Three engagements. Three different accounts. The same lab approach applied every time: go in, find the gaps, run the experiments, compound the findings. Here is what happened.
Client details withheld by request. Industry and revenue ranges shared with permission. Specific figures to be updated with verified client data.
The brand had made a considered decision (financial) to move from Bloomreach to Klaviyo. The migration was handled competently – flows rebuilt, data transferred, the platform working as it should. But rebuilding is not optimising. Everything was live and technically functional. Nothing had ever been looked at properly since go-live.
The abandoned cart flow was firing 4 hours after abandonment. The welcome series was a single email. Post-purchase, win-back, and browse abandonment flows were not configured. Segmentation consisted of date-based filtering with no behavioural logic.
The diagnostic audit identified a significant monthly revenue gap – the difference between what email was making and what it should have been making on a brand at this level.
We started with the abandoned cart trigger because the data said it was the highest-leverage change. Moving the first email from 4 hours to 1 hour catches customers while intent is still warm – before they have bought elsewhere, talked themselves out of it, or simply forgotten. The recovery rate nearly doubled. On a brand at this revenue level, that single timing change was worth a meaningful amount every month.
We then rebuilt the welcome series from one email to four, each with a specific role: setting expectations, telling the brand story, building social proof, and a soft conversion push at the right moment. Revenue per recipient improved approximately four times over.
Over 90 days we worked through post-purchase, win-back, and browse abandonment flows – none of which had been configured. By the end of the third month the programme was generating five figures per month more than it had been at the point of the audit. The migration from Bloomreach had been done well. Now it was being used properly.
The brand had a list of 94,000 subscribers and was sending the same weekly campaign to all of them. Open rates were around 18%, click rates around 2.1% – numbers that looked acceptable until the diagnostic showed what was actually happening underneath.
Inbox placement had been declining for six months – from close to 98% down to around 91%. At that level more than 1 in 12 emails was not reaching the inbox at all. The root cause: a significant portion of the list was disengaged, and mailing them weekly was quietly damaging sender reputation with every send.
The second problem was equally significant. The top customers – a small percentage of subscribers accounting for a disproportionate share of revenue – were receiving identical emails to someone who had bought once more than a year ago. Fundamentally there was no acknowledgement of who they actually were.
We started with deliverability because everything else depends on it. Before you can optimise revenue you need your emails to land. It doesn't matter how good the offer is or how good the email is. We built a sunset flow to identify and suppress chronically disengaged subscribers, established engagement-based segments, and shifted campaign sends to the engaged portion of the list only – roughly a third of the total.
Sending to a smaller, engaged audience improved both deliverability and revenue per recipient simultaneously. Inbox placement started recovering within the first two weeks and was above 99% within 60 days. Revenue per recipient on campaigns improved significantly – sending less frequently to the right people consistently outperformed blasting the full list.
We then built the RFM segmentation model – eight segments based on recency, frequency, and monetary value. For the Champions segment we built a VIP early access sequence: new arrivals and exclusive offers sent before the general list. Open rates and conversion on these sends were significantly above the brand average. By month 4, email-attributed revenue had increased by 68%.
This engagement was different from most. The brand wasn't neglecting email – they were doing it actively, regularly, with genuine effort. Six flows live, weekly campaigns, A/B testing on almost every send. Email attribution at 31% of total revenue, which most brands would consider strong.
The diagnostic audit found something more nuanced than a neglected account. The A/B testing programme, which the team was proud of, had a fundamental flaw: average sample size per variant of 340 contacts, average test duration of 6 hours, estimated statistical power of approximately 32%. Which meant roughly 68% of their tests were producing results statistically indistinguishable from noise.
The predictive analytics module in Klaviyo had been enabled for 14 months and never reviewed. Sitting inside it: predicted lifetime value scores, next purchase probability, and churn risk indicators for their entire list of 180,000 subscribers. Completely unused.
We started somewhere unusual: we retested their most recent "winning" subject line variant – the one selected on a small sample over a few hours and rolled out to their entire list. Looking at the sample size and test duration, the result was almost certainly noise. We ran it properly with a significantly larger sample over multiple send cycles. The previous "winner" actually performed worse when tested with sufficient data. This was uncomfortable to hear. We showed the full picture. The team understood immediately and it changed how they thought about every test result they had ever acted on.
We rebuilt the approach to testing from the ground up. Every change is now considered carefully before being made, with a clear view of what success looks like and enough data to know if it is real. In parallel, we built a segmentation model using the predictive analytics data that had been sitting unused in Klaviyo for over a year – identifying high-value customers showing early signs of disengagement and building a retention sequence specifically for them.
The incremental revenue from that segment alone was meaningful month on month. Over six months, with continuous attention and improvement across all flows and campaigns, email attribution moved from 31% to 39% of total revenue. The compounding effect of genuinely paying attention, every month.
The diagnostic audit is where every engagement starts. 72 hours. Written findings. No obligation to continue.
Book the diagnostic audit