BakingAdsA bad day
A holiday, a payment outage or a tracking incident hit every user at once.
Tell Me Why
When a number moves, know why.
Every daily total mixes users acquired today, last week and months ago. Tell Me Why decomposes each day’s result by cohort age, so you can tell a calendar event from a change in the quality of the users you are buying.
The problem with daily totals
Revenue dropped 18% yesterday. Four very different stories fit.
A single line on a chart cannot say which one is true. Each story calls for a different action.
Fewer new users
Acquisition volume fell, so there was less first-day revenue to count.
Weaker new cohorts
Volume held, but recently acquired users monetize less than before.
Older cohorts fading
Mature users are spending less, while new ones behave as usual.
The period × age matrix
One grid. Every day, split by how old its users are.
Rows are calendar days. Columns are the age of the users generating the result, in days since acquisition. Each cell is the contribution of users of that age on that date, and every row adds up to the daily total you already know.
- Metrics: counts, revenue, ARPU, LTV and ROI.
- Built on your events: the same acquisition and revenue events used across BakingAds.
- Saved configurations to monitor the same decomposition over time.
| Date ↓ · Age → | D0 | D1 | D2 | D3 | D4–7 | D8+ | Total |
|---|---|---|---|---|---|---|---|
| Mon 14 | 1.9k | 0.9k | 0.6k | 0.5k | 0.8k | 0.9k | 5.6k |
| Tue 15 | 2.0k | 0.9k | 0.6k | 0.5k | 0.8k | 0.9k | 5.7k |
| Wed 16 | 0.7k | 0.4k | 0.3k | 0.2k | 0.3k | 0.4k | 2.3k |
| Thu 17 | 1.8k | 0.9k | 0.6k | 0.5k | 0.8k | 0.9k | 5.5k |
| Fri 18 | 2.2k | 1.0k | 0.6k | 0.5k | 0.8k | 0.9k | 6.0k |
Wednesday drops at every age. The cause sits on the calendar, not in the traffic you bought.
How to read it
Three shapes. Three different causes.
Because a cohort acquired on a given day moves one column to the right every day it ages, the direction of a pattern in the grid points straight to its cause.
A calendar event
Every age moves together on the same date. Think holidays, outages, tracking gaps or a platform-wide change. Fix the incident; don’t cut the traffic.
One cohort is different
The same group of users stands out day after day as it ages. A specific campaign, source or creative brought in stronger or weaker users.
A change at a given age
From one day onward, users of the same age behave differently, for example all new users on their first day. Look at onboarding, pricing or offer changes released that day.
Decomposition over time
See the total, and what it is made of.
Stack each day’s result by cohort age and the story behind the curve appears. A one-day dip that hits every layer is a calendar event. A thinning base layer that persists means new cohorts are monetizing less, even while the total still looks acceptable.
D0Acquired that dayD1–7First weekD8–30First monthD31+Mature users
Last 14 daysD0D1–7D8–30D31+
How it works
From alert to answer in four steps.
Choose the metric
Counts, revenue, ARPU, LTV or ROI, built on the acquisition and revenue events you defined.
Pick the period
Select the calendar range around the change you want to explain.
Read the shape
Row, diagonal or column: the pattern in the grid points to the cause.
Save and monitor
Keep the configuration and reopen it whenever the same question comes back.
Use cases
Questions Tell Me Why answers.
“Revenue fell in the weekly report.”
Check whether the drop is concentrated on specific dates or spread across the youngest users before touching budgets.
“We scaled a campaign. Is it still worth it?”
Scaling can bring in lower-value users. Watch whether the youngest ages keep their contribution as volume grows.
“Did tracking break?”
A gap that hits every age on the same day points to collection, not to customers. Verify events before drawing conclusions.
“We changed pricing on the 10th.”
Compare the same ages before and after the release date to see who reacted, and how.
Part of the platform
The last step of the investigation.
Views finds the segment, Cohort shows how it matures, the ARPU forecast estimates what it will be worth. Tell Me Why explains why a number moved.
The next decision starts here
Stop guessing why the numbers moved.
See every daily result split by cohort age, and act on the real cause.