WhatsApp Marketing Metrics: What to Track and How to Read It

How do I track opt-outs?

Treat opt-outs as a metric and a stop signal. Count customers who ask to stop and customers who stop marketing in WhatsApp.

Meta's documentation describes a user setting for marketing messages. If a customer has stopped marketing templates and you send one, the API accepts the request but does not deliver it, and a status webhook reports error code 131050. You can subscribe to the user_preferences webhook to learn when a customer stops or resumes marketing messages. Meta also notes that a customer choosing "Not interested" does not trigger that webhook, but it can affect per-user marketing limits.

Add your own handling for typed requests such as "stop", record the consent change, and exclude those numbers from future audiences. See opt-in and consent. A rising opt-out rate after a campaign means the audience, frequency or content was wrong.

Which health metrics protect my sending?

Quality rating and template status are health metrics: they change before delivery does.

• Phone-number quality and messaging limits control how many people you can reach. Meta's mechanism is covered in messaging limits and quality rating; this article does not repeat it.
• Template quality is rated GREEN, YELLOW, RED or UNKNOWN (new templates without enough feedback). WhatsApp Manager shows approved templates as Active with pending, high, medium or low quality.
• Template pausing. Meta says quality ratings factor into template pacing and pausing, which can affect delivery, and a template whose status is not APPROVED cannot be sent in template messages. For pause durations, read Meta's current template-pausing documentation.

Review these before every large send. A template with a falling rating should be rewritten before the next campaign, not after it is paused. See template categories and approval.

How do I attribute sales to a WhatsApp campaign?

Give every campaign a way to be recognised downstream, then match it in your analytics or CRM.

1. UTM parameters on each link: source, medium and a campaign name that matches your campaign label.
2. Unique codes or links per segment so a purchase can be traced without cookies.
3. Phone-number matching. If the customer replies or checks out with the same number, join orders to the campaign recipient list.
4. Click-to-WhatsApp referral data. When a chat starts from a click-to-WhatsApp ad, Meta's inbound message webhook carries a referral object with the ad's source_id, source_url, headline and a ctwa_clid click ID. Meta omits ctwa_clid for ads placed in WhatsApp Status. Store these fields with the conversation. See click-to-WhatsApp ads.
5. A review window. Decide in advance how many days after a send a purchase counts, and keep it the same across campaigns.

For link tracking mechanics, the same principles apply as in SMS link tracking.

What does a sample WhatsApp campaign dashboard look like?

One row per campaign and template. The figures below are placeholders for the layout, not benchmarks.

Add a second tab with failures grouped by error code, and a third with cohort conversion by days since send. OmniChannels.ai offers delivery reports in summary and advanced views with filters and CSV export, and the bulk-insights endpoint for per-recipient WhatsApp status.

Worked example: a Riyadh electronics retailer

"Najd Electro" (synthetic) sends a weekend offer to 20,000 opted-in customers in two Arabic template variants.

1. Each variant has a campaign label and a UTM campaign name, such as weekend_a and weekend_b.
2. Delivered counts come from status events; the team computes delivery rate and sees one segment with many failures, which they trace to numbers that had not messaged them in a long time.
3. On day 2, they export template analytics, because read and click data are kept for only 7 days.
4. Variant B has a higher click-through rate but also more opt-outs. They compare orders per delivered message, not clicks alone.
5. The winning template goes into the next campaign. The losing one is rewritten after a quality check.

What are the common mistakes?

• Using sent as the denominator. Undelivered messages inflate the apparent weakness of the content.
• Treating read as total attention, and dismissing a campaign because reads look low.
• Waiting past 7 days to export template read and click data.
• Quoting industry benchmarks without a source. Compare against your own history.
• Ignoring opt-outs and quality. A good click rate that damages quality costs you reach later.
• Last-click only. A WhatsApp message may prompt a later purchase on another channel; use a consistent attribution window and say what it is.
• Limits: Meta can change analytics definitions and retention. Confirm them in Meta's documentation.

How OmniChannels.ai helps

OmniChannels.ai campaigns run on every channel with quick send, bulk, personalised and auto-schedule modes, and delivery reports carry normalised statuses into one webhook format so WhatsApp, SMS and RCS can be reported side by side. Reports come in summary and advanced views with filters and CSV export, and the shared inbox keeps replies attached to each customer. For Meta analytics fields beyond these, ask our team. See campaigns, the WhatsApp channel and the webhooks reference.