Barton Springs Mill · Fluid IQ Digital
Klaviyo earned $454,841 in the last twelve months. Sending has been stopped for a week, two automations were never switched on, and the welcome series is reaching about one in twenty new subscribers.
Email is one of the largest revenue channels on this account — $454,841 over the last twelve months by Klaviyo's own attribution, across 171 campaigns and five live automations. It is also the channel getting the least attention, and three specific things are leaking money right now.
Note on the revenue figure: $454,841 is Klaviyo's own attribution, which uses a wider window than Google Analytics. It is the right number for comparing emails against each other, which is how it is used throughout this document. There is a separate tracking issue we're handling on our side that currently prevents analytics from seeing the channel at all — it doesn't change any of the findings below.
Campaign volume has been stable for over a year — 11 to 20 a month, every month. Then nothing between 5 and 12 August. Sending restarted this afternoon with two recipe emails, but the three campaigns written on 7 August have now sat for five days.
| Campaign | Written | Status |
|---|---|---|
| 8.9 sourdough | — | Sent 5 Aug |
| — seven days, nothing sent — | ||
| 8.15 pasta fredda | 12 Aug | Sent 12 Aug |
| 8.16 baguettes | 12 Aug | Sent 12 Aug |
| August Mill Digest | 7 Aug | Still draft |
| August BOGO Class Sale | 7 Aug | Still draft |
| August BOGO Class Sale Ending | 7 Aug | Still draft |
The recipe emails move; the promotional ones don't. The average campaign here is worth about $2,280 — the best ones clear $15,000 — so a week of silence is not a rounding error.
The account is being actively worked in. Two of these were still in draft when this review began and were sent while it was being written.
July: 0.021% spam complaints and 0.37% bounces across 52,944 delivered. Both are comfortably inside safe limits — the practical ceiling for complaints is 0.10%.
That matters because a list which has been dormant for months needs a slow, staged warm-up before you can safely send to it again. A seven-day gap on a list this healthy does not. The drafts can go out immediately.
Short answer: they were never switched on. Nothing paused them. Both were built during the original 2022 setup, never finished, and never touched again.
| Automation | Recipients | Orders | Revenue | Per person | Last edited |
|---|---|---|---|---|---|
| Abandoned Cart Reminder | 17,754 | 365 | $32,505 | $1.83 | Jun 2023 |
| Browse Abandonment | 5,526 | 202 | $20,541 | $3.72 | Feb 2025 |
| Customer Thank You | 17,180 | 132 | $9,796 | $0.57 | Jun 2022 |
| Welcome Series | 645 | 27 | $2,082 | $3.23 | Sep 2024 |
| Customer Winback | — | 0 | $0 | — | Oct 2023 |
| Customer Thank You – New vs. Returning | never launched | Jun 2022 | |||
| Test flow at 5 orders | never launched | Jun 2022 | |||
Trailing 12 months, Klaviyo attribution. Total across all automations: $64,924.
Roughly 11,100 people subscribed to email marketing in the last twelve months. The welcome series sent to 645 of them — about one in twenty.
That is the most expensive thing in this document. The welcome series earns $3.23 per recipient, the second-best rate of any automation you run. At the rate it already performs, reaching everyone who signs up would be worth roughly $36,000 a year — from an email that is already written.
Likely cause: the flow triggers on being added to one specific list, and most sign-ups are landing somewhere else — there are 203 lists on the account, mostly one-off class sheets. Worth checking which list it watches before assuming anything, but the size of the gap points strongly at this.
It shows as running, and it returned zero attributed revenue and zero orders. Either almost nobody meets its entry condition, or the condition is written in a way that never matches.
A live automation producing nothing is worse than a switched-off one, because everyone assumes win-back is covered. It isn't. This should be rebuilt around each customer's own repurchase interval rather than a fixed window.
