Four emails, three of them fired by behaviour rather than the calendar. The one most products never send is the one that matters most.
A trial email sequence exists to get someone to first value, then to a decision. Most sequences do neither, because they are timed on days elapsed rather than on what the person has actually done.
Three of those four are behavioural triggers, not scheduled sends.
The distinction matters more than the copy. A message saying "you signed up three days ago" to someone who activated on day one reads as noise, and noise gets filtered.
The standard drip runs day 1, day 3, day 7, day 12. It treats a power user and someone who never logged in identically, which means half your sequence is wrong for whoever receives it.
Behavioural triggers fix this without much extra work. You need one event instrumented — your activation moment — and everything else follows from whether that event has fired. The onboarding framework covers naming it.
| Trigger | Send | Do not send |
|---|---|---|
| Signed up, no activation, 48h | The one-step nudge | Feature highlights |
| Activated once, nothing since | A second use case | A setup guide they finished |
| Using it regularly | Expiry reminder only | Onboarding tips |
| Never logged in at all | One message, then stop | The full sequence |
One action. Not a welcome, not a tour, not your founding story.
The single most common mistake is listing everything the product does. A new signup has one question — what do I do first — and a list of nine features does not answer it. Name the one thing that produces the first useful outcome and link straight to it.
Send it from a personal address. At solo scale, replies to this email are some of the most useful research you will get, and a noreply address throws that away.
Fires only if someone has not reached first value within 48 hours. For most early SaaS this is the single most valuable message in the entire lifecycle, because it catches the largest single group of people you are losing.
Keep it to one obstacle and one link. "Most people get stuck on X — here is the two-minute version" outperforms a general offer of help, because it names something specific and implies you know what happens.
Asking a question works too: what were you hoping to use this for? Replies to that question are the raw material for your positioning, and the people who answer are disproportionately likely to convert.
Three days out. Two things in it: what they have actually done in the product, and what happens on the expiry date.
The usage recap does the persuading. "You have created 14 invoices and saved roughly three hours" is a stronger argument than any feature list, and it only works because it is true and specific to them. If you cannot generate that line, you do not have the instrumentation and that is the first thing to fix.
Be explicit about the mechanics — whether the card is charged, what happens to their data, how to cancel. Ambiguity here produces support emails and cancellations from people who were otherwise happy.
Short. What they no longer have access to, and one link to continue.
Do not offer a discount here. A discount at expiry teaches everyone who hears about it to let their trial lapse, and it converts people at a permanently lower price who would often have paid full. If conversion at this point is poor, the problem is upstream — see trial-to-paid conversion.
Four emails is the whole sequence. Adding a fifth and sixth reliably reduces reply rates and raises unsubscribes, and the marginal conversion after email four is close to zero. If four is not working, the fix is activation, not volume.
The four triggers stay the same; only the expiry timings move. On a seven-day trial the pre-expiry email lands on day four. On a 14-day trial, day 11. On 30 days you need one extra mid-trial check because the dead middle is long enough for people to forget entirely.
If you are choosing the length itself, that decision follows from time-to-value rather than from the sequence — the trial length playbook covers it.
You do not need a marketing automation platform. Most billing and email tools can fire on a custom event, and the whole sequence is four templates plus one instrumented event.
Build it in this order: email 2 first, then 3, then 1, then 4. That is descending order of impact, so if you stop halfway you have still built the parts that matter — SaaS business automation covers the wider set of things worth automating first. For email templates for every situation beyond just the trial sequence — cancellations, outages, price changes — the full set is worth having ready before you need it.
Four: one on signup with a single action, one when someone has not activated within 48 hours, one three days before expiry showing what they have done, and one on expiry. Three of the four should fire on behaviour rather than on elapsed days.
Behaviour-based wherever possible. A standard day 1, 3, 7, 12 drip treats a power user and someone who never logged in identically, so half the sequence is wrong for whoever receives it. You only need one instrumented event to fix this.
The non-activation nudge at 48 hours. It reaches people at the moment they are deciding whether signing up was worthwhile, it catches the largest group you are losing, and most products never send it at all.
No. A discount at expiry teaches everyone who hears about it to let their trial lapse, and it converts people at a permanently lower price who would often have paid full. Poor conversion at expiry is usually an activation problem upstream.
Four. Adding a fifth and sixth reliably reduces replies and raises unsubscribes, and marginal conversion after the fourth is close to zero. If four emails are not working, the fix is activation rather than more volume.
Tell Marcus your activation rate and trial length. You get the trigger to build first and the message to put in it.
Try GhostCoach free →14-day free trial · cancel anytime · 30-day money-back on Lifetime