Deliverability Case Study: "Segment the Groove"
Track 07 of Click Through is the album's pivot. The domain is rebuilt. The list is still forty thousand names, most of which have never meaningfully engaged with anything Marcus sent. Track 06 gave Marcus the insight — engagement is the signal, not the tab. Track 07 is where he operationalizes it: sort the list by who actually moved, not by who nominally subscribed.
The good practice here is the most important shift in the album's second half: Marcus stops using opens as the primary engagement signal and starts using first-party behavioral data that cannot be faked.
Verse 1: The Problem with Open Rate as a Segmentation Signal
"Not the ones who clicked — the others never showed me"
Before Marcus can build better segments, he has to stop using the signal that Track 06 — and Apple Mail Privacy Protection — revealed as unreliable. Open rate was the default segmentation signal for email marketers for two decades. It is no longer trustworthy.
In September 2021, Apple rolled out Mail Privacy Protection (MPP) for iOS 15. MPP pre-downloads email content — including tracking pixels — before the user reads the message, in a proxy environment that hides the user's actual IP address. The result: every email sent to an Apple Mail user on iOS or macOS registers as "opened" in ESP reporting, regardless of whether the user ever saw it.
Across most ESPs, Apple Mail users now represent 30–50% of a typical consumer email list. A list where half the "openers" are actually machine-pre-fetched pixels is a list where open rate is measuring Apple's infrastructure, not subscriber intent.
First-Party Behavioral Signals: What Actually Measures Intent
Marcus's new segmentation framework — clicks, site visits, logins, purchases — is the correct response to the MPP problem:
- Clicks: A click requires a human action. Even with Apple's Link Tracking Protection (which adds some obfuscation to link parameters), a click-through to your site is a significantly stronger intent signal than an open. Clicks cannot be faked by pre-fetch proxies. A subscriber who clicks is telling you something real.
- Site visits: If your ESP integrates with your website analytics (or you use server-side tracking), a site visit after a send — especially a visit to a product page or pricing page — is a confirmed engagement signal. The email opened a real browser session.
- Account logins: For SaaS products, e-commerce accounts, or any platform with an authenticated session, a login triggered after a campaign is the strongest possible engagement signal. The subscriber received the mail, acted on it, and engaged with the product.
- Purchases: The most direct revenue signal. A subscriber who purchased after a campaign is your highest-value segment for nearly every future send.
Bridge: The Contrast with the Old Approach
"I blasted forty thousand — never asked who cared / The list was big — the groove was never there"
The bridge is a clean articulation of the volume trap's core error: confusing list size with audience. A list of forty thousand passive subscribers produces lower engagement rates than a list of one thousand active ones. The signal — what ISPs are watching, what drives tab placement, what earns sender reputation — is the engagement rate, not the send count. A smaller, more engaged list outperforms a larger, passive one on every metric that matters for deliverability and revenue.
Segmentation built on behavioral signals rather than demographic data or subscription date produces cleaner sends, better deliverability, and higher revenue per campaign. The shift from open-rate-based segmentation to click-and-behavior-based segmentation is the most important tactical update email marketers can make in the post-MPP environment.
Building Behavioral Segments
The starting framework for behavioral segmentation uses four signal tiers, ordered by engagement strength:
- Purchasers (30-day window): Subscribers who completed a purchase after a send in the last 30 days. Highest-value segment. Send product-specific follow-ups, cross-sells, and loyalty content here.
- Clickers (60-day window): Subscribers who clicked a link in any campaign in the last 60 days. Strong intent signal. Most of your nurture and promotional content should go here first.
- Site visitors (90-day window): Subscribers who visited your site after a send in the last 90 days (requires ESP-to-analytics integration or server-side tracking). Lower than clickers but still behavioral — they acted on something.
- Openers only (90-day window, use cautiously): Subscribers who registered an open but no click or site visit. Post-MPP, this segment is unreliable for iOS/macOS users. Treat it as a low-confidence signal and do not use it as the primary engagement filter for high-volume campaigns.
What "Send to Who Moved" Means in Practice
"Send to who moved" is a shorthand for a single question before every campaign: does every address on this list have a behavioral reason to be here?
- Before each send, filter your list to subscribers who have clicked, visited, or purchased within the last 90 days. Everything outside that window is a re-engagement candidate or a suppression candidate — not a primary campaign recipient.
- For each campaign, match the content to the signal that earned the recipient's place on the list. A subscriber who clicked a product link belongs in a segment that gets product content, not a broadcast newsletter.
- Rebuild your segments from behavioral data after every campaign, not just at scheduled intervals. A subscriber who clicked last week moved; a subscriber who last clicked eight months ago did not.
Moving Away from Open-Rate-Based Decisions
Open rate is still a useful directional metric, but it should not drive segmentation, sunset decisions, or engagement scoring in the post-MPP environment:
- Replace "opened in the last 90 days" with "clicked in the last 90 days" as your primary engagement filter.
- Use click rate, click-to-open rate (CTOR), and reply rate as engagement health metrics instead of raw open rate.
- For re-engagement campaigns, trigger on absence of clicks (not absence of opens) over a 60–90 day window.
Educational content. Email deliverability evolves rapidly. Platform rules (Gmail, Yahoo, etc.), engagement signals, and ESP behaviours change frequently, and real-world issues often involve conflicting signals, data quality problems, and failure modes that general best practices can’t anticipate. Content on this site is provided for informational purposes only and does not replace a thorough analysis by a qualified deliverability professional.
Terms of Use 