How we scope a custom ERP, and why we sometimes say no

Every few weeks someone asks us to “build an ERP”. It is usually the most important software decision that company will make for a decade, so before we quote anything, we run the same exercise we have used for years.

1. We ask what spreadsheet it replaces

Almost every good ERP we have built started life as a spreadsheet that broke. If nobody can show us the spreadsheet (or the WhatsApp group, or the paper register), the requirement is usually not real yet. The spreadsheet tells us the fields, the users and the real workflow, more honestly than any requirements document.

2. We find the one process that pays for everything

An ERP proposal with twelve modules usually fails. We ask: which single process, if automated, saves the most money or prevents the biggest mistakes? That module gets built first and goes live alone. For a vacation-rental client that was booking and payout reconciliation. Everything else came later, funded by the time that first module saved.

3. We say no when off-the-shelf is honestly better

If a client's process is standard, standard software is cheaper and better supported. Custom software earns its cost only where your process is genuinely different from everyone else's. Telling a client to use Zoho or Tally for a part of their business has cost us some invoices and won us long relationships.

If you are weighing custom against off-the-shelf, we are happy to talk it through, even if the answer ends up being “don't hire us”.