Size recommendation: payback of the tool
Calculates whether an online size-recommendation tool (a widget, a measurement-based hint, a size chart) pays back: how many size returns it removes, what one return costs, the net effect per month and the payback of the launch cost. Shows the reduction in returns at which the tool breaks even.
At a glance
- What question does it answer?
- Whether a size-recommendation tool pays back and at which reduction in returns it breaks even.
- What data is needed?
- Online units per month, the share of tool users, the share of returns for the reason “size” without and with the tool (from a pilot).
- How to read the result?
- Look at the break-even reduction in returns: if the pilot showed more, the tool pays back. The sensitivity table answers what happens if the effect turns out a quarter or a half below the pilot. Example on the default values: Units with the tool per month — 3,600 pcs; Size returns removed per month — 180 pcs; Saving per return — RUB 605; Saving per month — RUB 108,900.
Why and how to reason
A size tool is paid for out of return savings, so the formula starts from returns: how many it removes and what each one costs. Units with the tool = sales × coverage (12,000 × 30% = 3,600). Returns removed = those units × the drop in size-return share (14% − 9% = 5 points), giving 180 per month. One return costs more than handling (RUB 380): part of returned goods cannot be resold at full price, a separate loss (25% × 900 = RUB 225). So RUB 605 per return and RUB 108,900 saved per month. Against that stand costs: 60,000 subscription + RUB 5 × 3,600 units = RUB 78,000. Net effect RUB 30,900 a month, and the RUB 240,000 launch pays back in 7.8 months. Break-even reduction is 3.58 points: below it the tool loses money.
When to use it
Use it for online apparel and footwear sellers where size returns are a visible cost line. You need an honest base: the share of size returns before the tool on the same range and a reasonable effect assumption, best taken from a pilot on part of traffic. It does not fit if returns are mostly about quality or color, if the range is mostly size-free accessories, or if coverage is a few percent: the fee will not pay back. Remember the drop in returns is not always caused by the widget: a change of season or range also moves it.
How to read the result
The key numbers are monthly net effect and payback of the launch. Under 6–9 months with a stable effect the decision is usually sound; over a year or a negative effect means price or coverage is wrong. The 75% and 50% pilot scenarios show the safety margin: if a month turns negative at half the effect, risk is high. Compare break-even reduction with what the vendor promises: if it claims 5 points and you need 3.6, margin is thin. After deciding, track not only return share but also turnover: the tool may have put some buyers off. Ask the tool vendor for pilot data on your own range and compare your estimate with it: the gap between promise and pilot is usually the costliest error in this calculation.
Common mistakes
Typical errors: taking the effect from a brochure instead of your own pilot; omitting the per-unit fee (only the subscription); entering return cost without the loss on unsellable goods; assuming 100% coverage when only a minority use the tool. Also comparing before and after shares across different seasons. If the subscription includes unlimited use, do not add the per-unit fee a second time.
What data is needed?
- Online units per month, the share of tool users, the share of returns for the reason “size” without and with the tool (from a pilot).
- Cost of one return, the share of unsellable returns and the loss on such a unit.
- Monthly fee, fee per unit and the one-off launch.
Inputs
| Field | Unit | Value in the example | Meaning |
|---|---|---|---|
| Units sold online per month | pcs | 12,000 | All units of the online channel in a month where the tool could work. |
| Share of units where the customer uses the tool | % | 30 | From the widget statistics: purchases after opening it divided by all purchases. Not every customer uses the tool, so the effect is counted only on this share. |
| Share of size returns without the tool | % | 14 | From your return-reason records (“wrong size”) in the control group without the tool. There is no single standard in Russia: enter your own data. For footwear and dresses online the return share is usually above the store average (trying several sizes); for children’s it depends on the child’s height, accessories have no sizes. We give no guide without your own data: take your own statistics by category and channel. |
| Share of size returns with the tool | % | 9 | Measure it in a pilot: some customers get the tool at random, some do not. Comparing those who opened the widget themselves with the rest overstates the effect: such customers are more careful anyway. |
| Cost of one return excl. the goods | ₽ | 380 | Return shipping, receiving, inspection, repacking, payment-return fee. Excl. VAT. |
| Share of returns that cannot be resold at full price | % | 25 | A creased item, signs of try-on, a missed season. Empty: all returns are assumed to resell. |
| Loss on such a unit | ₽ | 900 | Cost minus the price at which you still sell such a return (excl. VAT). Empty: 0. |
| Monthly fee | ₽ | 60,000 | Per the contract with the tool vendor, excl. VAT. Empty: 0. |
| Fee per unit with the tool | ₽ | 5 | If the vendor charges for each purchase with the tool. Empty: 0. |
| One-off launch cost | ₽ | 240,000 | Integration, size charts by style, team training. Empty: 0, the payback is zero. |
| Period in which the launch must pay back (optional) | months | — | Your decision. There is no single standard: leave empty if you have no threshold. |
How it is calculated
| Step | Formula |
|---|---|
| Units with the tool | units sold × share of tool users |
| Returns removed | units with the tool × (share without − share with the tool) |
| Saving | removed × (return cost + unsellable share × loss) |
| Break-even reduction | monthly cost ÷ saving per return ÷ units with the tool |
Worked numbers
Default values from the app; in the interactive version you enter your own.
- Units with the tool per month
- 3,600 pcs
- Size returns removed per month
- 180 pcsUnits with the tool × (return share without − with the tool).
- Saving per return
- 605 ₽Return cost + share unsellable × loss per unit.
- Saving per month
- 108,900 ₽
- Tool cost per month
- 78,000 ₽
- Net effect per month
- 30,900 ₽
- Payback of the launch cost
- 7.8 months
- Effect over 12 months including launch
- 130,800 ₽
- Break-even reduction in returns
- 3.6 p.p. (5 p.p. in the pilot)The reduction in the return share at which the saving equals the monthly cost.
Sensitivity to the pilot effect
| Scenario | Reduction in returns, p.p. | Returns removed, pcs | Net effect per month |
|---|---|---|---|
| as in the pilot | 5 | 180 | 30,900 ₽ |
| 75% of the pilot | 3.8 | 135 | 3,675 ₽ |
| 50% of the pilot | 2.5 | 90 | −23,550 ₽ |
A pilot with self-selection overstates the effect: compare random groups. Use the share of returns for the reason “size”, not all returns. A better recommendation also cuts abandoned baskets and exchanges, but that is not counted here to avoid overstating the result.
All money excl. VAT. Size charts of Russian and foreign brands differ: see the Syntha Lab size reference.
Nuances and limits
- Do not compare those who opened the widget with the rest: self-selection overstates the effect.
- Use only size returns, not all returns.
- Do not take someone else’s reduction: every brand has its own patterns and audience.