Estimate Storage Before Moving Custom Files to a Shopify App

Measure a representative set

Use the final exports your customers receive. Production projects can be much larger than deliverables, and a folder may contain several discarded revisions. Group materially different file types instead of averaging a tiny PDF together with a large video.

Record units consistently. Decimal megabytes and binary units can produce different displayed totals, so compare the provider’s measurement with your own calculation before using a forecast to approve charges.

A simple planning formula

Projected storage = historical eligible bytes
                  + new bytes per month × retained months
                  + retained revision bytes
                  + planning headroom

This estimates a steady workload. If the business is seasonal or file sizes are growing, model the actual months separately. Retention should follow a deliberate customer promise and approved policy, not merely the largest number an app advertises.

A fictional archive calculation

The subtotal is 3,860 MB. Adding a fictional 20% planning allowance produces 4,632 MB. The allowance is a business assumption, not a measured error margin or a recommendation that every shop needs exactly 20%.

Check what counts as stored

QuickAttach’s listing includes a storage allowance and a retention ceiling. Confirm the current billing rules and actual usage before migration. An external link leaves storage with its provider; it does not make the overall cost or availability responsibility disappear.

Also check individual upload size. QuickAttach support specifies files under 100 MB, which is a different constraint from total account storage. A plan with ample remaining capacity does not necessarily accept one oversized package.

Use the forecast to choose a batch

Move a verified sample, compare measured usage with the expected bytes, and explain differences before expanding. Do not delete old customer files simply to force the migration under an allowance without reviewing access commitments.

Keep three forecast scenarios: expected workload, seasonal peak, and longer retention. The useful output is a range of plausible storage needs with named assumptions, not a precise-looking bill derived from uncertain averages.