In nearly every discovery call, prospects tell us they are evaluating WAIR for one of two reasons: to reduce returns or increase conversion.

That makes sense. Those are common ecommerce mandates. But they are incomplete measures of success.

The more useful lens for evaluating almost any e-commerce tool is its impact on kept orders.

‍

What is a Kept Order?

A Kept Order is an order that remains with the customer after the brand’s defined return-observation window has closed.

It is different from a completed order. A completed order measures the checkout event. A kept order measures the portion of that demand that survives the post-purchase experience.

At the order level:

Kept Orders = Completed Orders − Returned Orders

‍

The corresponding rate is:

Kept-Order Rate = Kept Orders ÷ Sessions

‍

To normalize the metric across sites or test groups:

Kept Orders per 10,000 Sessions = (Kept Orders ÷ Sessions) × 10,000

‍

Because kept orders are completed orders adjusted for returned orders, the metric can also be expressed as:

Kept Orders per 10,000 Sessions = CVR × (1 − Order Return Rate) × 10,000

‍

For example, 10,000 sessions that produce 500 orders and 100 returned orders result in 400 kept orders, or 400 kept orders per 10,000 sessions. A second tool might produce 550 orders but 140 returned orders. It would win on conversion while producing only 410 kept orders. The additional 50 orders created at checkout would translate into only 10 additional kept orders.

Kept orders should also be distinguished from kept items:

Kept Items = Purchased Items − Returned Items

‍

The order-level metric asks whether an order remained with the customer. The item-level metric asks how much merchandise remained. Both can be useful, but they answer different questions.

For tool evaluation, the most useful comparison is incremental kept orders against a comparable control group:

Incremental Kept Orders per 10,000 Sessions = [(Test Kept Orders ÷ Test Sessions) − (Control Kept Orders ÷ Control Sessions)] × 10,000

‍

The definition should be consistent across every tool: the same return-observation window, denominator, treatment of exchanges and cancellations, and rule for whether an order with any returned item counts as returned. Used consistently, kept orders give ecommerce teams a way to evaluate whether a tool creates demand that survives the return window.

‍

What the WAIR Data Shows?

To see how WAIR impacted Kept Orders, we ran a Returns Impact Analysis across 30 brands over the last 90 days and found some really interesting stats:

• The WAIR segment produced 581.2 kept orders per 10,000 sessions (compared with 127.7 kept orders per 10,000 sessions when WAIR in the non-WAIR segment.
  
  • That is 453.5 additional observed kept orders per 10,000 sessions, or a 355.1% higher observed kept-order rate.

‍

Kept orders are calculated after accounting for order returns. In other words, the analysis asks a more complete question than “Did the shopper place an order?” It asks whether the shopper kept the order.

That distinction matters because ecommerce teams often evaluate technology through a single metric. A tool is credited for increasing conversion, AOV, revenue, or repeat purchase, while another is judged for reducing returns. These metrics are useful, but each captures only one point in the economic journey... and often tools may positively impact one while detracting from another.

‍

The problem with single-metric evaluation

Conversion rate measures the share of sessions that produce an order. It does not tell the team whether that order was returned.

AOV measures the value of the initial basket. It does not show the value that remains after refunds, returned units, discounts, and return-related costs.

Return rate measures returned orders or units. It does not show how many shoppers who would otherwise have abandoned were converted into additional kept orders.

Revenue per visitor is more comprehensive because it connects revenue to the full visitor population. Profitability per visitor goes further by accounting for the costs required to generate and fulfill that revenue. Additional kept orders per 10,000 sessions adds another useful lens: it shows whether a tool is creating more completed purchases that survive the return window.

Together, these metrics help answer the question that single dashboards often leave unresolved:

Is this tool creating profitable demand, or simply moving one part of the funnel in the right direction?

‍

How to evaluate the major ecommerce tool categories

Apparel companies use a wide range of tools to improve site performance. Each can be evaluated with the same broader scorecard.

Try-before-you-buy platforms (ex. BlackCart and TryNow) reduce purchase risk by allowing shoppers to experience products before deciding what to keep. Their benefit is demand creation. Their tradeoff is structural: products are shipped before the final decision, which creates additional fulfillment and return activity. Evaluate them on kept revenue and profitability, not only conversion or gross items shipped.

Reviews, ratings, and UGC platforms (ex. Bazaarvoice, PowerReviews, Yotpo Reviews, and Okendo) add social proof and fit context. Their impact can be valuable when shoppers find relevant feedback, but reviews are rarely personalized to the current shopper’s body and the specific product. Measure whether they produce more kept orders, not just more clicks or orders.

Search, recommendations, and personalization platforms (ex. Nosto, Constructor, and Rebuy) help shoppers find relevant products, complementary items, or personalized assortments. They can improve conversion and AOV, but product relevance does not necessarily establish size or fit. Their broader site value should be evaluated after returns.

Loyalty and retention platforms (ex. Smile, Yotpo, and LoyaltyLion) can increase repeat purchase, purchase frequency, AOV, and lifetime value. Their effect is usually downstream. A stronger evaluation connects future purchases to whether the first order was kept.

CRO, promotion, checkout, payment, and messaging tools can reduce transactional friction, improve affordability, or create urgency. They may increase conversion quickly, but the order is counted before the brand knows whether the product worked. Revenue per visitor and profitability per visitor reveal more than conversion alone.

‍

Where WAIR stands out

WAIR belongs in this broader site-value conversation because it influences the purchase before fulfillment and measures the outcome after fulfillment.

The mechanism is size confidence: helping a shopper answer, “Which size is most likely to fit me in this product?” The business outcome is kept revenue: more of the demand a site creates surviving the return window.

That makes WAIR more than a fit recommendation feature. It is a way to connect shopper body data, product information, size recommendations, purchases, returns, reviews, and fit feedback to the metrics ecommerce leaders ultimately care about: kept orders, kept revenue, and profitability per visitor.

The right question for every ecommerce tool is not simply, “Did it increase conversion?” It is: Did it create more profitable demand that customers kept?

‍

[Learn more about using WAIR to connect size confidence with kept-order performance.]

‍