Cohorts: reading accounts by entry month

Q1 is in the books, and the quarterly rollup says the program is fine: activity on target, conversations steady, meetings roughly where the plan put them. Now run one more report before you file the quarter away. Take every account that entered the funnel since January and split the results not by when the meetings happened but by the month the account first got touched. January entries in one column, February in the next, March in the third. In a surprising number of programs, that one pivot tells a story the totals had completely averaged away: the January cohort converted at twice the rate of the March cohort, and the quarter was fine only because early strength subsidized late weakness.

Why entry month is the honest axis

A monthly report answers what happened in March. A cohort report answers what became of the accounts you started working in March, however long that took. The difference matters because outbound results lag their causes: a meeting held in March might trace to a first dial in January, so March's healthy meeting count can be entirely the harvest of January's decisions while March's own plantings quietly fail. Totals mix vintages. Cohorts keep each vintage in its own bottle, so you can taste the one that went off.

What a weakening cohort is telling you

When later cohorts underperform earlier ones, the causes are usually one of three, and the cohort view is often the only place any of them show up early.

• List decay. The January list was built from the freshest, best-fit accounts, and each later batch reached deeper into the barrel. Conversion falls cohort by cohort even though reps, message, and effort never changed. This is the strongest argument that list-building is a continuous job rather than a January event.
• Message drift. The pitch that worked in January mutated by March, one small field edit at a time, and nobody noticed because no single week looked different. A cohort falloff that coincides with rising activity is frequently drift, and it is exactly the failure the 90-day message review exists to catch from the other direction.
• Seasonal texture. Sometimes the cohort is innocent: accounts first touched in the closing weeks of a buyer's quarter start slower than accounts touched at its open, a rhythm we saw from the buyer's side in the Q2 budget review. You only learn your market's texture by keeping cohort records long enough to see the pattern repeat, which is one more reason to start the discipline now rather than in a crisis.

Keeping it light enough to survive

Cohort reporting has a reputation for dashboard bloat it does not deserve. The working version is one small table: entry month down the side, the funnel stages across the top, refreshed monthly, with each cohort's rates filled in as they mature. Ten minutes a month once the habit exists. It complements rather than replaces the operating numbers you watch weekly; the weekly view steers the week, and the cohort view audits the quarter, in the same spirit as the two dashboards. The trap to avoid is judging young cohorts too fast: a March cohort is barely five weeks old in early April, and for most cycles its meetings simply have not happened yet. Compare cohorts at equal ages, or you will invent a collapse that is really just a calendar.

We keep entry-month cohorts on every CommandVA account, which is how a client learns whether a soft month means a soft market or a stale list, and how we know when a message needs surgery before the totals ever sag. If your Q1 rollup looks fine and you have never seen your numbers split by vintage, book a strategy call. Bring the raw funnel export, and we will build the first cohort table with you.