Guides
Measure a comment-to-DM campaign beyond comment counts
Separate requests, delivery, visits and actual outcomes. Use a small measurement plan to find the step worth fixing.
Decide what success means before the post
A recipe creator may want readers to receive a shopping list. A consultant may want a qualified enquiry. Choose one outcome before the campaign, then record the intermediate steps that tell you where people stop.
Sayw Statistics includes automation metrics such as resources delivered. Use the labels shown in the product carefully: a recorded delivery is not proof that someone read a document, booked a call or paid. Check commercial outcomes in the system that actually records them.
| Event | What it tells you |
|---|---|
| Keyword request | Someone asked for the offer |
| Resource delivered | The delivery step was recorded |
| Destination visit | The linked page recorded a visit |
| Verified booking or payment | A commercial outcome occurred |
Use the same cohort and time window
Compare events from the same campaign and period. Don't divide this week's deliveries by comments accumulated over several months. Also distinguish people from events: one person may click twice, and a page visit may arrive from a forwarded link.
The example below is fictional arithmetic, not a Sayw benchmark. Its purpose is to reveal the question to ask next. A lower ratio alone does not identify whether the cause is copy, eligibility, delivery or measurement.
Give the destination a recognizable campaign label
When linking to a site you control, add a campaign label your analytics can recognize. Keep names readable and consistent, for example utm_source=instagram, utm_medium=dm and utm_campaign=portrait_checklist. Verify that the destination preserves the parameters and records them as expected.
Do not put names, email addresses, private codes or conversation text into tracking URLs. Consent choices, blocked tracking and app browsers can leave gaps. A missing attributed conversion does not prove that nobody booked.
Fix the step with supporting evidence
If requests arrive but delivery is low, inspect a real failed flow before rewriting the offer. If delivery works but the resource disappoints readers, review whether it matches the promise. If qualified enquiries stall, check ownership and the next action in CRM.
Change one element and record the date. Compare a reasonably similar group later, and keep the explanation tentative when the audience or traffic source changed. Small samples can suggest a test; they rarely justify a claim that a new caption always performs better.
- Outcome and counting rules written down.
- Date range consistent across reports.
- Test traffic identified where possible.
- Bookings and payments checked against actual records.