Costs
Part of What matters most in email automation: five priorities
Email automation benchmarks: hard limits, planning bands, and your own baseline
Explains no universal email automation benchmark exists, covers fixed spam and bounce limits, planning ranges by journey, reading reports, and setting a baseline.
What to take away
- No universal email automation benchmark exists. A frozen internal baseline, plus an eligible holdout, is the substitute.
- Absolute limits do exist. Keep spam complaints under 0.1% of delivered mail and never above 0.3%. Keep hard bounces under 2%.
- Planning bands differ by journey type. Welcome, cart and win-back journeys do not share a click target.
- Print the denominator, window, version and date beside every figure, or nothing is comparable.
- Last-click automation revenue is an upper bound until a holdout shows the lift.
What is fixed, and what is relative
Mailbox providers set hard limits. Gmail and Yahoo bulk sender rules in force since 2024 require one-click unsubscribe and keep spam complaints below 0.3% of delivered messages, with 0.1% as the working target. Those figures are floors, not goals.
Consent and opt-out rules are equally absolute. The FTC's CAN-SPAM compliance guide states what a commercial message must contain and how an unsubscribe request must be honoured.
Everything above that line is relative. Clicks, revenue per send and retention depend on audience, offer, cadence and market. Email automation sorts out which of those variables you control and which you inherit from your list.
Planning bands for common journeys
These are typical planning ranges, not published vendor figures. Use them to sanity-check your own numbers, then replace them with your baseline.
| Journey | Metric | Typical range | Denominator |
|---|---|---|---|
| Welcome | Click rate | 5 to 15% | Delivered |
| Welcome | Open rate, directional only | 40 to 60% | Delivered |
| Abandoned cart | Click rate | 8 to 20% | Delivered |
| Abandoned cart | Recovered carts | 5 to 15% | Carts created |
| Browse abandonment | Click rate | 3 to 10% | Delivered |
| Post-purchase | Repeat order in 60 days | 5 to 15% | Buyers enrolled |
| Win-back | Click rate | 2 to 8% | Delivered |
| Win-back | Reactivation | 1 to 5% | Lapsed cohort |
| Any journey | Spam complaint rate | Under 0.1%, ceiling 0.3% | Delivered |
| Any journey | Hard bounce rate | Under 2% | Sent |
| All automations | Share of email revenue | 20 to 40% | Total email revenue |
The bands are wide on purpose. List age and offers move a click rate far more than copy does. A 15% cart click rate on a small, high-intent list is not comparable to 8% on a broad one.
How to read published benchmark reports
Provider benchmark reports are the nearest external reference. Check the sample size, the publication date, the metric definition, and whether campaigns and automations are pooled together. Open rates have run high since Apple's Mail Privacy Protection arrived in 2021, so treat them as directional at best. org/wiki/Email_deliverability).
Reports older than about 18 months rarely hold up. Targets for campaigns rather than journeys are gathered in email campaign planning benchmarks, which keeps the two reporting jobs separate.
How journey types differ
Trigger and timing separate them. Marketing automation platforms group these triggers differently, and this overview of marketing automation covers the shared feature set.
- Cart journeys earn the most revenue per send, because intent already exists.
- Welcome sequences carry the highest click rates, then decay after the third or fourth message.
- Post-purchase sequences rarely earn direct revenue. Score them on repeat orders and fewer support tickets.
- Win-back journeys cost little to send and post the lowest click rates. Judge them on reactivation.
- Browse abandonment swings on catalogue size, and volume moves more than quality.
Side-by-side comparisons of live programmes sit in email automation examples, which shows how one trigger performs under very different offers.
Example: setting a threshold for a win-back journey
- Freeze the populationone-time buyers with no order in 90 days and permission still valid.
- Split it. Two thirds enter the journey, one third becomes the holdout and receives nothing.
- Fix the window at 60 days after entry so every cohort matures the same way.
- Set the threshold from your own history, not from a vendor report.
- Compare reactivation in the journey group against the holdout. Report the gap as lift.
- Rewrite or retire the journey if lift stays at zero for two quarters. The recovery process is described in how to improve email automation.
Data quality behind the numbers
Broken events make honest benchmarks impossible. Track expected against received events, missing fields, duplicates, late arrival, order reversals, version mismatch and unidentified profiles.
[figure 1]
Event and identity quality checks
- Expected versus received events
- Missing fields and duplicates
- Late arrival and order errors
- Reversals and version mismatch
- Unidentified profiles and incorrect merges
- Reconciliation with source systems
Eligibility decisions need the same check: correct entry, send-time permission, market, preference, frequency, suppression and exclusions.
[figure 2]
Eligibility accuracy checks
- Correct entry decisions
- Send-time permission
- Market and topic eligibility
- Preference and frequency rules
- Suppression and service state
- Exclusion decisions
- Queued messages canceled after eligibility changes
