See which influencer creators actually drove Shopify sales

On the third of every month at 8am, work out which creators in our Later influencer program actually drove sales last month, not just impressions, and post the answer to our influencer channel in Slack. Use the previous full calendar month as the date window, written as YYYY-MM-DD in UTC, and apply that same window on both sides of the join so the two systems are comparable.

Start in Later. Use Get Campaign Performance for last month's campaigns and Get Creator Performance for per creator numbers along with each creator's profile and social handles. Later requires a metrics parameter on every performance route and returns only the metrics you ask for, so request impressions, engagements, and the link click metrics, which are prefixed trackingLinks and affiliateLinks. The date basis defaults to post date, which is what we want here. Campaign ids are capped at 50 per request, so batch them if the program is larger, and note that the reporting token expires 12 hours after it is issued.

Then pull the matching orders from Shopify with List Orders, filtered to the same created-at window. Count paid orders only, exclude test orders, and net out refunds rather than taking the gross order total. Read the discount code straight off each order rather than enumerating discount codes separately, because Shopify scopes discount codes under price rules and walking them is the slower and less reliable path.

Join the two sides on the discount code. Our codes follow a naming convention where the code contains the creator handle, so match each order's discount code back to a creator handle from the Later creator data, comparing case insensitively and ignoring separators like dashes and underscores. Later is a reporting system that holds no revenue of its own, so this join is entirely code based and there is no revenue field in Later to fall back on.

For every matched creator compute attributed revenue, order count, average order value, revenue per thousand impressions, and revenue per link click. Guard the efficiency numbers against division by zero, and mark them as not available rather than printing a misleading figure when a creator has no impressions or no recorded link clicks.

Post one Slack message to our influencer channel using Send a Message, in three sections. First, the creators earning their keep, ranked by attributed revenue, each line showing revenue, orders, average order value, revenue per thousand impressions, and revenue per link click. Second, the creators with real reach and no sales, meaning meaningful impressions or link clicks but no attributed orders, framed as candidates for a content or offer rethink rather than as failures. Third, any discount codes with sales last month that did not match a creator in Later, so we can fix the mapping.

State the match rate near the top of the message: how many of last month's paid orders carried a discount code we could tie to a Later creator, and what share of last month's revenue that represents. Nobody should read a partial join as the full picture, so when the match rate is low, say so plainly instead of burying it. Keep the message compact enough to read inside the channel, and when a section is empty, say so in one line rather than dropping the heading.