Most CRM and ESP setups share a quiet assumption that's costing brands more than they realise. If a customer hasn't purchased in 90, 180, or 365 days, they belong in the same bucket. Same segment. Same re-engagement flow. Same message.

The problem isn't the logic, it's that two customers who last bought 180 days ago could have almost nothing else in common. Treating them identically almost guarantees that at least one of them receives something entirely irrelevant.

74% of people dislike being shown irrelevant content, making it one of the biggest drivers of email list decay and unsubscribes.

The bucket problem

Most platforms weren't built to work any other way. Recency-based segmentation e.g. 90, 180, 365 days is the industry default because it's manageable and easy to set up. However, it flattens real behavioural difference into a single, blunt category.

Consider what lapsed 180 days might actually contain: a high-value seasonal buyer who's due to return, a repeat customer who's been browsing regularly but hasn't converted again, someone who bought once and has disengaged entirely.

One re-engagement email goes to all three. For one it's perfectly timed. For another it misses completely. For the third it's pure noise and an unsubscribe risk with no upside.

And the inbox is getting less forgiving

Email and SMS remain the most valuable owned channels a brand has and the only channels where you can reach a customer directly and without paying for the privilege (somewhat).

When irrelevant communications erode a subscriber's trust, they rarely tell you directly. They just quietly leave, and take their future purchase value with them. The damage isn't always visible in a single campaign report. It compounds over time.

What individual-level behaviour actually looks like

At Noa, we work differently. Rather than grouping customers by how long it's been since they last bought, we look at the full picture of how each person has been engaging: browse patterns, category affinity, response to past communications, purchase cadence and signals that indicate intent even when a transaction hasn't happened yet.

A customer who last bought 180 days ago but has browsed three times in the last month gets treated very differently to one who has been entirely silent. Because they are different. One is close. One needs a different conversation entirely or perhaps none at all right now.

The goal isn't to send more. It's to send what's right, to the right person, at the moment it's useful to them.

The channel is worth protecting

Email delivers an average return of around $55–$65 for every $1 spent, still one of the strongest ROI figures in marketing. That number assumes the channel itself remains trusted and healthy.

An inbox full of irrelevant messages is one people stop opening. An audience that stops opening eventually stops being an audience at all.

The brands that will protect that channel long-term are the ones investing now in understanding customers as individuals, not as cohorts defined by a timestamp. The infrastructure to do this at scale, without the manual overhead, is exactly what most teams have been missing. That's the gap we're closing.