Dynamic segmentation turns a segment into a rule instead of a list. The rule says who belongs, for example "top-tier members with no purchase in 45 days." The system checks every member against it again and again. Members move in and out on their own as their data changes. The same idea goes by two other names: real-time segmentation and dynamic customer segmentation.
What is dynamic segmentation?
A dynamic segment is a saved query on member data. Its members are worked out again when the data under it changes, or on a set schedule. Each segment has entry rules and exit rules. A member who meets the entry rules joins. A member who meets the exit rules leaves, and no one has to rebuild a file.
The phrase means something else in map software. There, it names the process of working out where events stored in a table sit on a map (Esri). This page covers the loyalty meaning. For the types of segment and the models behind them, read the customer segmentation guide.
How is a dynamic segment different from a static list?
A static list is a snapshot. It holds the members who matched on the day someone exported it. A dynamic segment holds the members who match right now. The table compares the two on the points a loyalty team has to manage.
| Static list | Dynamic segment | |
|---|---|---|
| How members get in | Exported once from a query, or uploaded as a file | Checked against the saved rules each time |
| When it changes | Only when someone rebuilds it | When member data changes, or on a schedule |
| What it reflects | The day it was built | Each member's state today |
| Best use | A fixed test group or a one-time notice | Offers, journeys and tier nudges that run all year |
| Main risk | It goes stale and keeps members who no longer fit | A small rule edit moves thousands of members at once |
| What to audit | Who built it, from which data, on which day | Which rule version placed each member, and when |
Static lists still have a job. A campaign measured against a fixed group needs that group to stay fixed. Anything that should react to member behavior belongs in a dynamic segment.
What moves a member from one segment to another?
Two kinds of change move members: events and the passing of time. Events are things a member does, or things that happen to the account. Time matters because many useful rules depend on nothing happening at all.
- Purchases. One transaction updates spend, visit count and the last purchase date in a single step.
- Points activity. Earning, redeeming and expiring points change the balance. That drives segments such as "has enough points for a reward."
- Tier changes. A move up or down a tier is an event that other segments listen for.
- Engagement. App logins, offer views and replies to a survey show interest in the gaps between sales.
- Profile and consent. A new address, a birthday or an opt-out changes which segments a member can enter.
- Model scores. A churn or propensity score from a data team moves members when it crosses a set line.
- Time. "No purchase in 45 days" turns true on day 45 with no event at all. So the system has to rerun these rules on a clock as well as when new data lands.
Good segments also set exit rules a little looser than entry rules. For example, a member joins "high spender" at $1,000 of spend and leaves only below $900. That gap stops members near the line from flipping in and out each day.
How does dynamic segmentation drive offers and tiers?
A segment is worth building only if something happens to its members. Offers are the plainest example. The segment becomes the rule for who can get the offer. A member who joins gets it, and a member who leaves stops getting it. The exit matters as much as the entry, because it stops the program from paying a bonus to a member who already came back.
Tiers work the same way. Tier qualification is a dynamic segment with a status attached: cross the line and the member moves up. Programs also build segments around the tiers. "Within 10% of the next tier" gets a nudge to spend a little more. "Set to drop a tier at year end" gets a reminder while there is still time to act.
For illustration, a program with 2 million members targets "top-tier members with no purchase in 45 days." On the day the team exports the list, 30,000 members match. Over the next 14 days, 6,000 of them buy again on their own, and 4,500 more members cross the 45-day line. A static list keeps sending the $10 win-back credit to the 6,000 who came back. If half of them use it, the program gives away $30,000 on visits it was getting anyway. It also misses all 4,500 new lapsers. A dynamic segment drops each returning member at the next purchase and adds each new lapser on day 45.
How do you govern and test dynamic segments?
Because segments update themselves, a mistake updates itself too. A changed threshold or a broken data feed can move thousands of members overnight, and no one sees it. Four habits keep that in check.
- Give each segment an owner and a treatment. A segment with no offer or message attached is a report. Retire it.
- Version the rules. Record who changed each rule and when. Keep the old wording too. Then any member's place in a segment can be explained later.
- Set an order of priority. One member often fits several segments at once. Decide which offer wins, and cap how many messages one member gets in a week.
- Watch segment size. When a segment doubles overnight, check the data feed before you credit a change in behavior.
Testing shows whether a segment earns its cost. Hold back a random share of each segment from the offer and compare the two groups, as in control group testing. Run a new rule against past data before you switch it on, to see how many members it would catch. Then track lift by segment while the campaign runs, so weak offers stop early.
Privacy law applies too. Under the EU's GDPR, profiling means using personal data in an automated way to judge or predict things about a person. Examples include what they like and how they act (GDPR Article 4). People can object at any time to direct marketing that uses their data, and that includes profiling. Once they do, their data can no longer be used for it (GDPR Article 21). So put the consent flag inside the rules of every marketing segment. One objection then takes the member out of all of them at once.
How GRAVTY handles dynamic segmentation
GRAVTY is the Loyalty Juggernaut platform. Segments are written in GRAVTY Visual Rules, a patented no-code rules engine, and they read live member data and model scores. Every offer, reward and message on the platform can target the same definitions. When a member crosses a threshold mid-transaction, the offer they see reflects it in that session.
Scores built by an outside data team slot in as member attributes. So a churn score becomes a segment a campaign can target, with no rebuild. Rules run in a simulation before they go live, and the team ships them without an IT ticket. Each rule keeps its past versions, with one-click rollback. Agentic AI Compass compares results across any segment in plain language. GRAVTY runs 500M+ members in production.
Frequently asked questions
Is dynamic segmentation the same as real-time segmentation?
Nearly: a real-time segment is a dynamic segment that updates the moment new data comes in. Dynamic on its own just means no one has to rebuild the list by hand. A segment that refreshes once a night is dynamic, but it is not real-time.
When should a loyalty program use a static list instead?
Use a static list when the group must stay fixed, such as a test cohort or a one-time legal notice. An event guest list is static too, because the invitation went to named people on a set day.
How often should a dynamic segment refresh?
Refresh a segment at least as often as the action it drives. An in-store offer needs the segment to be current during the sale. A weekly email needs it current on the morning the email goes out.
Do you need machine learning for dynamic segmentation?
No, rules on purchases, points and profile fields are enough to start. A churn or propensity score adds a forward-looking signal once it is stored on the member record, and the segment rules can then read it.
What does dynamic segmentation mean in GIS?
In GIS, dynamic segmentation is the process of computing the map location of events stored in an event table (Esri). It lets several sets of attributes attach to any part of a line feature on a map. It has no link to customer segments.