Lead Cohort
A lead cohort is a group of leads that share a defining characteristic, such as the week they were captured, their quiz score band, or their acquisition source, analyzed together to spot patterns over time.
Key takeaways
- Membership is fixed at capture; the defining trait must never change afterwards.
- Compare cohorts at equal age, not on the same calendar date.
- Widening the period buys statistical stability and costs you timing detail.
- Structured quiz fields let one cohort be cut by tier or answer.
- A young cohort looks worse than it is until the cycle completes.
In depth
A cohort is defined by a trait that cannot change after capture, most often the period in which the lead entered. Membership is fixed at that moment, which is the whole point: the group is then followed forward and each measurement is taken at the same age rather than on the same calendar date. Day thirty for a January cohort and day thirty for a March cohort are comparable, even though the calendar dates are two months apart.
Cohort size and cohort purity trade against each other. Weekly cohorts react quickly to a change but each contains too few leads for a difference to mean anything; monthly or quarterly cohorts are stable but hide what happened inside them. Splitting by a second trait, source as well as period, multiplies the number of groups and shrinks each one. A workable rule is to widen the period until a typical cohort contains enough conversions to notice a change you would act on.
The practical use is before-and-after comparison of a change you made. Ship a new headline, a new offer or a shorter form, then compare the cohort captured afterwards with the one captured before, at equal age. Quiz funnels make this unusually clean because every respondent carries the same structured fields, so a cohort can be cut by score tier, by answer to one question, or by the version of the quiz they saw, without reconciling different form schemas.
Cohorts assume that everything except the defining trait stayed roughly constant, and in marketing it rarely does. A cohort that converts worse may have arrived during a seasonal lull, a competitor's campaign or a pricing change, none of which the grouping controls for. Very recent cohorts are the sharpest trap: they look bad simply because most of their conversions have not happened yet. Never read a cohort younger than a typical sales cycle as a finished result.
Example in practice
How to measure it
Read a cohort as a curve, not a number. For each cohort, plot the share that has reached a given stage at each week of age: contacted, meeting held, closed. The shape tells you two different things, how high the curve eventually goes and how fast it gets there, and a change can improve one while worsening the other.
Alongside the curve, keep each cohort's size and its cost. A cohort that converts at a higher rate but contains a third as many leads may still produce less revenue than a weaker, larger one. Divide revenue attributed to the cohort by what was spent to acquire it, and compare that ratio across cohorts of the same age rather than comparing conversion rates alone.
Common mistakes
The most frequent error is putting a fresh cohort next to a mature one in the same table. The newest column is always lowest, and someone reads that as a decline. Show cohorts as a triangle where each column is an age, leave the incomplete cells blank rather than filling them with partial numbers, and refuse to compare any two cohorts until both have reached the same age.
The second is letting a lead move between cohorts. If the grouping is by source and a lead later re-enters through a different channel, updating their source rewrites history and the old cohort silently changes size. Stamp the cohort key onto the record once, at creation, and never update it. Later channel activity belongs in a separate field, where it can be analysed without disturbing the original grouping.
Frequently asked questions
How is a lead cohort different from a lead segment?
A segment groups leads by static traits like industry or company size at any moment, while a cohort fixes a shared starting point, often a time period, and tracks that group forward. Cohorts are about behavior over time; segments are about who a lead is right now. Many teams use both together.
What can cohort analysis reveal about lead quality?
Cohort analysis shows how different groups of leads convert, retain, or stall over time, exposing which sources and quiz segments produce real revenue. It can reveal that a high-volume channel actually delivers low-quality leads. That insight helps you reallocate budget toward what genuinely works.
Why are quiz funnels good for building cohorts?
A scorecard quiz captures consistent, structured data on every respondent, including score, segment, and timestamp, so cohorts are clean and comparable. There is no patchy or missing data to normalize first. That makes it easy to compare, for example, hot-tier leads across consecutive weeks.
What is the difference between a lead cohort and a lead segment?
A cohort has fixed membership and is followed over time; a segment is a live filter whose membership changes as records change. Everyone captured in March stays in the March cohort forever, while a segment of leads with score above eighty gains and loses members daily. Use cohorts to judge whether something changed, segments to decide who to contact today.
How big does a lead cohort need to be?
Big enough that the number of conversions, not leads, is stable. If a cohort of two hundred leads produces four closed deals, one extra deal shifts the rate by a quarter and nothing you read is reliable. Widen the period, or measure an earlier stage such as replies or meetings booked, where counts are larger and move sooner.
Should cohorts be weekly, monthly or quarterly?
Match the period to your sales cycle and volume. If deals close in weeks and you capture hundreds of leads a month, weekly cohorts are readable. If the cycle runs several months, weekly groups will still be half-empty when you want an answer, so monthly or quarterly is more honest. Many teams keep weekly cohorts for early-stage signals and monthly for revenue.
Can I build cohorts on something other than capture date?
Yes, as long as the trait is fixed at capture and observable for every lead. Score tier, quiz version, campaign, first landing page and company size band all work. What does not work is anything that changes later, such as lifecycle stage or current owner, because the cohort would then be defined by an outcome you are trying to measure.
How long should I track a lead cohort?
Until the conversion curve flattens, which is roughly one and a half times your typical sales cycle. Stopping earlier undercounts slow deals and makes long-cycle sources look worse than fast ones. Keep the cohort open in the report after that, but stop treating late additions as meaningful; by then the differences you see are noise rather than the effect you tested.
Why does every cohort look worse than the one before it?
Usually because the newer cohorts are younger, not worse. In a table sorted by date, recent rows have had less time to convert and always sit lower. Check by comparing each cohort at a fixed age instead of at today's date. If the gap survives that correction, look for a real cause such as a change in traffic mix or offer.