We started a Klaviyo audit for a European pet-care DTC brand expecting the usual list of small fixes — a subject line here, a segment there. What we found instead was a flow generating essentially zero revenue for reasons the account owner had no way of seeing from inside the Klaviyo UI, plus a chain of smaller structural problems that were quietly taxing every other flow in the account. Here's the diagnosis, what we fixed over the following two months, and what actually moved — including the parts that didn't.
What did the initial audit find?
The account had 18 active flows and zero global exclusion logic between any of them — no rule anywhere saying "if this customer is already in Flow A, don't also enter Flow B." That absence turned out to be the root cause of the single biggest finding: a Returning Customer flow (10 emails, built to re-engage repeat buyers) and a Post-Purchase flow were both triggered by the same event. Klaviyo's own Smart Sending protection — designed to stop customers getting flooded with email — was quietly suppressing the second flow's messages before they ever sent. Total attributed revenue across that entire 10-email sequence: roughly the cost of a single ad click. It had been running that way for months.
That wasn't the only structural issue. The Welcome Series fired an email and an SMS at the exact same moment with an identical discount code — a duplicate incentive that also made it impossible to measure which channel actually drove the click. The Abandoned Checkout flow split traffic into "high value" and "low value" tracks at a threshold roughly double the account's actual average order value, so the high-value branch was starved of traffic it should have been getting. And on the subscription side, a broken entry filter on the subscription welcome flow had let through exactly one profile in the flow's entire lifetime — every other new subscriber had been silently excluded since the flow was built.
What did we fix first?
Month one was foundation work, not new campaigns. We added purchase-verification checks and global suppression logic to the highest-revenue flows so customers stop re-entering sequences for things they've already bought. We rebuilt the Welcome flow with a staggered send (SMS first, email a few minutes later, not simultaneous) and cut it from a long nurture sequence down to three focused emails over two days, moving the lower-performing content to a separate track. We reset the Abandoned Checkout's value-tier split to match the account's real average order value instead of the inflated threshold it had been running on. And we built two flows that hadn't existed at all: browse abandonment and cart abandonment, both from scratch.
Month two shifted to the flows that needed more than a logic fix. The 10-email Returning Customer sequence — the one Smart Sending had been silently blocking — was retired and replaced with a tighter 4-email structure (welcome back, value content, an offer, a final nudge) with conditional splits that remove anyone who's already converted. We found and fixed the broken subscription filter. We built a dedicated first-order retention flow for new customers specifically, separate from the generic post-purchase content that had been serving everyone the same message regardless of purchase history. And we built the account's first real segments — a first-time-buyer segment and an at-risk segment (customers 60–90 days past their last purchase with no recent activity) — neither of which had existed before.
What actually changed?
The Abandoned Checkout flow — fixed in month one, redesigned visually in month two — saw revenue per recipient grow by roughly 27% month over month, with total flow revenue holding steady while efficiency per send improved. That's the cleanest read of the logic fix working: the same audience is now worth meaningfully more per email sent.
The two flows we built from nothing — browse abandonment and cart abandonment — launched in their first full month and generated just over €1,100 in combined attributed revenue with zero prior optimization, purely from being live. That's not a huge number in isolation, but it's real revenue from two flows that captured zero dollars the month before, because they didn't exist.
What we can't yet report: a clean "after" number for the Returning Customer rebuild, because the Finnish-language version was still pending translation sign-off at the two-month mark — it hadn't been live long enough to generate a comparable result. Same for the new first-order retention flow. We're not going to backfill numbers that don't exist yet.
What's still unresolved?
Plenty, and we'd rather list it than pretend the account was "done" at two months. The at-risk segment we built (a meaningful chunk of the list) doesn't have an automated re-engagement flow running against it yet — that's queued for month three. An order-reminder flow for subscription customers turned out to be almost entirely non-functional (a routing condition sent effectively everyone down a dead-end branch) and is scheduled for a fix next. The loyalty-program flows haven't been touched yet. And a cost-efficiency item flagged in the original audit — an SMS plan sized well above actual usage — doesn't have a confirmed resolution in what we've reported so far, which is worth being honest about rather than quietly dropping.
None of this is a dramatic before/after screenshot with a single headline number — it's a handful of concrete, verifiable fixes, some of which paid off within weeks and some of which are still mid-build. That's what a real audit looks like in practice: less "10x your revenue," more "here's a flow that was silently making nothing, here's why, here's what replaced it." If you suspect your own account has a version of the Returning-Customer problem — a flow quietly capped by another flow's Smart Sending, or a segment that's never actually been built — a SciGrowth Klaviyo Audit is built to find exactly that, flow by flow.
- How can a Klaviyo flow generate close to zero revenue without anyone noticing?
- The most common cause is two flows sharing the same trigger event. Klaviyo's Smart Sending feature — a safeguard against over-emailing — will suppress a second flow's messages if a customer already received an email from a related flow recently, even if that wasn't the intent. The dashboard for the suppressed flow still shows normal open and click rates on the few emails that do get through; it just doesn't show how few customers are actually reaching it, which is why the problem hides in plain sight.
- Why would an "Abandoned Checkout" flow's value-tier split be set wrong?
- Usually because it was set once, early, based on an assumption or an outdated average order value, and never revisited as the business changed. A threshold set at double the real AOV means the "high value" track — often the one with the strongest incentive — barely gets used, while most traffic goes through the generic low-value path regardless of what they actually almost bought.
- Is it normal for a Klaviyo audit to take more than a month to show results?
- Yes, and it's worth being suspicious of anyone who promises otherwise. Foundational fixes (suppression logic, broken filters, flow restructuring) typically show up in month one; flows that were rebuilt or newly launched need a full send cycle before their numbers mean anything, and multi-language accounts add translation sign-off as another real bottleneck. A two-month case study with some results and some still-pending items is a more honest picture than one that claims everything is solved immediately.
- What's the first thing to check if a flow looks like it's underperforming?
- Whether it shares a trigger event with another active flow, and whether Smart Sending could be suppressing it. It's a five-minute check in Klaviyo's flow settings and it catches a surprising number of "why is this flow so quiet" cases before you touch a single subject line or piece of copy.
Source:
SciGrowth client Klaviyo account audit and two-month implementation, anonymized. Figures rounded; exact client identity withheld.
0 comments