Lifecycle Segmentation
Lifecycle segmentation groups contacts by the stage they occupy in the customer journey, from anonymous visitor through lead, opportunity, customer, and advocate.
Key takeaways
- Every stage needs a defined entry event, exit event and an owning team.
- The field is single-valued, so a contact occupies exactly one stage at a time.
- Suppression matters as much as targeting: customers should never receive acquisition messages.
- Without demotion rules or a lapsed state, contacts only ever move forward.
- Loop-shaped buying and product-led trials need separate account and contact stages.
In depth
Each contact record carries a single stage field, and the value changes only when a defined event fires. A form submission promotes an anonymous visitor to lead, a qualification threshold promotes lead to marketing qualified, an accepted meeting promotes it to opportunity, a signed contract to customer. The definitions matter more than the labels: every stage needs an entry event, an exit event and an owner. Because the field is single-valued, a contact occupies exactly one stage, which is what makes suppression rules and reporting unambiguous.
Two forces pull the model out of shape. Adding stages increases precision but multiplies the transitions you must automate, and every unautomated transition becomes a manual update that nobody performs. Leaving stages loose has the opposite cost: contacts pile up in the middle, and campaign audiences quietly include people who bought months ago. There is also a direction question. Most implementations only move contacts forward, but real buyers stall and churn, so the model needs demotion rules or a separate lapsed state.
Operationally, the stage decides three things: which message a contact may receive, which team owns them, and whether they count in a funnel report. Suppression is the underrated half; a customer should never receive the acquisition sequence. Quiz funnels use stage as the entry condition rather than the outcome, so a first-time visitor meets an orientation assessment while a trial user meets an activation one, and the completed quiz can itself be the event that advances the record.
Stage models assume a sequence, so they misrepresent buying that loops. A customer who evaluates a second product is simultaneously a customer and an early-stage prospect, and one field cannot hold both. Product-led motions break the model further, because the trial user is using the product while still counted as a lead. In committee purchases the account moves through stages while individual contacts do not, which is why many teams keep a contact-level stage and an account-level stage separately.
Example in practice
How to measure it
Read the model through stage-to-stage conversion and time in stage. Divide contacts entering the next stage by contacts entering the current one to get the transition rate, and track the median days a contact spends before moving. A stage where the rate is very high adds no filtering and probably should be merged; a stage where the median duration keeps growing is where handoffs are failing.
Data hygiene needs its own measurement. Count records that have sat in one stage beyond twice its median duration and express them as a share of the database, because that number reveals how much of your reporting is fiction. Watch it alongside the share of contacts whose stage was last changed manually, since heavy manual editing means the automated transitions are not covering real buyer behaviour.
Common mistakes
The dominant failure is stale stages. Nobody builds the transition that fires when a customer churns or a deal dies, so records sit in opportunity indefinitely, funnel reports show phantom pipeline and nurture emails reach people who left. Write the exit event for every stage at the same time as the entry event, and run a monthly audit listing contacts that have not changed stage within the expected duration for their current one.
The second failure is inventing stages that describe internal process rather than buyer state. Stages named after handoffs or approval steps change meaning whenever the team reorganises, and they cannot be triggered by anything the buyer does. Anchor each stage to an observable buyer action instead. A related error is running lifecycle stage and lead score as competing fields, when the stage should say where someone is and the score how ready they are.
Frequently asked questions
How do I keep lifecycle stages accurate?
Automate stage transitions based on real signals like demo bookings, trial starts, and purchases, and add rules that move stale contacts back or out. Periodic audits prevent reports and automations from drifting.
How many lifecycle stages should we use?
Four to six covers most businesses: visitor, lead, qualified, opportunity, customer and lapsed. Each additional stage must have a transition you can automate and a team that owns it. If a proposed stage cannot be entered or exited by an observable event, it will be maintained by hand and will drift out of date within a quarter.
How is a lifecycle stage different from a lead score?
A stage says where a contact is in the journey; a score says how ready they appear to buy. Stages are discrete, ordered and change on events. Scores are continuous and change on accumulated signals. A contact can hold a high score while sitting in an early stage, which is exactly the combination worth routing to sales first.
Should stages ever move backwards?
Yes. Opportunities die, customers churn and qualified leads go quiet, so define demotion rules or a lapsed state. Without them the model only counts progress and the reporting inflates. Many teams prefer a separate lapsed stage over reversing the sequence, because it preserves the history of how far the contact got before stalling.
Who owns updating lifecycle stages?
The system should, through automated transitions tied to events in the CRM and marketing platform. Assign a human owner per stage for exceptions and for approving the definitions, but any stage that depends on routine manual updates will decay. Reserve manual changes for correcting errors and record who made them so you can spot automation gaps.
How do lifecycle stages interact with quiz funnels?
In both directions. The current stage decides which quiz a contact is offered, so a first-time visitor sees an orientation assessment and a trial user sees an activation one. The completed quiz then acts as a transition event, promoting the contact and passing its score and answers into the routing rule that decides whether sales follows up now.
Do we need separate account and contact stages?
In committee-driven B2B, yes. The account can be an active opportunity while most contacts at it remain untouched leads, and forcing one field to represent both produces misleading counts. Keep a contact stage for messaging and suppression and an account stage for pipeline reporting, with a rule defining how contact activity rolls up to the account.