looking is learning — so is your data.
a first read of five years of orders, customers and products — fifteen findings on where brand x is quietly winning, and where there’s revenue waiting to be picked up.
five years of compounding growth
across the full history, brand x has taken 30,018 paid orders generating £4.34m in revenue, from a customer base of 13,000 people. the last twelve months alone brought in £1.5m across 10,417 orders.
the headline story is encouraging: revenue has grown every single year, from £326k in 2020 to £1.5m in 2024 — a 4.6× increase. unlike many brands at this scale, growth is broad-based: order volume, customer count and basket composition are all moving in the right direction at once. the opportunities below are about doing more of what’s already working, with very little extra effort.
reading this report: two lenses. five years of data tells two kinds of story. each finding is labelled lifetime (the structural truths — scale, loyalty, seasonality — that need years of data to be meaningful) or last 12 months (what’s true now — current bundling, recent customer value, today’s reorder timing). where they differ is often the most interesting part.
sales & customer findings
fifteen findings drawn from the order, customer and product data. each carries a recommendation — the specific, low-effort action the data points to.
your top 10% of customers fund a third of the business
the most valuable 1,300 customers have each spent an average of £1,201 — and together account for 36% of all revenue. the remaining 90% average £215. this is a healthy concentration for fashion: a loyal core returns season after season while a wider base buys once or twice.
that top tier is small enough to treat specially — early access to new collections, a thank-you with their next order, or a simple “refer a friend” ask. protecting and growing this group is the single highest-leverage thing the data points to.
multi-item orders are worth 151% more
single-item orders average £94.51. orders with two or more items average £237.70 — a 151% uplift. yet only 34.9% of orders currently contain more than one item.
- classic white tee + cotton canvas tote
- classic white tee + stripe breton top
- classic white tee + oversized linen shirt
- cotton canvas tote + stripe breton top
- classic white tee + cotton canvas tote
- classic white tee + stripe breton top
- classic white tee + oversized linen shirt
- classic white tee + relaxed chino trousers
the tee is the gateway product — it anchors nearly every top pairing. surface the tote, the breton top and the linen shirt as “complete the look” add-ons on the tee product page. nudging the multi-item rate from 35% toward 45% has an outsized revenue effect, because the second item more than doubles the order.
knitwear and outerwear carry the value; tops carry the volume
the top products by revenue are dominated by the higher-priced knitwear and bottoms lines. the ribbed knit jumper alone has generated £417k. but the tops range moves the most units — it is the volume engine that brings customers in.
the longline wool coat converts far less volume (1,433 units) but carries a much higher unit price (£298 vs £45 for a tee) — it’s an under-exploited aov lever. featuring outerwear as the natural “step up” from tops and knitwear, especially to returning customers, is a clean upsell path.
half of customers come back — strong for fashion, with room to grow
50.6% of customers have placed more than one order. for an apparel brand this is a healthy repeat rate — well above the gifting-led pattern seen in many categories — but it is still the clearest growth lever in the dataset, because acquiring these customers has already been paid for.
the loyalist segment (1,642 customers) averages 5.1 orders each and an LTV of £917 — nearly 8× the £116 LTV of a one-time buyer. moving even a small share of one-time buyers into the repeat column compounds powerfully.
the range already spans entry-level tees through to premium outerwear. the opportunity is a structured “next purchase” journey — when a customer buys a tee, they’re a natural audience for the pieces that pair with it. see the reactivation window in finding 12.
aov is stable — the growth is all volume
aov has held remarkably steady between £142 and £148 across all five years. crucially, this is not price inflation masking a decline — it’s consistent basket behaviour at scale. the entire revenue story is being driven by reaching more customers, not by charging existing ones more.
flat aov alongside strong volume growth is the clearest signal in the data: the brand has cracked acquisition but has not yet pulled the basket-size lever. that’s the opportunity finding 2 quantifies — and it’s pure margin, because the traffic is already arriving.
november and december are the peaks; mid-summer is the trough
aggregating every year, demand builds through spring (mar–apr), dips through mid-summer, then climbs to a pronounced gifting peak in november and december. july and august run roughly 40% below the november high.
stock and campaign planning should lean into the spring (mar–apr) and peak (oct–dec) windows. the july–august trough is an opportunity: a deliberate summer campaign, or using the quiet period for operational work that’s harder to run at peak.
bottoms lead revenue; tops lead units
the catalogue splits cleanly into a volume tier (tops, accessories) and a value tier (knitwear, outerwear). bottoms sit in the middle and lead overall revenue. understanding which tier a customer enters through is the foundation of the cross-sell logic in part two.
accessories generate the most units after tops but the least revenue — they’re a natural low-friction add-on rather than a destination purchase. position them at cart level to build baskets rather than as standalone hero products.
five behaviours, five different responses
the 13,000 customers don’t behave uniformly. the order history splits them into five groups with materially different value, frequency and risk profiles — each of which warrants a different treatment.
the discount-led group is the one to watch: 2.4 orders each but an LTV barely above occasionals, because the margin is being given away. the lapsed group is the opportunity — high historical value (finding 9). treat each segment as a distinct audience rather than emailing all 13,000 the same thing.
