Wildberries API TypeScript SDK / computeROAS

Function: computeROAS()

function computeROAS(stats: 
  | FullStatsItem
  | DaysV3Item[], options?: ComputeROASOptions): ROASResult;

Defined in: utils/roas.ts:80

Compute ROAS (Return on Ad Spend) from WB fullstats per-day data.

Semantics (confirmed empirically against a live cabinet, task-136 AC#1):

• sum = ad spend (Затраты, RUB) — proportional to clicks × cpc, same-day.
• sum_price = order revenue (Сумма заказов, RUB) — attributed to the click day.

ROAS = Σ(sum_price) / Σ(sum) over a rolling window.

Why exclude the freshest day: WB's sum_price finalization lags ~1-2 days, so the most recent day undercounts revenue → same-day ROAS is a known footgun (Q6). The default excludeLastDays: 1 drops it. (0-rev-on-clicked-days is mostly genuine zero-conversion, not lag — but the freshest day is where the undercount concentrates.)

Parameters

Parameter	Type	Description
stats	| FullStatsItem | DaysV3Item[]	Either a FullStatsItem (uses its .days) or a DaysV3Item[] directly (e.g. FullStatsItem.days).
options?	ComputeROASOptions	ComputeROASOptions.

Returns

ROASResult

ROASResult. roas is null when spend === 0 (no ad spend) or when the window is empty.

See

PromotionModule.getAdvFullstats for the source data.

Example

const stats = await sdk.promotion.getAdvFullstats({
  ids: '36508180',
  beginDate: '2026-06-28',
  endDate: '2026-07-11',
});
const { roas, revenue, spend } = computeROAS(stats); // default: exclude freshest 1 day
if (roas !== null) {
  console.log(`ROAS ${roas.toFixed(2)}x (spent ${spend} RUB, earned ${revenue} RUB)`);
}