Segmentation gets sold as a way to lift open rates. That is the least interesting thing it does. Its real function is to stop you sending irrelevant mail to people who will punish you for it, and since February 2024 that punishment has a number attached: Google and Yahoo cap bulk senders at a 0.3% spam complaint rate, and irrelevance is how you get there.
The measured effect, with its age stated
The most-cited study on this is Mailchimp’s analysis of segmented campaigns, drawn from around 2,000 accounts sending roughly 11,000 segmented campaigns to nearly 9 million recipients. It reported opens 14.31% higher than non-segmented campaigns, clicks 100.95% higher, unsubscribes 9.37% lower, bounces 4.65% lower, and abuse reports 3.90% lower.
+100.95%
Clicks on segmented vs non-segmented campaigns
−9.37%
Unsubscribes on segmented campaigns
−3.90%
Abuse reports on segmented campaigns
Source: Mailchimp, Effects of List Segmentation on Email Marketing Stats (data as of 2017).
That data is from 2017, which is worth saying out loud rather than laundering the number as current. Two things follow. The open-rate lift is the least reliable figure in it, both because it predates Apple Mail Privacy Protection and because clicks were always the better measure. The complaint and unsubscribe reductions are the ones that have grown more valuable since, because the enforcement environment changed underneath them.
Start with the segment that protects you
Before any clever behavioral targeting, split your list on engagement: people who opened or clicked in the last 90 days, people who did not. Send your regular campaigns to the engaged group and route the rest into a re-engagement flow. This single split does more for deliverability than every other segment combined, because mailbox providers score you on how recipients react and dormant subscribers react by ignoring or complaining.
One complication: with Apple Mail Privacy Protection loading pixels automatically, "opened in the last 90 days" now includes people who never saw the message. Define engagement on clicks, site visits, and purchases where you can, and treat opens as a weak supporting signal rather than the definition.
The segments worth building next
| Segment | Data needed | What it changes |
|---|---|---|
| Engaged vs dormant | Click and open history | Protects deliverability. Build this first. |
| Lifecycle stage | Order count | Separates the pitch to a prospect from the pitch to a repeat customer. |
| Signup source | A tag set at opt-in | Nearly free if captured at the start, impossible to reconstruct later. |
| Purchase category | Order line items | Stops you sending the wrong catalogue to the right person. |
| RFM: recency, frequency, monetary | Full order history | Identifies the small group producing most of the revenue, and the ones lapsing out of it. |
| Location | Address or IP | Only useful if you have location-specific offers or events. Otherwise noise. |
RFM, without the spreadsheet
Recency, frequency, and monetary value is the oldest segmentation model in direct marketing and still the most useful for a small business, because it answers a question you can act on: who bought recently, who buys often, and who spends most. You do not need scoring quintiles. Four buckets will do: new customers, repeat customers, high-value customers, and lapsing customers who used to be one of the other three. Each gets a different message and a different frequency.
Behavior over demographics
What someone did predicts what they will do better than what category they fall into. A subscriber who clicked three emails about a product line is a stronger target than one who matches an ideal customer profile on paper. Behavioral data also has the advantage of being first-party, observed, and current, which matters increasingly as third-party data becomes both less available and less legally comfortable to rely on.
Keep segments dynamic so membership updates as behavior changes, and revisit the definitions quarterly. A segment defined around a product you no longer sell will keep quietly filling with people and quietly receiving nothing.
Key takeaways
- Mailchimp’s 2017 analysis found segmented campaigns produced roughly double the clicks and materially fewer unsubscribes and abuse reports.
- The complaint reduction matters more now than the open-rate lift, because Google and Yahoo cap bulk senders at a 0.3% spam rate.
- Build engaged versus dormant first. It protects deliverability more than every other segment combined.
- Define engagement on clicks and purchases, not opens, since Apple Mail Privacy Protection loads pixels automatically.
- A segment that would receive the same email as another segment is not a segment. Over-segmenting costs writing time and test validity.
Related reading
Sources
- Effects of List Segmentation on Email Marketing Stats, Mailchimp (2017)
- Email sender guidelines (spam rate threshold), Google Workspace Admin Help (2024)
- Use Mail Privacy Protection on iPhone, Apple Support
- Email Marketing
- Strategy
- Personalization
Written by
Valter Brandt
Chief Marketing Officer
Valter Brandt is the Chief Marketing Officer of ThisCom, working with clients across the United States and internationally. He has led marketing strategy through the major shifts in social advertising, mobile, content marketing, programmatic media, and marketing automation.
All articles by Valter Brandt →
