Keeping data up to date

Product data is never really finished. Prices change, products sell out, new items launch, descriptions get rewritten. A feed that was perfect yesterday can be wrong today, which means getting your data right once isn't enough. You have to keep it right. The good news is this is mostly a job you set up once and let the platform run for you.

You can have data that's complete, accurate, and perfectly formatted, and it can still be wrong tomorrow. That's why freshness is part of what makes product data good. A channel showing a price that no longer matches your shop, or an item that's actually sold out, gives shoppers a bad experience and can earn you penalties or suspensions from the channel. Keeping data current protects both your sales and your standing.

Remember the journey from how a feed moves through Productsup: every site run imports your latest source data, optimizes it, and exports it to your channels. Keeping data fresh really just means running that process regularly, so the newest prices and stock levels flow all the way through to every channel. You don't do this by hand. You set it on a schedule.

The Scheduling feature lets you set site schedules so your site runs automatically at the interval you choose, no clicking required. Set one up in Set up scheduling and triggering processes.

The key is to match the cadence to how often your data actually changes. Fast-moving stock and pricing might need hourly runs, while a stable catalog might be fine once a day. Don't over-schedule, though: every run is a full run that costs processing time, so running far more often than your data changes just burns resources, as covered in data processing performance.

Schedules handle "run at this time." Trigger events handle "run after that happens." You can have one site's run automatically kick off another's, so when an upstream site finishes, the sites that depend on it refresh right away instead of waiting for their own clock. It's how you keep a chain of linked sites consistent, and it's set up in the same Scheduling feature.

Refreshing often raises a fair worry: isn't resending your whole catalog every run wasteful? It would be, which is why many channels support delta updates. Instead of the full feed each time, the platform works out what's new, changed, or deleted since the last run and sends only that slice. It's faster, lighter, and easier on a channel's API limits, the same efficiency idea from data processing performance.

Delta only works if the platform can reliably tell which item is which from one run to the next, and that comes back to the unique item identifier. When identifiers are stable, a price change lands on exactly the right product. When they wobble, the platform can't tell a changed item from a brand-new one, which is one more reason identifiers matter so much.

Fresh data only helps if the run actually succeeds. If an import quietly fails or a source file shows up half-empty, an automated run could push bad or stale data without you noticing. Monitor watches for exactly this, flagging things like a sudden drop in item count, a failed import, or unmapped values before they reach a channel. Set it up in Monitor, and keep an eye on the overall picture from your dashboard.