Neither is a fault and neither was switched off by anyone — they simply never launched, and nothing has touched them in four years. The more useful finding is the one next to them: "Customer Thank You" has not been edited since June 2022, and it is the worst performer you have at $0.57 per recipient. The unfinished upgrade sitting beside it was the fix.
Abandoned cart, browse abandonment, thank-you, win-back and welcome. What is missing matters more than what is stale — see the recommendations below, particularly replenishment, which does not exist at all and is the natural fit for a mill.
Campaigns sent per month
August 2026 is 1–12 Aug (partial, shown hatched).
Revenue by automation — trailing 12 months
Klaviyo attribution. Two automations never launched; one is live and earning nothing.
The welcome series gap
New email subscribers vs. people who actually received the welcome series, trailing 12 months.
In order. The first two are worth more than everything below them combined, and neither needs new creative.
The Mill Digest and the BOGO sale have been finished since 7 August. The list is healthy and other campaigns went out today, so there is no technical reason these can't follow. The BOGO sale is time-bound and has already lost five days.
Check which list the series watches and widen it so every new subscriber enters, regardless of where they signed up. Worth roughly $36,000 a year at the rate it already performs, using an email that already exists.
Winback is live and earning nothing — rebuild it to trigger on each customer's own repurchase gap rather than a fixed window. And the New vs. Returning version of Customer Thank You is already built apart from its email; finishing it replaces your weakest automation ($0.57 per recipient) with one that treats first-time and repeat buyers differently.
Flour is a consumable bought on a predictable rhythm, and there is no automation reminding anyone to reorder. This is the single biggest gap in the automation set and it fits this business better than any other tactic in this document. Trigger on each customer's typical interval, not a fixed 30/60/90 days.
The affinity segments most brands never build already exist here and are barely used:
| Segment | People | What to do with it |
|---|---|---|
| All Profiles — Newsletter | 41,785 | Main send audience — and big enough for real A/B testing |
| sour dough | 11,386 | Lead with sourdough content instead of sending everyone the same digest |
| 100 Miles of Mill | 7,906 | Local — classes, tours, events, pickup |
| pizza geeks | 4,477 | Pizza flour launches and recipes |
| Previous class / tour purchasers | 4,262 | The obvious audience for class sales |
| Placed 5+ orders | 3,738 | VIP — early access rather than discounts |
Two worth adding: lapsed / at-risk (bought before, nothing in twice their usual gap) and one-time buyers (the biggest single revenue lift available on a food brand is turning a first order into a second).
With 41,785 on the main newsletter segment you can split roughly 20,000 per arm, which is enough to detect real differences. Most brands can't test properly and have to guess. Test one thing at a time, in this order:
| # | Test | Why here |
|---|---|---|
| 1 | Subject line | Biggest swing, cheapest to produce two of |
| 2 | Send day & time | Free, and often a bigger lift than new creative |
| 3 | Offer framing | Free shipping vs % off vs bundle — moves revenue, not just opens |
| 4 | Primary call to action | Turns the open into a click |
| 5 | Template / design | Where the redesign gets proven — see below |
Two rules that keep the answers honest: decide the finish line before you send (24 hours for opens, 72 for revenue — checking early and stopping when one side looks ahead invents winners that don't repeat), and judge on revenue per recipient, not open rate. Your open rate reads 78% because Apple Mail opens messages automatically on behalf of a large share of subscribers. It's a tiebreaker, not a verdict.
Worth doing. Three things worth building into it:
Recipe emails went out today; the Mill Digest and the BOGO sale from 7 August didn't. Knowing whether that's an approval, a date or a pricing decision tells us whether we can release them now.
This is the $36,000 question. If sign-ups from the site are landing on a different list than the one the series triggers on, that's a one-setting fix.
Zero revenue in twelve months. Useful to know whether it ever produced, or whether the entry condition has always been too narrow.
All of it builds on data Klaviyo already holds. We'll bring the replenishment automation and a first subject-line test as soon as the drafts are moving.