560 customers spent £595 each, then went quiet
the lapsed segment placed an average of 3.6 orders before drifting away more than 18 months ago. their combined historical spend is £332,922. these are not cold prospects — they know the brand, they bought repeatedly, and reactivating them costs a fraction of acquiring someone new.
a targeted three-email reactivation sequence to this list, sent outside peak season to avoid cannibalising full-price demand, is the lowest-effort revenue in the dataset. at a 15% response rate this is £12k+; at 25% with a two-order assumption it exceeds £40k. zero acquisition cost.
discounting is controlled — but a dependent core is forming
15.7% of orders use a discount code, spread fairly evenly across six codes — healthy, with no single code dominating. the concern is the 1,965 discount-led customers (finding 8) who rarely buy at full price. they’re engaged, but their margin is structurally lower.
worth testing whether the discount-led segment is genuinely incremental (winning hesitant buyers) or simply discounting people who’d have bought anyway. a controlled test — varying the offer for one segment — would show whether that revenue could carry a higher margin. a clear margin question worth answering.
1.52 items per order — the headroom is in the second item
the average order contains 1.52 items, and this figure has barely moved in five years. the bimodal pattern from the bronze analysis holds: most orders are a single item (often a tee or accessory), with a smaller band of genuine multi-item baskets. the gap between 1.52 and a realistic 1.8–2.0 is the single biggest structural opportunity.
items-per-order is the lever beneath the aov. because price-per-item is stable, the only way to grow aov is to grow basket size. every finding about bundling, cross-sell and category pairing ultimately points back to this one number.
when repeat buyers come back — and when to nudge
among customers who return, the median gap between first and second order is 164 days. 32% reorder within 90 days, 54% within six months, 84% within a year. this is the single most actionable timing insight in the report.
the fastest quarter of returners come back within 66 days — the keen early repeaters — while the median is pulled out to 164 days by a longer tail. the practical read: the first nudge should land early (around 60–90 days) for the eager group, with a second wave around 150–180 days for everyone else.
a well-timed email sequence — a first touch around 75 days, a second around 160 — targets customers exactly when the data says they’re most receptive. this directly attacks the 49% one-time-buyer rate and turns it into the structured journey described in part three.
returning customers quietly underwrite each year
although new-customer acquisition drives the headline growth, returning customers contribute a disproportionate share of revenue relative to their numbers — the loyalist and occasional segments together (5,939 customers, 46% of the base) account for roughly two-thirds of lifetime revenue. the business is more retention-dependent than the acquisition story suggests.
this is the flip side of the growth narrative: the brand is doing well on acquisition, but its revenue base rests on the returning core. a small slip in retention would be felt immediately. protecting the repeat rate (findings 1, 4, 12) is defensive as well as offensive.
the ability to reach customers is the multiplier
almost every recommendation above — reactivation, next-purchase journeys, loyalty rewards, the summer-trough campaign — depends on being able to reach customers directly. the size of the marketing-consented list is therefore the enabling constraint behind the entire plan. for a brand with a 50% repeat rate, every point of consent gained makes all the other levers work harder.
prioritise consent capture at checkout and post-purchase with a soft, on-brand opt-in. this is the unglamorous enabling move: without a reachable list, the reactivation window and next-purchase journey stay theoretical.
the natural path: tee → accessories → knitwear → outerwear
pulling the threads together, the data describes a clear customer journey. people enter through the tee (the hero, accessible-price product), add a tote or breton top in the same basket, and — for the most valuable customers — step up to knitwear and the higher-value outerwear pieces over time.
entry: classic white tee, stripe breton top (volume tier)
same-basket add-ons: cotton canvas tote, oversized linen shirt, leather belt
step-up / returning customers: ribbed knit jumper, merino cardigan, relaxed chino trousers
premium / loyalists: longline wool coat, quilted gilet, dresses
designing the site’s cross-sell and email flows around this exact sequence — rather than generic “you may also like” — would align merchandising with how customers already behave. the path is in the data; it just needs to be made deliberate.
in summary
website opportunities
the data tells us what customers do. a quick review of the store shows where the site could make it easier for them to do more of it. each theme connects directly to a finding above.
