Introducing Cresca AI Features
Cresca's AI features are now available on every paid plan. This is an introduction to what was built and, more usefully, how to adopt it without breaking the campaigns you already have running.
What shipped
Six capabilities, each aimed at a specific bottleneck in campaign production:
- Campaign generation from a written brief, producing a draft with subject lines, body copy and layout.
- Subject line variants for testing, generated in multiple angles rather than variations on one.
- Template generation from a description, best suited to emails you send repeatedly.
- Send time optimisation, selecting a per-recipient send window from historical engagement.
- Plain-language segmentation, which builds a filter from a described intent.
- Personalisation across the contact data you already hold.
Why these six
Each of these corresponds to a step where campaign work actually stalls. Writing the first draft is the slowest part of producing a campaign. Choosing a subject line is where teams debate longest over the smallest text.
Building a layout is mechanical. Picking a send time is a guess for most teams. Writing a segment rule in boolean terms invites silent errors.
And personalisation is abandoned by most teams not because it is hard, but because doing it correctly requires data hygiene work that never gets prioritised.
None of these are novel problems. The point of building them into the platform is that the fix lives where the campaign is built, rather than in a separate tool you have to remember to use.
Adopting it without disrupting live campaigns
Do not switch everything at once. The sensible sequence is to start where the cost of being wrong is lowest.
- Start with templates. Generate the layout for a recurring email you send every month. If it is wrong, you still have time to rebuild it by hand before the send.
- Next, campaign drafts. Use AI generation for a regular newsletter where a mediocre draft is cheap to improve. Do not start with your most important send of the quarter.
- Then segmentation. Describe a segment you would normally build manually, generate it, and compare the resulting count against what you expected. That comparison is the entire value of the exercise.
- Send time optimisation last. It changes when contacts hear from you, which makes its effect hardest to attribute. Enable it once you have a stable baseline to compare against.
What to watch
Two things will determine whether this helps you.
Data quality. Personalisation and segmentation both depend on the accuracy of your contact data. A segment built from a described intent is only correct if the underlying fields are populated and current. Check your data before enabling either.
Metric choice. Click and conversion, not open rate, are what tell you whether generated content worked. Open rates are inflated by image proxying and prefetching, so a subject line that "wins" on opens may simply have been delivered to a client that prefetches images.
What has not changed
Delivery, automation and analytics behave exactly as before. AI features generate content and suggest configuration; they do not alter how messages are sent, how workflows execute, or how events are recorded. Existing campaigns and automations continue to run unchanged, and nothing needs to be migrated.
You can use the AI features entirely, not at all, or selectively by campaign type. The platform does not require it.
Getting started
Create a campaign, describe what you want to send, and read the draft. That is the whole onboarding. If the output is not to your taste, the fastest correction is to be more specific in the brief: audience, purpose, tone and the single action you want the reader to take.
Pricing
AI features are included on every paid plan. Tiers differ by scale:
| Plan | Price | Contacts | Emails / month |
|---|---|---|---|
| Free | $0 | 50 | 50 |
| Professional | $29/mo | 5,000 | 5,000 |
| Premium | $49/mo | 25,000 | 25,000 |
| Ultra | $99/mo | 55,000 | 55,000 |
What each feature costs you in time, honestly
Adoption guidance is more useful with expectations attached.
- Template generation. Saves the most time on the first build and almost none afterwards. Build once, then reuse.
- Campaign drafts. Saves the most time when you have a clear brief, and close to none when you do not. A vague brief takes longer than writing from scratch.
- Subject line variants. Saves the debate time, not the writing time. The value is unblocking a team, not producing a better line.
- Segment construction. Saves the most time on complex filters. For a simple rule, building it directly is faster.
- Send time optimisation. Saves no time at all; it is a performance feature, and it needs historical data to do anything.
Setting expectations with a team
The most common failure after enabling these features is the belief that campaigns no longer need review. They do, and the review is a different kind of work than before: rather than writing, you are checking. Teams that skip the check ship confident errors, usually a factual claim about the product or a personalisation field that renders empty.
A useful norm is that generated content is a draft from someone who has not been in the room. It has read the brief and nothing else. Treat it accordingly.
Rolling it out across a team
Start with one person and one campaign type. Establish what a good brief looks like and what the review checklist is, then share both. A team adopting this simultaneously without an agreed brief format produces inconsistent results and concludes the feature is unreliable.
Document the brief format once. It is the single artefact that determines whether adoption works.
The short version
The AI features remove mechanical work from campaign production: drafting, subject line selection, layout building, send scheduling, segment construction and personalisation. Adopt them in order of rising risk, start with templates and drafts, check your contact data before enabling anything that depends on it, and measure results on clicks and conversions rather than opens. Expect review work rather than no work, agree a brief format before rolling it out to a team, and remember that cheaper production does not create a reason to send.
Continue learning
Related Cresca resources
External references