Marketing pressure: halve your unsubscribe rate without giving up revenue

With Batch AI Predict’s propensity score, you can halve your unsubscribes while keeping 95% of your campaign revenue, by targeting only the contacts who genuinely care about the theme. The message gets diluted, and your least relevant contacts grow tired. If that situation sounds familiar, this use case is for you.

It’s one of the most common use cases for predictive AI in CRM, part of a series breaking down, one by one, the concrete uses of Batch AI Predict. To understand why a predictive score only creates value once it’s activated, start with our framing article.

Why does sending to everyone dilute the message and tire the base?

Unable to pinpoint the customers genuinely interested in a theme, the reflex is to send wide. The “home linen” campaign goes to the whole base, so does the “summer collection.” It feels reassuring: you maximize reach. But it’s costly in the long run.

Every irrelevant send tires the contact it doesn’t concern a little more. Over time, the least interested unsubscribe, mark as spam, or simply stop opening. Overall deliverability degrades, which penalizes even the campaigns aimed at the right customers. You favored volume; you lose relevance and base health.

Let’s ask the question that matters right away: how do you know the problem is solved? Here, success isn’t measured in revenue but in base preservation.

• The primary metric is the unsubscribe rate (the direct marker of fatigue) with deliverability as the underlying stake.

• Watch closely alongside it: the click rate and the share of maximum revenue you still capture despite a smaller audience.

We’ll come back to it at the end of this article.

How does Batch AI Predict reduce pressure without giving up revenue?

With Batch AI Predict, you flip the “everyone” logic right inside the platform. The approach is simple:

• For each customer, the theme propensity score sits on the profile.

• Rather than sending to the whole base, you set a threshold and include in the campaign only the customers above it.

• The audience narrows to the most relevant profiles, and unconcerned contacts are spared.

The gesture is familiar to a CRM team: you filter on an attribute, exactly as you’d filter on a purchase category or a segment. Except here, the filter rests on a real probability of interest, not an approximate rule. The result: a message that reaches those it concerns, and a base you preserve.

How does the propensity score identify who’s genuinely receptive to a theme?

The score powering this use case is the product propensity score. For each customer, it predicts the probability they’ll buy within a given product target over the next X days.

That probability, between 0 and 1, is computed by a Machine Learning model trained on all the data available about customers:

• behavioral data (purchases, browsing, engagement: opens, clicks, unsubscribes)

• product catalog

• socio-demographic data

• and more

That richness is what allows fine estimation of who is receptive to a theme, well beyond what a history-based rule could isolate.

Worth noting: this same propensity score also serves to maximize a themed campaign’s performance, by building the highest-performing audience rather than narrowing pressure. One score, two complementary use cases: optimize performance or preserve engagement, depending on your priority.

What results can you expect from propensity-based targeting?

The unsubscribe rate can be cut in half, for instance moving from around 0.44% to 0.22%. The click rate, meanwhile, improves markedly, up to a fourfold increase, since the message now reaches only concerned contacts.

And counterintuitively, narrowing the audience doesn’t sacrifice revenue: by targeting only about the top 30% most-appetent customers, you capture around 95% of the campaign’s theoretical maximum revenue.

These figures answer the starting problem point by point:

• The halved unsubscribe rate proves you’ve stopped the fatigue.

• The higher click rate shows the message is relevant again.

• And the 95% of revenue captured on a fraction of the base demonstrates the key point: you protect engagement and deliverability without giving up revenue.

In short: what does propensity-based targeting deliver for marketing pressure?

Sending themed campaigns to the whole base seems to maximize reach, but it dilutes the message and tires the least relevant contacts. Batch AI Predict’s product propensity score lets you narrow the audience to genuinely interested profiles, right inside the platform. Fewer unsubscribes, more clicks, and most of the revenue preserved.

To activate it on your base, talk to a Batch expert.