How to measure WhatsApp campaign ROI: metrics, UTM links and attribution that holds up

WhatsApp campaign ROI answers one question: did the money and attention you spent on a campaign come back as profit? The formula is simple, (revenue caused by the campaign − campaign cost) ÷ campaign cost, but both halves are easy to get wrong. This article explains which numbers a WhatsApp campaign report gives you, how to connect them to sales with UTM links, coupon codes and a holdout group, and how to track Click-to-WhatsApp ads back to purchases. It is written for marketers and owners who already send campaigns and want numbers they can defend.

What the campaign report tells you, and what it does not

Every WhatsApp campaign produces message-level statuses from Meta's webhooks. In Broadchat, open CRM → Bulk Campaign and click a campaign to see:

Each recipient row also shows the sent, delivered, read and replied times. You can filter by Read Status, Reply Status, Replied In (24h, 48h or 7d) and Reply is (Stop Notification or Other), then Retarget the filtered group in a follow-up campaign. Export CSV downloads every message of the campaign, and Export Report via Email sends the same file to your login email.

None of these is revenue. A campaign with a high read rate can sell nothing; a campaign with few replies can drive many website orders. Treat the report as the top of the funnel and bring in sales data for the rest.

Tag every link with UTM parameters

Most campaign revenue happens off WhatsApp: on your website, app or store. To connect it, add UTM parameters to the URL in your template's URL button. Google Analytics and most other analytics tools read them automatically. Broadchat does not track link clicks itself, so this tagging is what makes the click and the order visible in your analytics.

Meta's template components allow two kinds of URL button:

• Static URL: the full link is fixed in the template, for example https://rangolihome.in/diwali?utm_source=whatsapp&utm_medium=broadcast&utm_campaign=diwali_2026.
• Dynamic URL: the template stores the base URL and a {{1}} suffix you fill per send, for example https://rangolihome.in/{{1}} with the suffix diwali?utm_source=whatsapp&utm_medium=broadcast&utm_campaign=diwali_2026&utm_content=segment_a. This lets one approved template carry a different campaign or segment tag each time.

Keep a naming convention and stick to it:

Use coupon codes for sales you cannot tag

UTM links miss orders placed by phone, in the shop or by replying I'll take the blue one. Give each campaign its own coupon code, such as WADIWALI26, and use it nowhere else. When the code appears on a bill, you know where the customer came from. For replies that turn into orders in the chat, tag the contact with the campaign name in the inbox so the order can be linked later.

Measure lift with a holdout group

Your best customers often buy during Diwali whether or not you message them. If you count all their orders as campaign revenue, ROI looks better than it is. A holdout group fixes this:

1. Build the campaign segment of opted-in customers in CRM → Segment.
2. Randomly set aside about 10% of it, for example by exporting the list and moving every tenth row to a separate tag, and send the campaign only to the rest.
3. After the campaign window, compare the purchase rate of the two groups from your sales data.
4. The difference in purchase rate, multiplied by the size of the sent group, is the number of orders the campaign caused; multiply by the average order value for revenue.

A holdout costs you some sales in the short term, so use it on recurring campaigns where the learning pays off, not on a one-off launch.

Attribute Click-to-WhatsApp ads to purchases

When someone taps a Click-to-WhatsApp ad, their first message carries a referral from the ad, including a click ID (ctwa_clid). Broadchat saves the ad details and the click ID on the contact. When the lead later buys or qualifies, the chatbot step Send Conversion Event reports it to Meta's Conversions API with an event such as Purchase, LeadSubmitted or QualifiedLead, plus optional value and currency. The dataset it reports to is set on the channel (Channels → Conversions API dataset).

This matters because Ads Manager otherwise only sees chats started, not sales. With conversion events coming back, you can compare ads by purchases and let Meta optimise towards people who buy. Setup details are in our guide to setting up Click-to-WhatsApp ads.

A worked ROI example (illustrative)

Illustrative figures. Rangoli Home sends a Diwali bedsheet campaign to an opted-in segment of 10,000 customers, holding back 1,000 as a holdout. It sends to 9,000; 8,700 are delivered.

Revenue caused. The extra purchase rate is 1.5 percentage points. 1.5% of 9,000 is 135 orders caused by the campaign, or ₹3,17,115 in revenue. At a 40% gross margin, that is ₹1,26,846 in gross profit.

Cost. Meta’s rate card effective 1 Oct 2026 lists ₹0.8631 per delivered marketing message to India, excluding GST, so 8,700 deliveries cost ₹7,508.97 in Meta charges. Add the provider’s markup and a share of the subscription (say ₹2,000 here), creative work (₹3,000) and the 10% discount, which every one of the 315 buyers used: ₹73,994 on ₹7,39,935 of orders. Total: about ₹86,503.

ROI. (₹1,26,846 − ₹86,503) ÷ ₹86,503 ≈ 0.47, or 47%. Counting all 315 orders as campaign revenue (₹2,95,974 gross profit) would have shown about 242%. The gap comes from two things a simple report hides: 180 of the buyers would probably have ordered anyway, and they still received the discount.

Run your own cost side through the WhatsApp pricing calculator, and see WhatsApp API pricing for how wallet charges work.

Pitfalls that distort ROI

• Counting reads as results. Reads are undercounted (read receipts) and are not purchases.
• Ignoring failures. Per-user marketing limits cause some 131049 failures on most sends; those people saw nothing, so do not count them in the sent group. See error 131049 explained.
• An attribution window that is too long. Choose the window before the campaign (for example 7 days for a festive sale) and stick to it.
• Leaving out the discount. A 20% coupon on orders that would have happened anyway is a cost, not a gain.
• Ignoring opt-outs. Check the Reply is → Stop Notification filter after every send; a rising stop rate is a cost to future campaigns.

Building a measurement habit

Measure every campaign the same way: report export, UTM-tagged links, a coupon code, and a holdout on recurring sends. After four or five campaigns you will know your normal reply and purchase rates and can spot a weak campaign within a day. This is part of the wider WhatsApp marketing strategy; for audience and timing choices, read the broadcast best practices. Agencies reporting ROI to many clients can run campaigns and exports under their own brand through the partner program.