Signups in a key market dropped 12% week-over-week

Diagnose a problemMedium

Problem. In one large country, new paid signups fell 12% week-over-week, while engagement of existing members looks normal. What happened, and what should we do?

Before you reveal: say your answer out loud, as if you were in the real interview — get your reasoning across clearly first. There is no single correct answer: reading what the interviewer is really after and defending your own thinking is what makes an answer strong.

Clarify the question first

First make sure you understand the prompt itself before solving:

"How are we counting signups — gross new paid signups, or net adds after cancels?"
"Is the 12% week-over-week this week versus last week, and versus a normal week or a promo-inflated one?"
"By one country do you mean a single market, across all plans and platforms?"

Lay out your approach

Rule out a measurement artifact first. A logging/attribution change, a billing-cycle timing shift, or a country-mix change can fake a 12% drop before any real demand moves.
State a hypothesis before slicing. With engagement normal, my lead guess is a leak at one funnel stage — most likely payment — not broad demand. Name the cut that would confirm it.
Decompose the funnel. Visits → plan-select → payment-auth → activated. Find which stage moved before guessing why.
Internal vs external. Internal / fixable: a price change, a checkout or payment regression, a dropped payment method, an app-release bug. External: a competitor, seasonality, a marketing cut, an FX shock.
Attribute and size it. Once you find the stage, attribute the points across causes so they add up, and pick the cheapest decisive check.

Then check in: "Does that plan sound right, or should I start somewhere specific?"

Simulate the back-and-forth

Interviewer

Okay, where would you start?

You

First I'd sanity-check it's real — could a logging or attribution change have faked the 12%? Then, since engagement looks normal, my hunch is a payment problem, not demand. I'd decompose the signup funnel to see which stage moved. Do we know if it's broad or at one step?

Interviewer

Say it's all at the payment step — plan-selection is flat, but fewer people complete payment. What does that tell you?

You

Then it's not demand — people are trying to pay and failing, so it's a checkout or payment-method problem. I'd line it up against anything that shipped recently and split by platform and payment method to localize it.

Interviewer

Suppose it's concentrated on iOS, starting a few days ago. How do you read that?

You

That points to a release regression — most likely the iOS billing flow that shipped around then. I'd size it: if iOS is a big share of that market, a sharp drop there alone could explain most of the 12 points. I'd also check the decline-reason mix in case a card processor is adding to it.

Interviewer

How do you know it's real lost revenue and not just broken tracking? What do you think?

You

I'd compare client logs to server-side authorization records. If the client shows the drop but server-side auths are flat, it's a tracking bug, not lost signups — a totally different fix. If both moved together on iOS from the release date, it's real.

Land a clear conclusion

You

So my read: it's a self-inflicted, reversible checkout regression on iOS, lined up with a recent billing release, not falling demand. I'd roll back or hotfix the release, expect most of the drop to come back within a day or two, and validate on the iOS auth rate — rather than reach for a promo, which would only mask it. The one thing that flips me is if server-side auths never moved, in which case I'm fixing the logging pipeline, not checkout.

Key summary — what really matters.

Rule out a metric/logging artifact before trusting the 12%.
Hypothesis before slicing — normal engagement plus a sharp drop smells like payment, not demand.
Decompose the funnel to find the stage that moved, then attribute the points so they add up.
Separate self-inflicted from market causes — fix a regression, don't paper over it with a promo.
End with a decision and the check that would flip it.

Run the investigation (illustrative numbers).

Decompose first — which stage actually moved?

Funnel step	Last week	This week	Δ
Visit → Plan-select	41%	41%	flat
Plan-select → Payment-auth	78%	69%	−9 pts
Auth → Activated	95%	95%	flat

Only payment authorization moved. That alone kills the "demand is down" story: visits and plan-selection are unchanged, so people are trying to pay and failing.

Attribute the 12 points — don't stop at the first story.

Suspected cause	Points of the 12%	The cut that shows it
iOS in-app billing SDK update (shipped Tue)	~7	drop is iOS-only and starts on the release date
Card-network decline spike (one LATAM processor)	~3	decline-reason mix, concentrated in one method
Organic seasonality	~2	matches the same week last year

The three add to 12 — a complete account, not a single guess that only covers 3 of the points.

iOSWeb

Illustrative: daily payment-auth rate. iOS steps down on the SDK release date; web stays flat — a release regression, not a gradual demand decline.

Real vs instrumentation. Compare client logs to server-side authorization records. If the drop shows in client logs but server-side auths are flat, it's a tracking bug, not lost revenue — a completely different fix.

Distribution check. Is the drop broad, or concentrated in a few high-volume devices/affiliates? A power-law concentration points to one broken integration rather than a market-wide shift.

Recommend. "The drop is concentrated in payment-auth on iOS, consistent with the Tuesday SDK update (~7 of the 12 points); the decisive check is the iOS auth rate against the release. Roll back / hotfix and expect ~7 points back within 48 hours. It is self-inflicted and reversible, so ship the fix now — not a promo, which would only mask it."

Dimension	What interviewers look for	Below the bar ❌	Above the bar ✅
Problem framing	Whether you take a messy business question and break it into clear parts, and say what you think is going on before you dive in.	You jump straight to a metric or a model without understanding the goal, and skip clarifying questions.	You clarify the goal and scope first, lay out a simple structure, and state a clear guess to test.
Methodology and math	Whether your methods, metrics, and numbers are sound, and you can say why you chose them.	You use a heavy method without saying why, make basic math or estimate errors, or forget a control group.	You pick an approach that fits the problem, define the metrics you will judge it on, and handle the obvious edge cases.
Business judgment	Whether you understand why the question matters and tie your answer to real outcomes.	Your idea is technically fine but ignores cost, what is feasible, or what the company actually cares about.	You connect your work to real levers like revenue, growth, or retention, and you name the main risks.
Deployment thinking	Whether you think about how the solution actually runs and lasts in the real world, not just on paper.	You treat it like a class exercise and never say how it would work once it is live.	You think about data quality, scale, monitoring, what happens when it fails, and how you keep it working over time.
Communication	Whether your thinking is easy to follow and you treat the interview as a conversation.	You ramble, lose the thread, or talk at the interviewer instead of with them.	You restate the problem, walk through your logic in order, check in, and sum up your main point at the end.

Interview tips

Strong-answer signal: stating a hypothesis before segmenting, then attributing the points (7 + 3 + 2 = 12) instead of stopping at the first plausible cause; separating self-inflicted/fixable from market causes. Common trap: jumping to "demand is down" and proposing a promo, when the real issue is a payment regression a promo would only mask. Trains: hypothesis-before-segment; funnel-attribution.