Measuring Results
Comment automation helps ads in two ways: buyers get answers while they’re still on the ad, and the thread under the ad stays clean. This page shows how to measure both on your own account.
Why it can move ROAS
- Buyers ask before they buy. In RYPL’s Q3 2026 data, 40% of analyzed comments on advertisers’ ads and posts were questions and 38% showed buying intent. An answer within a minute catches them; one the next morning often doesn’t.
- The thread is part of the ad. Meta’s quality ranking draws on feedback such as people hiding or reporting the ad, and engagement rate ranking counts comments. Spam and scam claims give viewers reasons to hide the ad; useful answers keep them engaging.
Published results: in a controlled A/B test by BrandBastion at MindValley, comment management alone delivered +48% ROAS, the upper end of published results. Across RYPL accounts, brands saw about +10% ROAS in the first month (RYPL’s own measurement). Your result depends on your comments and how good your knowledge base is, so measure it.
Before you switch on
Write down, for the ads you’ll compare:
- ROAS, cost per result and conversion rate for the last 2 to 4 weeks;
- quality ranking and engagement rate ranking in Ads Manager;
- how long questions waited for a reply (check the last 50);
- how many spam or scam comments sat under the ad, and for how long.
Weeks 1 to 4
Week 1: get the replies right
Read every reply and every hidden comment. Fill knowledge base gaps. Don’t judge ROAS yet.
Week 2: tune
Run Suggest improvements, add rules for repeated questions, and check what your rules flagged in Review first.
Weeks 3 and 4: compare
Compare the same ads (same product, targeting and creative) with your baseline. If you can, run two similar ad sets at the same time with Rypl on only one channel.
What to compare
| Metric | Where |
|---|---|
| ROAS, cost per result, conversion rate | Ads Manager |
| Quality ranking, engagement rate ranking | Ads Manager → ad relevance diagnostics |
| Reply time to questions | Rypl inbox |
| Spam and scam comments hidden, and hides you reversed | Rypl inbox → All hidden or deleted |
| Questions left unanswered | Rypl inbox, and Suggest improvements |
| Hours your team spends on comments | Your team |
Reading the result
- Reply time fell but ROAS didn’t move: check reply quality. Fast, vague answers don’t sell.
- ROAS rose on some ads only: those are usually the ads with the most questions.
- Many hides reversed: a banned word or rule is too broad.
- Questions still unanswered: the knowledge base is missing facts.
More on rypl.is
- Meta ads ROAS in 2026: what’s realistic by industry, with a calculator
- Why comment response time affects ad performance: the case for answering fast
- Facebook comment engagement rates 2026: what’s normal by industry
- Ad comment statistics 2026: what people write under ads, RYPL data for Q3 2026
| Back to Best Practices | Back to Documentation | Visit Rypl |