Product-lag RCA

The question, verbatim: "Shilajit is lagging 30 behind. What are we doing? And why has this happened in the first place? Last time we saw it was 10. Simple question. Answer is — winners didn't came this time with their ads. The win rate is bad. Or creators didn't get produced this time because the agency delayed."

This is the "product is behind target" RCA. Every product, every week, every RCA call. The dominant answers (per the team's experience) are predictable enough that the platform can hit ~90% of them with a small set of branches.

Who asks

• Founder / function head (shantanu in the source meeting)
• Perf-marketing leader, defending the number
• Category lead, defending the input pipeline

Frequency

Weekly per product. Daily during sale weeks.

Data you need

• targets — monthly / weekly target per product
• orders and customers — actuals, with new-vs-repeat tag
• ads and ad_creatives — the active creative set per product
• creator_pipeline — planned vs delivered creative count per period (the upstream supply)
• Meta platform health signals (algorithm changes, account-level issues)

How to ask it

• "Shilajit is -30% vs target this week. RCA the gap. Decompose into ad winners, creator supply, and platform issues."
• "For each product behind target by >15% this week, give me the dominant cause."
• "Has the Shilajit ad-winner rate dropped this week vs the trailing 4 weeks? If yes, by how much, and which creators stopped winning?"

What you'll get back

A focused, structured RCA — not a chart dump:

Headline: Shilajit -30% vs ₹X target. Primary cause: ad win-rate dropped from 28% → 11% week-over-week.

Top 3 candidate causes, ranked:

1. Ad-winner shortage (most likely, 72% of gap) — only 2 of last week's 18 active ads cleared the NRoAS threshold, vs 7 of 18 the prior week. The 5 ads that flipped from winners to losers all came from the same creator batch produced 6 weeks ago — decay-period exhaustion.
2. Creative supply lag (18% of gap) — agency delivered 4 new creatives this week vs the planned 10. The gap left under-fed ad sets across BAU.
3. Platform health (residual) — no algorithm-level signal detected this week.

Recommended actions:

• Pause the 5 decayed ads (linked to Ad action decisions)
• Escalate the creator-supply gap to category lead with the missed-count chart
• Re-run RCA mid-next-week to confirm trajectory

The logic tree

This is the 70%-coverage tree the team validated. The platform extends it over time.

Product behind target by >X%?
├── Is the ad-winner rate down?
│   ├── Yes (>20% drop) → primary cause: winners. Surface decayed ads.
│   └── No → continue
├── Is creative supply behind plan?
│   ├── Yes (>30% gap) → primary cause: agency / category. Escalate.
│   └── No → continue
├── Did Meta have an algorithm event in the period?
│   ├── Yes → noise; downweight the WoW comparison
│   └── No → continue
├── Did the prior week's winners get over-scaled (saturating audiences)?
│   ├── Yes → expected decay; correction is normal
│   └── No → continue
├── Did landing-page CR for the product drop?
│   ├── Yes → not perf's call; route to product / web team
│   └── No → continue
└── Is the target itself ambitious vs trailing-month run rate?
    └── If yes → calibrate target conversation, not RCA

How to make it recurring

Save as a tracker with:

• Cadence: weekly, Monday morning (before the marketing RCA call)
• Scope: every product, separately. Use a tracker template applied per product.
• Threshold: alert if any product is below target by > 15%
• Linked decisions: link to past lag-RCA decisions on the same product, so the platform surfaces "we said it was creator supply last time too — has that been fixed?"

Pitfalls

• Conflating WoW noise with trend. A single -30% week on Shilajit might be a one-week miss, not a structural problem. Always read the trailing-4-weeks line, not just this week.
• Blaming the agency for everything. Half the time the creative-supply branch is real; half the time it's a convenient story. The platform's job is to keep the ranking honest.
• Ignoring the target conversation. If a product missed target every week for 8 weeks, the target is wrong — escalate that, not the operational pieces.

Related

• NRoAS drop RCA — the channel-level cousin
• Ad action decisions
• Creator pattern trends
• Decisions — log the RCA outcome as a decision