Send time optimisation gets a lot of airtime in email marketing circles. Nearly every major platform now offers some flavour of AI-predicted send timing, and the marketing copy around it implies it's one of the highest-leverage levers you can pull.
That claim deserves scrutiny. This post looks at what send time optimisation actually is, what the evidence says about its impact, when it genuinely helps, and when it's a distraction from the things that actually move revenue.
What Send Time Optimisation Actually Does
Send time optimisation (STO) — sometimes called predictive send time or smart send time — works by analysing each subscriber's historical engagement behaviour to predict when they're most likely to open an email. Instead of sending a campaign to your full list at 10am Tuesday, the platform staggers delivery across a window (typically 24 hours) so each recipient gets the email at their personalised optimal time.
The underlying model is trained on things like: what time a subscriber has opened past emails, what day of the week they tend to engage, and — in more sophisticated implementations — how that behaviour compares to similar subscribers when individual data is sparse.
It sounds compelling. In practice, the results are more nuanced.
What the Data Actually Shows
The headline claim from most STO implementations is a 2–5 percentage point improvement in open rates. Platform data and independent studies suggest campaigns using send time optimization see open rates roughly 3 points higher on average.
That sounds meaningful — until you account for two things.
First: open rate is broken as a metric. Apple Mail Privacy Protection has been inflating opens since 2021. Gmail's Gemini AI summaries have added a second layer of pixel pre-rendering since early 2026. Across most ecommerce lists, reported open rates are running 20–30 percentage points above the true human-read rate. A 3-point improvement on an already-inflated metric is a small signal buried in a lot of noise.
Second: open rate doesn't pay for inventory. The question isn't whether STO gets more pixels loaded — it's whether it generates more revenue. That connection is much harder to establish from platform marketing materials, because the studies rarely isolate STO as the sole variable, and they're conducted by the same companies selling the feature.
The more reliable signal: click rate and revenue per recipient. There's limited independent evidence that STO meaningfully moves either at the population level, though individual cases vary significantly.
When STO Helps (And When It Doesn't)
Send time optimisation is not a universal lift. Its value depends heavily on three factors:
List size and data quality
STO requires enough historical engagement data per subscriber to make a reliable prediction. Platforms typically need 3–5 prior interactions with a subscriber before their individual model is meaningful. Below roughly 5,000 subscribers, the list is too small for the per-subscriber predictions to be statistically robust — you're essentially sending at an educated average rather than a genuine personalisation.
If your list is under 5,000, skip STO. Focus on growing the list and improving offer quality instead.
Audience composition
STO tends to perform best when your list has diverse engagement patterns — some subscribers active in the morning, others at lunch, others in the evening. If your audience is relatively homogeneous (say, a B2B tool where most subscribers check email during UK business hours), the variance in optimal send times is narrow and STO adds little over a well-chosen fixed window.
For consumer ecommerce with a geographically diverse list, the timing variation is higher and STO has more to work with. Understanding your different subscriber types can help determine whether STO is worth implementing.
Campaign type
Automated flows — abandoned cart, post-purchase, welcome series — are typically triggered by behaviour rather than broadcast at a fixed time. STO in the traditional sense doesn't apply to most flows. For broadcast campaigns, the value depends on your list size and composition as above.
The General Benchmarks (When You Don't Have STO)
If you're sending without STO or your list is too small for it to be meaningful, here are broad industry benchmarks for ecommerce email in 2026:
| Day |
Performance vs. Average |
| Tuesday |
+Above average |
| Wednesday |
+Above average |
| Thursday |
+Slightly above average |
| Monday |
Roughly average |
| Friday |
Slightly below average |
| Weekend |
Below average (category-dependent) |
| Time Window (Recipient Local Time) |
Performance vs. Average |
| 10am – 12pm |
Above average |
| 1pm – 3pm |
Above average |
| 8am – 10am |
Roughly average |
| 6pm – 8pm |
Slightly below average |
| After 9pm |
Below average |
These are averages across industries from aggregated email platform data. They should be treated as starting points, not gospel. A list of shift workers, parents of young children, or subscribers in a different timezone will behave differently. Your own historical data — even two or three campaigns' worth — is more actionable than industry averages.
The Honest Priority Order
If you're trying to improve email revenue, send time optimisation belongs low on the list. The honest priority order for most ecommerce merchants:
- Offer quality. A mediocre offer at the perfect time still underperforms a strong offer at a decent time.
- Segmentation. Sending the right email to the right segment — purchasers vs. non-purchasers, high-value vs. new subscribers — moves revenue more than timing.
- Automation coverage. If you don't have a functioning abandoned cart sequence, a post-purchase flow, and a welcome series, these will generate more per-email revenue than any send time tweak to your broadcast campaigns.
- Subject line and preview text. These are the actual determinants of whether a human opens and engages. They're worth more testing time than delivery timing.
- Send time. Once the above are in reasonable shape, STO is worth enabling if your platform supports it and your list is large enough. It may give you a small lift. It won't transform underperforming campaigns.
The Attribution Problem With STO Testing
One practical note if you're trying to measure STO's impact yourself: it's genuinely hard to isolate. The correct test is an A/B split where one cohort gets STO delivery and a matched cohort gets a fixed send time, with revenue per recipient as the outcome metric (not open rate).
Most platform-native testing doesn't set this up cleanly. The easy test — comparing open rates before and after enabling STO — conflates timing changes with list maturation, seasonal effects, and offer variation. Treat any internal "STO improved our open rate by X%" claim with appropriate scepticism unless you can trace it to a controlled experiment.
Key Takeaways
- Send time optimisation predicts each subscriber's optimal engagement time and staggers delivery accordingly
- Evidence suggests a 2–5 percentage point improvement in open rates on average — but open rate is an unreliable metric in 2026 due to privacy-related inflation
- STO is most valuable on lists over 5,000 with diverse engagement patterns
- For most ecommerce merchants, segmentation, automation coverage, and offer quality will move revenue more than timing
- If you're not using a platform with native STO, Tuesday–Thursday, 10am–3pm in your audience's local timezone is a reasonable default
- Measure STO impact via click rate and revenue per recipient, not open rate
Related Articles
