Ask most association managers which members are at risk of lapsing and you get an honest answer: the ones who do not renew. That is not a prediction, it is a post-mortem.
The alternative usually gets pitched as a big analytics project. Dashboards, data warehouses, maybe a consultant. It stalls, because nobody has the budget or the appetite.
There is a middle path. You can build a useful engagement score from data your organisation already collects, in an afternoon, on a spreadsheet if you have to. It will not be perfect. It will be dramatically better than guessing.
Pick five signals you already have
An engagement score is just a way of turning scattered activity into one comparable number. Do not overthink the inputs. Most associations already capture everything they need:
Event attendance. Did they attend anything in the last twelve months, and how recently?
Portal or website logins. Have they signed in this quarter?
Email engagement. Have they opened or clicked anything in the last 90 days?
Volunteer or committee involvement. Are they on anything, in any capacity?
Payment history. Did they renew on time last cycle, or did it take three chases?
That is it. Five signals, all of which sit in systems you are already paying for.
Weight them by what actually predicts renewal
Not all signals are equal, and the differences matter. In member organisations the pattern is fairly consistent.
Anything involving a human commitment predicts renewal far more strongly than anything passive. A member who volunteered on a committee is enormously more likely to renew than one who opened three newsletters. Attending an event in person beats attending a webinar, which beats reading an email.
So weight accordingly. Something like: committee or volunteer role, 40 points. Attended an event in the last year, 25. Logged in during the last quarter, 15. Engaged with email in the last 90 days, 10. Renewed on time without chasing, 10.
The exact numbers matter less than the ordering. Give the deep signals most of the weight and the shallow ones very little. Email opens in particular are noisy and increasingly unreliable thanks to privacy features, so do not let them drive the score.
Use recency, not just totals
A member who attended four events three years ago and nothing since is not an engaged member, but a lifetime-total score will say they are. Cap every signal to a recent window, twelve months for events, ninety days for digital activity. You are measuring their current relationship with you, not their history.
This one change is usually what separates a score that finds real risk from one that just reflects tenure.
Look at the trend, not the number
The score’s absolute value is less interesting than which way it is moving.
A member sitting at 30 out of 100 who has been at 30 for three years is a low-engagement member who nonetheless keeps paying. There is a decent chance they value the directory listing, the accreditation, or the sector advocacy, and they will renew quietly forever. Chasing them is often wasted effort.
A member who was at 75 last year and is at 35 now is the one to call. Something changed, and they have not left yet. That drop is the most actionable signal in your entire dataset, and it is invisible to anyone only watching renewal dates.
Run the score quarterly and flag anyone who has fallen more than 20 points. That list is usually short enough for a real person to work through with a real phone call.
Do something with it
The score is worthless if it only lives in a report. Tie it to specific actions:
Big drop, high value member: a call from a staff member, not an email.
Consistently low, never attended anything: a targeted invite to one thing you think they would genuinely value, based on their sector.
Consistently high: ask them to speak, mentor, or join a committee. Your most engaged members are your recruitment pipeline for volunteers, and they are usually flattered to be asked.
New member below the average for their cohort at 90 days: your onboarding did not land. Intervene now, not at renewal.
Accept that it will be roughly right
Some members will score badly and renew happily. Some will score well and leave anyway because they changed jobs. That is fine. A score that is directionally right on most members beats a renewal report that is precisely right on all of them, too late to do anything about it.
If your member activity is spread across four systems and a spreadsheet, see how My Member Buddy brings engagement data into one place.