"Should we buy or build?" is the question behind half our discovery calls. Off-the-shelf ERP promises speed. Custom ERP promises fit. Both can be right — and both can become expensive mistakes if you choose for the wrong reason.
This guide is the framework we use with retail, logistics, and operations teams evaluating custom ERP development versus SaaS suites like NetSuite, SAP Business One, or vertical retail packages.
Buy when your workflows are industry-standard
If 80%+ of your operations match how a mature SaaS ERP expects you to work — standard PO flows, generic warehouse picking, chart of accounts that fits out of the box — buying is usually correct. You get SOC2, upgrades, and a consultant ecosystem on day one.
Buy signals: low differentiation in back-office ops, predictable SKU models, minimal consignment or vendor-specific rules, and a team willing to adapt process to software.
Build when operations are your competitive edge
Build when proprietary workflows drive margin or customer experience — multi-vendor consignment, custom reorder logic per SKU, field ops that do not map to desk-first ERP UIs, or finance close rules your accountants already perfected in spreadsheets.
Meridian Retail is the pattern: five tools and five spreadsheets meant "buying NetSuite" would force painful workarounds. A purpose-built retail inventory management system on one ledger delivered faster ROI than customizing a suite.
The 5-year TCO math most teams skip
SaaS looks cheap at signing. Hidden costs accumulate: per-seat licensing as headcount grows, integration middleware for every adjacent tool, customization consultants billing hourly forever, and data export fees when you outgrow the platform.
Custom ERP has higher upfront build cost but zero marginal seat tax and architecture you control. Run TCO over five years with realistic growth — not year-one sticker price. Include internal time lost to workarounds; that line item often exceeds license fees.
Integration debt: the silent killer
Many teams "buy" then spend 18 months gluing Shopify, 3PL EDI, barcode hardware, and finance into a fragile integration layer. That layer becomes the real system — undocumented, untested, owned by one senior engineer who is quitting.
Custom ERP development should include integration design from week one: which systems are source of truth, which events sync bidirectionally, and what happens when the 3PL API is down. Buying without an integration plan is not buying — it is deferring a build with extra vendors in the middle.
Key takeaways
- Buy for standard ops; build when proprietary workflows drive margin or UX.
- Model 5-year TCO including seats, consultants, middleware, and workaround labor.
- Treat integration architecture as day-one scope — not a phase-two surprise.
- Phased rollout beats big-bang: inventory first, finance second, analytics third.
The right answer is often "neither pure buy nor pure build" — buy commodity modules, build the 20% that defines how you operate. If you are unsure, run a two-week discovery mapping handoffs and data sources. The answer usually becomes obvious before anyone signs an ERP license.