Three patterns in this data are costing $475K a year.

Three patterns are quietly draining about 6% of revenue. None of them shows up in a standard dashboard — which is exactly why they persist.

Pattern one — paid acquisition funds customers who never become profitable

One in three acquired customers — the lowest first-order tier — generates $53 in lifetime revenue on average. At an industry-typical $40 CAC, 98% of them fail to cover acquisition. Roughly $300K of the total.

Pattern two — the 3% repeat rate is the wrong number to track

The right number is the revenue weight behind it, and the unmeasured upside of moving it. Even a one-to-two point lift produces $73K–$250K a year.

Pattern three — freight is silently eating margin

About 16% of orders carry freight costs exceeding half the product price. Depending on absorption policy, that's $55K–$95K a year in recoverable margin.

A note on this sample. The analysis uses a representative DTC dataset of 96,478 orders over 24 months. Specific category and freight patterns would shift on a real Shopify store. The framework — first-order behavior as the dominant LTV signal, revenue-weighted retention, mix-shift detection, low-intent acquisition — applies to any consumer or digital business of similar scale.

Revenue grew for 24 months. The repeat rate never moved.

Solera Co. sells home goods through Shopify across 70+ product categories. No single category exceeds 9.2% of revenue; the top five combine to just 39%. Most DTC brands at this scale have one hero category driving 25–40%. Solera doesn't — its flat mix reads more like a curated marketplace than a focused brand, and that structural fact shapes every finding below.

The customer base

93,358 buyers. 90,557 of them — about 97% — placed exactly one order. The remaining 2,801 averaged 2.11 orders, yet spend a smaller basket per order than first-timers ($146 vs $161). The repeaters are valuable because they return, not because they spend more. Most of the revenue is single-transaction, and most of the marketing that produced it is being asked to compound into retention — and it isn't.

One observation worth holding onto

The largest revenue event in the dataset — Black Friday 2017 — produced a sharp spike with no sustained lift to baseline afterward. Acquisition events don't compound into retention here. The rest of the report quantifies why.

The Repeat-Rate Trap

Finding  The repeat rate is 3%; repeat customers generate 5.6% of revenue. More striking — their per-order basket is smaller than one-timers', $146 versus $161. They're valuable for coming back, not for spending more when they do.

Evidence  Of 93,358 customers, 90,557 placed exactly one order. AOV by acquisition cohort is stable across the whole window — $146 to $175, inside normal variance. Frequency, not basket size, is the lever.

Action  Track revenue-weighted repeat rate weekly. Run a 30/60/90-day reactivation campaign aimed at the 90,000+ customers who ordered exactly once.

What Actually Predicts Lifetime Value

The obvious analysis would show small-electronics customers worth $368 and telephony customers worth $91 — a 234% gap — and conclude you should shift acquisition toward the high-value categories. The obvious recommendation would be wrong.

Repeat rates are virtually identical across every category — 1.02 to 1.07 orders per customer. Categories don't differ in whether they produce loyal customers. They differ in what those customers spend on the first order. The category-LTV gap is just first-order price wearing a costume.

The data is clear about which dimension matters. Category accounts for 13% of the variation in customer value. First-order spend accounts for 26%. Combined, they explain 37%. Category is the weaker predictor by a factor of two.

Action  Treat first-order value as the primary acquisition-quality signal — ahead of channel, audience, or category. Re-segment paid channels by the first-order AOV they actually produce.

What You Should Be Watching

Three standard early-warning metrics — 60-day cohort retention, first-order AOV trend, and tertile mix — all show stability or mild improvement across the last 18 months. None of the movements is statistically significant. All sit inside normal variance.

The absence of a trend is not permission to ignore. It means there's no system to detect deterioration when it begins — no dashboard light that turns from green to yellow. Whatever breaks the current acquisition pattern (creative fatigue, an algorithm change, channel saturation) won't show up in the P&L for 60 to 90 days.

A note: this sample uses Brazilian e-commerce data, where 2017–2018 inflation ran 3–4% annually. Some of the nominal +4.3% AOV trend reflects price-level effects. On U.S. Shopify data, this confound wouldn't apply.

One in Three Customers Loses You Money

There's a specific activity destroying margin: paid acquisition of low-intent customers. 98% of customers in the lowest first-order tier — first order $73 or less — fail to generate enough lifetime revenue to cover acquisition at any plausible CAC. The math is not marginal.

31,089 customers — exactly one-third of the base — fell into this tier. Their average lifetime revenue is $53. At a 30% gross margin and a typical $40 CAC, each needs $133 of revenue to break even; 99% fall below that. Even at an aggressive $30 CAC, 98% stay underwater.

Why it persists  Founders measure CAC against blended AOV, which looks healthy at $160. But low-tertile customers generate one-third the average lifetime revenue. The blended number hides the problem because the high tertile ($332/customer) mathematically masks the low ($53/customer). It's the wrong denominator.

Action  Within 30 days, audit paid acquisition by landing-page first-order AOV cohort. Pause or restructure campaigns producing more than 40% low-tertile customers — typically broad-audience promos, off-brand creative, price-led offers. Reinvest into channels that produce mid-and-high tertile customers.

Three actions, in priority order.

Action 1 is the primary lever — start with its 7-day audit step. It bundles the levers from Insights 2 and 4 into a single acquisition restructure to avoid double-counting.

Built by people who read the data, not just chart it.

Real Analytical delivers decision-ready analytics for businesses without a data team. You send your data; we send back a written report identifying what's actually driving your growth and what to do about it — the two or three patterns that move the P&L, not a dashboard of forty metrics.

The framework behind this report — first-order behavior as the dominant LTV signal, revenue-weighted retention, mix-shift detection, low-intent acquisition — is applied to every engagement, and it holds across any consumer or digital business of similar scale.