All posts

Eoin Cambay

Increased conversions or reduced returns?

A successful AI fitting room impacts two key metrics. First, it reduces the proportion of items that are returned because they didn’t fit as expected. Second, it increases conversion rates by addressing the doubt that every online shopper feels when choosing a size: “Which size will fit me how I want it to?”

But improvements to one of these metrics will make a much bigger difference to your revenue than improvements to the other – and it’s not the metric that often gets the most attention.

A typical scenario

Let’s assume an online store has:

  • 100,000 monthly visits
  • a conversion rate of 2.7% (global average for fashion ecommerce as of Q4 2022)
  • an average garment value of just €30
  • an overall returns rate of 26% (2021 average, according to Statista)
  • 52% of these returns are for reasons of wrong size/fit (Shopify, Aug 2021)
  • and an average cost of processing a return is €3

This scenario yields 2,700 sales per month. But:

  • 702 of these sales (26%) come back as returns, of which
  • 365 (52%) are for size/fit reasons.

This leaves us with net sales of 1,998 units which, at €30 each, yields €59,940 in revenue. However, once the €2,106 of direct costs of processing returned garments is taken into account, this reduces to €57,834.

This is our basic, no-AI-fitting-room scenario.

The impact of increased conversions

Let’s assume that:

  • a modest 30% of shoppers use your AI fitting room to help determine their size; and that
  • using your virtual fitting increases their conversion rate by 50% (i.e., from 2.7% to 4.05%).

In our scenario, ceteris paribus, total revenue rises to €66,509 - a 15% increase.

Note: there is evidence that AI fitting rooms improve conversion rates for shoppers that use them by 250% or even 300% - far higher than the 50% we have used in our scenario!

The impact of reduced returns

Let’s assume:

  • a modest reduction in returns for reasons of wrong size/fit of 36%; and that (as above)
  • 30% of shoppers use your AI fitting room to help determine their size

In our scenario, ceteris paribus, the overall returns rate falls from 26% to 21% and total revenue (still accounting for the costs of processing returns) rises to €59,135 – a 1% increase.

The value of reducing returns increases dramatically if your costs of processing is a higher proportion of the order value. For example, if your AOV is €89 but you spend an average of €11 processing each return, then the 1% increase in monthly revenue described above becomes a 5% increase in monthly revenue.

Note: studies suggest that a AI fitting room can reduce returns by up to 50%, considerably higher than the 36% we have used in our scenario.

So, which should be your priority?

Very clearly, the impact of increased conversions far outweighs the impact of reducing returns. In reality, however, your AI fitting room provides both benefits.

In our scenario above, a 50% increase in conversion rates for virtual fitting room users, and a simultaneous reduction of 36% in “wrong fit” returns, yield:

  • a reduction in the overall returns rate from 26% to 21%; and
  • a rise in monthly revenue of over 18%

If you’d like to plug your own figures into our model, do get in touch. We’d be happy to show you the maths.