this is a qualitative review of the public-facing store. the highest-value next step would be pairing it with on-site analytics — conversion rates, drop-off points, device split and page speed — which would turn these themes into measured, prioritised fixes.
make the “complete the look” saving impossible to miss
given that multi-item orders are worth 151% more, any bundle or set saving is one of the strongest purchase nudges available. if it sits in body copy below the price, it’s doing its work quietly.
use the shipping threshold as a basket-builder
with aov at £144 and free shipping at £100, most orders already clear the bar — but single-item orders (avg £94.51) sit just below it. a customer at £94 is one accessory away from qualifying.
surface the products that sell together now as add-ons
finding 2 shows today’s natural pairings are the tote, breton top and linen shirt alongside the tee. these aren’t currently presented as obvious add-ons, so customers find them on their own. there’s an execution question worth naming: choosing an upsell mechanism that looks native rather than bolted-on.
bring email sign-up out of the footer
almost every growth lever depends on being able to reach customers (finding 14). if the newsletter sign-up lives in the footer, few people see it.
build the “what’s next” path on-site
the data found a precise reactivation window (finding 12) and a clear step-up sequence (finding 15). the site has the category structure, but nothing actively guides a tee buyer toward the knitwear and outerwear that returning customers gravitate to.
segmentation & retention
the sales data shows where the opportunities are. this section is about who to focus on and how to keep them — turning the five segments from finding 8 into distinct, actionable journeys.
give high-value customers a reason to feel valued
the top 10% drive 36% of revenue (finding 1) and the loyalist segment averages £917 LTV (finding 8) — yet they likely receive the same broadcast emails as everyone else. no personalised flows, no recognition, no referral mechanic. for a group this valuable, that’s a significant gap.
turn first-time buyers into a lifecycle journey
4,536 customers have ordered exactly once. the reactivation window (finding 12) tells you precisely when they’re most receptive to a second purchase. capturing the entry category lets you tailor the next-stage prompt to what they actually bought.
recover £333k of dormant value
the 560 lapsed customers (finding 9) bought repeatedly before going quiet. they’re the warmest audience the brand has that isn’t currently buying.
protect margin without losing the customer
the 1,965 discount-led customers (findings 8, 10) are engaged but structurally lower-margin. the question is whether their discounting is winning incremental sales or simply eroding margin on sales that would have happened anyway.
turn the quiet months into a planned moment
july and august run 40% below the november peak (finding 6). rather than accept the dip, it can be planned for — a deliberate summer edit or a loyalty-only early access using the reachable list.
technical & discoverability
beyond sales strategy and on-site experience, there’s a layer of technical groundwork that quietly governs how easily Google — and increasingly, AI shopping tools — can find and recommend brand x products. largely invisible to customers, but directly affecting free traffic and shopping visibility.
this is an external review based on the public store and product export. a full audit — with access to Google Merchant Center, Search Console and the live product feed — would confirm exact error counts and let each fix be measured. that access is the natural first step of any engagement.
add GTINs so Google can match and surface products
for branded, barcoded apparel, Google expects a GTIN on each variant. missing identifiers are a leading cause of products being suppressed or disapproved in Shopping. where a GTIN is genuinely absent, the correct signal is to mark identifier_exists = false rather than leave the field blank.
check the product schema isn’t quietly failing
a common Shopify theme issue is product structured data (JSON-LD) emitting empty GTIN values, which Google reads as “identifier claimed but invalid” — triggering bulk disapprovals that look like a feed problem but are really a code problem.
keep feed prices in lock-step with the site
with promotions running (the six discount codes from finding 10), there’s a real risk of the feed price drifting from the landing-page price — one of the most common disapproval triggers.
restructure product titles for how people search
Google Shopping rewards brand + product type + key attributes (material, colour, fit). a title like “brand x ribbed knit jumper — merino wool, oatmeal” widens the range of searches a product can appear for — achievable in the feed without changing the friendly on-site display names.
use precise taxonomy and apparel attributes
apparel benefits from an accurate Google product category plus size, colour, gender and age_group attributes. getting these right sharpens how Google matches products to searches — a small, finite data task with a clear visibility payoff.
~6/10 — a healthy site with fixable gaps
based on this external review, an indicative discoverability health score sits at roughly 6 out of 10 — a fundamentally sound store (strong brand, clean catalogue) held back by fixable technical gaps, chiefly product identifiers and feed setup. a full audit with Search Console and Merchant Center access would turn this into a measured baseline to improve against.
this is a full silver example, built on a synthetic dataset. the real thing reads your own numbers - same depth, your store.
book the silver audit →