Lifecycle email case studies, taken apart

The day authentication stopped being optional

Rule change

Google and Yahoo · 1 February 2024

What was done: Google and Yahoo began requiring bulk senders to authenticate with SPF, DKIM and DMARC together, keep spam complaints under 0.3%, and offer one-click unsubscribe honoured within two days.
What was right: Both published the requirements months ahead, with thresholds and dates. This was one of the best telegraphed changes in email, and the documentation is still there.
What went wrong: Senders who had run for a decade on one SPF record kept sending exactly as before. There is no bounce for failing DMARC alignment, so open rates fell quietly and teams blamed subject lines.
What should have happened: Check all three records before the deadline, not after the numbers move. Two out of three counts as none, and passing SPF for your tool’s domain does nothing if your From address is your own.

What the right plan does to the economics

Our own arithmetic on a modelled sender: 80,000 emails a month, 38% open rate, 2.1% click, 3% of clicks converting at 210 dollars. That is roughly 40,000 dollars a month attributable to the channel. A silent 30% delivery loss removes about 12,000 dollars a month and shows up nowhere in the sending dashboard, because the dashboard counts sends.

Google bulk sender guidelines and Yahoo sender requirements, effective 1 February 2024 · revenue arithmetic ours

The second email that goes to everybody

Automation

Any sequence · ongoing

What was done: A five step welcome sequence is built. The first email is written carefully. Steps two to five are sent to the whole list on a fixed schedule, whatever anybody did.
What was right: Having a sequence at all beats having none, and the first email is usually genuinely good because somebody cared about it.
What went wrong: People who already bought are asked to buy. People who never opened receive the same message again with a new subject. Both groups learn that your mail is not about them, and the unsubscribes cluster on step two.
What should have happened: Add one condition after email one: opened or not. Two branches, different content. It is ten minutes in any tool and it is the cheapest improvement available.

What the right plan does to the economics

Our own arithmetic on a 24,000 person list: a 0.4% unsubscribe rate on step two is 96 people per send. Branching typically halves that in our modelling. Over twelve sends a year that is roughly 576 subscribers kept. At a modelled 4.80 dollars annual value per subscriber, about 2,765 dollars a year recovered from a change that costs one afternoon.

Our own arithmetic on a modelled list, offered as a worked example rather than a case

The warm lead nobody owned

Handover

Any B2B team · ongoing

What was done: A lead magnet converts well. Completed forms land in a shared inbox that both marketing and sales can see, and everybody agrees somebody will pick them up.
What was right: The magnet itself worked and the volume was real. The failure is downstream of a genuine success, which is what makes it so common.
What went wrong: A shared inbox is not an owner. Leads sat two to three days, by which point the person had spoken to somebody else, and both teams honestly believed the other was handling it.
What should have happened: Name a person per day, not a team. Measure the time between the form landing and a human replying, and put that number where both teams can see it.

What the right plan does to the economics

Our own arithmetic: 60 warm leads a month, a 22% close rate when contacted the same day and roughly 9% after two days, at a 3,400 dollar average deal. Same-day contact is about 44,900 dollars a month, two-day is about 18,400. The gap is roughly 26,500 dollars a month, and the fix is a name on a rota.

Our own arithmetic on a modelled B2B team; close-rate spread modelled, not measured by us

Dates and thresholds under Google and Yahoo come from their published guidelines. All revenue figures are our own arithmetic on modelled senders and are labelled as such.

Three sequences, taken apart.

The day authentication stopped being optional

The second email that goes to everybody

The warm lead nobody owned

One of these is happening to you?