Who they are

Etta Loves is a science-led baby brand, best known for its sensory muslins and playmats designed around how a baby's vision actually develops. The products are a firm favourite among new parents - and, just as importantly, among people buying for new parents. It is a brand people discover through a friend, a gift, a new arrival.

That gifting dynamic sits underneath everything in the data, and it shapes how the business needs to read its own numbers. A large share of orders are presents, not self-purchases - which means the people buying are often not the people using, and a "one-time buyer" is frequently a delighted gifter rather than a lapsed customer.

The brand arrived with a rich, healthy dataset and a fair question: with ten years of orders to look at, which patterns still reflect how the business behaves today - and which are simply history?

What we did

A fixed-scope sales and customer insight audit built on the brand's own Shopify exports - orders, customers and products - covering the full trading history from launch to the present day.

Most reporting looks at either all-time totals or a recent snapshot. We deliberately did both, and put them side by side. A decade of data is excellent for the structural truths that need scale to be meaningful - overall loyalty patterns, geography, seasonal rhythm. But products, collaborations and customer behaviour all shift over ten years, so those same long-run averages can quietly mislead when used to decide what to do next month.

To keep the analysis honest, we built the audit around two lenses and labelled every finding with the period it drew on:

• The lifetime view - the big, slow-moving truths across the brand's entire history, where years of data give a reliable picture.
• The last-twelve-months view - what is true right now: current bundling behaviour, today's most valuable customers, the reorder timing that reflects how people actually shop the brand at present.

Where the two lenses agreed, we could state a finding with real confidence. Where they diverged, that gap itself became one of the most valuable parts of the audit - a signal that a long-held assumption had quietly gone out of date.

Every figure was analysed on aggregated data only, with no individual customer names, email addresses or personal details reproduced anywhere in the work.

What we found

All revenue figures, order volumes, customer counts and financial metrics remain confidential. What follows are the strategic findings only.

The best-selling bundle had quietly changed

The single clearest example of why the two lenses matter. Across the full ten-year history, one accessory dominated nearly every basket - it looked, unambiguously, like the product to push. But viewed over the last twelve months, it had dropped out of the top pairings entirely, having migrated to a small checkout add-on. An entirely different set of products now drives multi-item orders. A recommendation built on the lifetime data alone would have pointed the brand at yesterday's winner.

The "buy once" story is more encouraging up close

Measured across all time, the brand looked like most gifting-led businesses: the majority of customers buy a single time. But among customers active in the last year, a materially higher share are repeat buyers. The recent cohort is more loyal than the lifetime average implies - a more encouraging, and more reachable, picture than a decade-wide number suggests.

Rising order values were real, not inflation

Average order value has climbed steadily for years - the kind of trend that is easy to dismiss as simply the effect of price rises. The data showed the opposite. Average price per item had stayed essentially flat, while the number of items per order grew. The uplift was genuine basket growth driven by better merchandising, not customers simply paying more - a far more valuable and more defensible kind of progress.

There is a precise, predictable moment to re-engage

By mapping the time between first and second orders, we could identify exactly when returning customers tend to come back - and, crucially, that there are two distinct groups: an eager set who return quickly, and a longer tail who take considerably more time. That points to a two-stage re-engagement rhythm rather than a single generic reminder, timed to how customers genuinely behave rather than to a marketing convention.

The gift dynamic is an untapped strategy, not just a quirk

Because so many purchases are gifts, every order potentially creates a brand-new person holding the product who may never have visited the site. Recognising gifters and recipients as distinct from self-purchasers - and treating them accordingly - opens up segmentation, packaging and lifecycle opportunities that a single undifferentiated "customer" view completely hides.

A baseline, not just a verdict

Because the audit was built on two lenses, it does more than list opportunities - it establishes a current benchmark the brand can measure against. The lifetime view sets the long-run context; the twelve-month view becomes the live baseline. Run the same analysis again in a year, and progress can be measured like-for-like on the metrics that actually reflect how the business behaves today.

That distinction - history for context, recent data for action - is also the strongest argument for keeping this kind of analysis current rather than one-off. Data shifts. The bundle that led for a decade stopped leading. A snapshot taken at the wrong moment, or read through the wrong lens, can send a good business in the wrong direction.

Who this applies to

Etta Loves sells baby products, but the underlying challenge is common to any brand with real history behind it. If any of the following sound familiar, your all-time reporting may be quietly masking what is true today:

• You have several years of sales data, and you are not sure which patterns still reflect how customers behave now.
• A large share of your orders are gifts, and you report on gifters and self-purchasers as if they were the same audience.
• Your headline metrics look flat or discouraging across all time, even though recent trading feels healthier.
• You have a "best seller" or "go-to bundle" that has been true for years, and you have never checked whether it is still true.
• You want to know not just what happened, but what to do next month - which needs recent data, read on its own terms.

In each case, a single all-time view captures the history but obscures the present - and the present is where the decisions are made.