. One person can message you in Direct and duplicate it on WhatsApp two days later. Without merging, you'll count two leads and corrupt CPL for both channels at once.

One more thing that usually surfaces in month two: repeat inquiries from existing customers. They belong in a separate pipeline, otherwise they inflate lead counts and understate CPL where advertising had nothing to do with it.

Automatic cost import from ad accounts at campaign level, plus manual entry for whatever doesn't import: Telegram channel placements, influencers, offline materials. Skip this block and you're calculating CPL for three channels out of five, with a distorted picture as a result.

Fixed sales team costs belong here too. Enter them once a month as a single line; the system handles allocation across channels by the rule you set.

The final table has to refresh itself and be available without asking a marketer for it. The minimum column set is the one used above: spend, inquiries, CPL, qualified, CPQL, deals, revenue, CAC, LTV/CAC.

Sensible review cadence: weekly for operational metrics (inquiries, qualification) and monthly for channel economics. More often is pointless — the sample is too small and you'll mistake noise for a trend.

We have a separate piece with dashboard examples on which reports a manager genuinely needs and which ones gather dust after week two.

1. Last-click attribution. A customer sees a reel, follows you, googles the brand two weeks later and writes on WhatsApp. Last click hands the deal to branded search, and Instagram — which actually created the demand — gets zero. At minimum, read the report in two models.

2. Salaries left out of CAC. The most common mistake of all. Advertising is 30–50% of real acquisition cost. Without the rest, CAC is understated by nearly half and every decision built on it is wrong. Easy check: multiply CAC by deals closed and compare against actual marketing and sales spend. If it doesn't reconcile, something is missing.

3. Month-on-month math with a long sales cycle. If deals close in 45 days, July's spend produces August's deals. Dividing August deals by August budget is arithmetically incorrect. Use cohorts: take July's leads and look at what happened to them over 90 days.

4. Counting every lead indiscriminately. Spam, competitors, misdirected inquiries and repeat requests from existing customers inflate the denominator. CPL looks excellent and never reaches a deal. This is the case where improving the metric means damaging the business.

5. Ignoring repeat sales and referrals. A channel can deliver expensive customers who then return on their own and bring friends. On CPL it loses; on LTV it wins comfortably. That's exactly why the summary table ends in LTV/CAC and not CAC alone.

6. Comparing channels with different jobs in the funnel. A content-led Telegram channel and search advertising solve different problems — one creates demand, the other harvests it. Putting them in the same column on deal conversion is like comparing a truck and a sports car on top speed.

The most common reaction to the table is "switch off the expensive channel." Almost always wrong. Four typical situations.

Low CPL, high CAC. The channel brings in a lot of irrelevant inquiries. The problem isn't the channel, it's audience setup and the first touch. Change creatives and add qualification to the automated message first; only if nothing shifts within a month should you cut the budget.

High CPL, low CAC. The channel is underfunded. Raise budget in steps of 30–40% and watch whether conversion holds. It will break at some point — the only question is where, and your job is to find that ceiling before the money runs out.

Both metrics high. Look at average order value and LTV. If the order value runs 20% or more above your company average, the channel can stay even with an elevated CAC. If not, it's the first candidate to switch off.

A channel that closes almost nothing directly but shows up constantly in first touches. The classic role of Instagram and a content-led Telegram channel. Judge it on assisted conversions and keep a separate top-of-funnel budget for it.

And one rule that saves a lot of frustration: change one variable at a time. If you simultaneously reallocate budget, rewrite scripts and hire a rep, you'll never know which one worked. In the case above we changed two — and separating the effect of the budget shift from the new automated message took a dedicated test.

One more prerequisite: your CRM pipeline stages have to match the actual sales process. If the system has three statuses — "in progress," "thinking about it" and "won" — you won't see where Instagram leads are falling out. We wrote separately about structuring sales pipeline stages in a CRM so they produce usable analytics.

The sequence we run with clients at project kickoff. Realistic if the company already has a CRM and someone owning its configuration.

Days 1–2. Lock the single UTM scheme in a document. List every entry point: every link in a bio, every button in an ad, every QR code on printed material. This stage typically surfaces five to eight channels nobody remembered.

Days 3–5. Connect messengers to the CRM through official integrations. Move Telegram to a bot with the start parameter, WhatsApp to the Business API, Instagram to click-to-Direct with tags and automated replies under comments.

Days 6–7. Create the required fields: source, campaign, qualification status, loss reason. Configure auto-population and lock the source field from rep edits.

Days 8–9. Write qualification criteria in three or four bullet points and train the team. Without a shared understanding of what counts as qualified, next week's data is garbage.

Days 10–12. Connect cost import from ad accounts, enter fixed sales team costs, and configure the allocation rule.

Days 13–14. Build the report with the columns from the table above. Validate it against last month: lead counts by channel must reconcile to the CRM total, and spend must reconcile to actual charges.

The first trustworthy conclusions arrive one full sales cycle plus two weeks later. In most niches that's 6–10 weeks. Don't make sharp budget moves before then.

Go through the list and answer honestly. Every "no" is a place where your calculation falls apart.

1. Source is filled automatically and reps can't change it.
2. Leads from Instagram, Telegram and WhatsApp reach the CRM without manual entry.
3. Duplicates by phone number and username merge automatically.
4. There's a qualification status with a fixed list of values.
5. Loss reason is selected from a list, not typed as free text.
6. Ad spend imports into the system at campaign level.
7. Fixed sales team costs are entered every month.
8. The CAC-by-channel report opens without asking a marketer for it.
9. You can see message history in the deal record, not only on a rep's phone.
10. Repeat inquiries from existing customers go into a separate pipeline.

Seven or more "yes" answers means you only need to finish the reporting layer. Fewer than four means there's nothing to calculate from yet — close the basics first.

If you don't have the time or the person to do this in-house, we handle end-to-end sales analytics setup end to end: tracking, messenger integrations, CRM fields, reporting, and team training.

We at Brutal Marketing will select the best CRM program for you to use in your business. We will be happy to tell you about the program's capabilities and show you which settings will exactly help you achieve the desired financial results.

We'll show you, on your own data, which messenger brings cheap leads and expensive customers — and build a report that updates itself, without a monthly spreadsheet exercise.

Request an end-to-end sales analytics setup using form below.