What measurements are necessary for a reliable size recommendation?

It is possible, however, that the reliability of a recommendation must differ clearly from the available data. For categories or models for which only a generic size table is known, it is better to provide limited help and not to display an erroneously accurate recommendation.

How many buyer measurements and previous choices are needed to recommend the size?

Only to the extent that it actually improves the size solution. If the recommendation has enough stature, normal size and desired fit, there is no need to create a comprehensive body profile. Previous purchases can only be used with a clear legitimate basis and benefit to the customer.

Do you measure the total return rate?

The overall indicator is too broad: the impact of a size decision requires a distinction between returns due to an undersized, oversized or otherwise faulty product and analysis of those categories where the recommendation has actually been shown.

How do you distinguish the reason for a customer's return from incorrect model measurements?

Returns are associated with that model and the size for which the recommendation was shown. Recurring non-conformity of one model is passed to the product data check; a single customer experience should not automatically change measurements for all buyers.

For which range might the size recommendation be insufficiently reliable?

For goods and models in a single or highly variable condition, the measurements of which are unknown. In this case, the more accurate presentation and consultation of measurements of the product may be more valuable than automatic sizing.

Apparel size recommendation tool

How the solution works

Checking measurements and shearing information for a specific model
The customer provides the data necessary for the selection
A size recommendation and its reliability limits are provided
Size-related returns are used to assess the quality of a recommendation

Key challenges

The customer struggles to choose the right size online

Solution capabilities

Specific model eligibility data

The recommendation is not based solely on the brand size table, but on measurements of a specific model and known shearing information.

Information provided by the customer

The questions must be as much as is realistically necessary for the size decision, avoiding excess personal data.

Purchase and Returns Signals

Once the client has been identified, previous successful and unsuccessful size choices can be used.

Recommendation and explanation

The client is presented with a size, confidence level, and clear argument, not just an inexplicable automatic choice.

Business context

The customer struggles to choose the right size online: The general size table does not explain measurements and shearing of a specific model, and the recommendation does not show its limits. The customer chooses the wrong size or orders several sizes for comparison.
It's easier for the buyer to decide on size: The uncertainty about how a particular model will fit can stop the purchase or lead to ordering multiple sizes. The explained recommendation helps to choose based on the dimensions of the product and the buyer's data. Its benefits must be assessed together by the completion of the purchase and the reasons for the returns.

Core features

Specific model eligibility data
Information provided by the customer
Purchase and Returns Signals
Recommendation and explanation

Key integrations

Product options information system: Model measurements, size structure and product facts.
E-commerce platform: The product selected by the customer, the submission of a recommendation and the actual ordered size.
Order and Return Data: Actual purchase, reason for return and change to another size.
Client profile when used: Service is allowed to use previous purchases information and customer preferences.

Potential impact (%)

The ranges indicate an illustrative relative change in the metric under the stated assumptions. Results depend on the starting position and actual use of the solution. Percentages for different metrics must not be added together.

Part of returns due to inadequate size or suitability

2–12%Decreasing

This illustrative scenario assumes that 10-30% of the metric is attributable to addressable planning and execution shortcomings. That share is assumed to fall by 20-40%. Company data is needed to verify both the addressable share and the resulting change.

The proportion of returns for size or suitability is compared for the same categories and groups of models before and after the recommendation. The percentage is calculated on the number of units sold in the same group.

Part of orders with multiple sizes of the same model

2–12%Decreasing

This illustrative scenario assumes that 10-30% of the metric is attributable to addressable planning and execution shortcomings. That share is assumed to fall by 20-40%. Company data is needed to verify both the addressable share and the resulting change.

Orders are counted, in which the same model is purchased in several close sizes for the same customer. The percentage is calculated on the number of all confirmed orders of the same type.

Part of the sessions completed with the purchase of clothing

4–19%Increasing

Sample starting portion - 5%. Assumption: 2-5% of the remaining cases relate to the uncertain size of a particular dress pattern; a solution would help solve 10-20% of these cases.

Sessions with a completed clothing purchase are divided by all comparable product selection sessions. Comparisons are made between buyers who have received the recommendation and those who have not received it; returns are checked.

Conditional calculation scenarios. The assumptions have not been validated against client measurements.

When this solution is relevant

A significant portion of returns are associated with inappropriate size or shearing
In e-commerce, customers often buy multiple sizes of the same model
Available measurements of specific models or can be started
There are enough purchases and returns to verify the history recommendation

Implementation requirements

Size recommendations require reliable model measurements and separate recording of the reasons for resizing. The evaluation methodology has to compare the results of similar purchases and the use of the recommendation without transferring a total conversion change to it.

Further development options

The client's selected size available to the consultant
Analysis of discrepancies in different model sizes