AI analytics

AI analytics and reporting for email

Predictive send-time, churn-risk scoring and a plain-English weekly report. Since Apple broke the open rate in 2021, we lead with click-to-open rate and revenue per recipient, the numbers that reflect what people actually did.

Good email analytics answers one question: what do we change next, and what will it earn? We get there by reporting metrics that survived Apple's privacy changes and letting a strategist, not a dashboard, make the call. The model finds the pattern; the human decides what to do with it.

Why is open rate unreliable now?

Apple Mail Privacy Protection, launched in 2021 with iOS 15, pre-loads tracking pixels and registers opens that never happened. By August 2021, Apple Mail already accounted for about 49.8% of all opens per Litmus, so roughly half the open figure is machine-generated noise. Reporting open rate as a success metric means optimizing for a number Apple invents on your behalf.

What do we report instead?

Click-to-open rate and revenue per recipient. CTOR measures the share of people who, having seen the email, actually clicked, and it runs around 8.6% all-industry per GetResponse. It is harder to fake because a click is a real action. Revenue per recipient ties every send to money. Those two numbers tell you whether the email worked, which is what the report leads with every week.

How does predictive send-time work?

The model learns each subscriber's engagement pattern and schedules their send for the window they are most likely to act, instead of one fixed time for the whole list. It is a per-person decision made across the list at once. In practice it lifts CTOR by reaching people when they are actually reading, which compounds with better revenue per recipient on the flows that already convert.

What does churn-risk scoring catch?

Subscribers drifting toward inactivity, flagged before they go fully cold. The score watches engagement trend, not a single open, so it spots a regular buyer slowing down while there is still time to re-engage. That feeds the win-back flow and the sunset policy: we mail the at-risk segment differently and retire the truly dead, which keeps the list clean and complaints low.

What does the strategist do that the model does not?

Judgment. The model scores send-time, ranks churn risk and surfaces anomalies in CTOR and revenue per recipient. The strategist decides which test to run next, which segment to cut, and which flow to rebuild, then writes the weekly read in plain English. The model is the instrument; the strategist reads it. This connects directly to lifecycle email automation, since the report tells us which flow to fix. See the method in how it works, the numbers in DTC skincare lifecycle rebuild, and what reporting each plan includes on pricing.

FAQ

AI analytics questions

Why do you ignore open rate?

Apple Mail Privacy Protection, launched in 2021, pre-loads tracking pixels and records opens that never happened. Apple Mail accounted for about 49.8% of all opens by August 2021, so roughly half the open data is fiction. We report click-to-open rate and revenue per recipient instead, which reflect what people actually did.

What is predictive send-time?

A model that learns when each subscriber tends to engage and schedules their send for that window, rather than blasting the whole list at one fixed time. Combined with churn-risk scoring, which flags subscribers drifting toward inactivity, it tells us who to re-engage before a sunset policy removes them.

What does the weekly report actually say?

Plain English, not a dashboard you have to interpret. Each week we report click-to-open rate against the roughly 8.6% all-industry benchmark, revenue per recipient by flow and campaign, and the one change we are testing next. A strategist writes the read; the model surfaces the patterns.

Does a human still make the decisions?

Yes. The model scores send-time, flags churn risk and surfaces anomalies. A strategist decides what to do about them, which test to run, which segment to cut, which flow to rebuild. We treat AI as the instrument and the strategist as the one reading it. That split is where the judgment lives.

Start here

See where your email program is leaking revenue.

Book a strategy call and we’ll run a free audit of your authentication, list health and flow coverage — you keep the findings either way.