How do I replace ecommerce ops spreadsheets with internal tools my team can actually maintain?

Replace the copy, not the question. A spreadsheet is a private ledger of numbers that already exist somewhere else. An internal tool is a view (and sometimes a small write) on the system of record. If the numbers do not already agree, a prettier sheet will not save you. Twenty-five small tools beat twenty-five tabs. Zero tools on top of a fight between Shopify and the 3PL is still a fight.

Start with three weekly jobs

Start with three jobs that rot on a weekly export. Do not start with a hero dashboard or a banner. Start with a list a shift lead will open on Monday.

Picker scoreboard: lines or units per hour from actual pick events, by person and by wave, for yesterday. Not a whiteboard photo typed into a sheet. The tool reads completed picks. It does not invent a target. Supervisors still coach. If pick events do not exist as rows, you do not have a scoreboard. You have a guess.

Supplier OTIF: promised receipt date versus actual receipt date, by PO and SKU, using real lead time. On-time is a receipt, not an ASN email. The tool flags POs that will miss the wholesale window. A buyer still sends the chase. Late is a date math problem. It is not a slide that says OTIF.

Channel margin: net revenue minus WAC minus outbound minus fees minus a returns allowance, by class (Shopify, Amazon, wholesale). Same GL accounts, tagged. The tool does not allocate rent. If WAC is garbage, the dashboard is garbage. Fix valuation first. A class P&L that pulls Shopify payout fees from one app and WAC from a spreadsheet is two products arguing.

Keep the tools maintainable

Maintainable means one source, few columns, a named owner, and no nightly CSV glue. If you need a contractor to change a filter, you built a second product. If the tool writes, gate it: draft a count, draft a PO, never post a journal alone. Kill the sheet the week the tool is trusted, or both will drift. Do not rebuild the warehouse and the books in a low-code app. A tool that re-keys on-hand so the scoreboard looks right is a second ledger. Fix the first one.

The scoreboard, the OTIF list, and the class P&L only work because pick, receipt, and invoice already live in one place. A ranked list of late POs is a tool. A slide that says OTIF is not. Build the view after the event exists. Not the other way around. If a job needs a nightly export to exist, it is not ready for a tool. It is still a reconciliation.