It reads the signals. Then it writes the email.
Segmentation tells you who to send to. AeroCampaign’s engine goes the step nobody automated: it decides what each group should actually read, grounds it in what is moving in their industry, and drafts the copy for your review.
Same campaign, three contacts, three emails
Each of these contacts is on the same list and would receive the same message from a conventional platform. Switch between them to see what the engine does with their history instead.
- Opens, last 30d
- 9 of 11
- Pricing page clicks
- 4
- Recency
- 2 days ago
- CRM stage
- Evaluation
Diesel surcharges down 6% QoQ — cost-per-load is the sector's live conversation.
You've been back to the pricing page four times this month, so I'll skip the pitch and go straight to the arithmetic.
At Northsea's volume, the difference between bundled sending and sending at cost is roughly $18,400 a year. With surcharges easing 6% this quarter, that's margin you can actually keep.
I've put the breakdown on one page — no call needed.
What counts as a signal
Nothing here requires new instrumentation. These are the events an email platform already records, read properly instead of reduced to an open rate.
Behavioural
What the contact has actually done, weighted by recency.
- Open history
- Rate and streak over rolling windows
- Click topology
- Which links, in what order, how deep
- Cadence response
- Engagement change as frequency changes
- Dwell between touches
- Time from send to first interaction
Contextual
Who the contact is and how they arrived.
- List provenance
- The form, import, or campaign of origin
- Firmographics
- Company size, sector, geography
- CRM stage
- Pipeline position where a CRM is connected
- Declared preferences
- Topics and frequency the contact chose
Environmental
What is happening in the contact's market, not yours.
- Sector momentum
- Direction of travel in their industry
- Seasonal demand
- Where they are in their buying year
- Peer benchmarks
- How similar companies are behaving
- Topic salience
- What their sector is talking about now
Personalization has a cost. Spend it where it pays.
Generating a unique email for every contact on a 200,000-person list is possible and almost always wasteful. The engine chooses its resolution per segment based on how much signal it actually has.
Cohort
The default. Contacts cluster on behaviour, and each cluster gets its own subject line, preview text, and body. Cheap to generate, easy to review, and the bulk of the lift.
Micro-cohort
Where a cohort splits on a meaningful secondary signal — sector, lifecycle stage, or a specific click path — the engine forks a variant rather than averaging across the difference.
Individual
Reserved for contacts whose signal history is strong enough to justify it: high-value, high-intent, or high churn risk. The engine writes to that person's actual history rather than to a segment average.
An engine that proposes, not one that publishes
Nothing sends unreviewed
Generated variants land as drafts against their cohort. You approve, edit, or discard — the engine proposes, you ship.
Every variant is attributable
Each generated variant carries the signals that produced it, so a lift or a flop is traceable to a reason rather than a vibe.
Personalization respects suppression
The engine writes against the same suppression set the sender enforces. A suppressed address is never a cohort member.
See what your own list looks like segmented.
Connect a provider, import a list, and let the engine show you the cohorts hiding inside it.
No credit card. Bring your own SendGrid or Amazon SES account.