Users groan: “My Excel file crashes”, “I waste hours copying data between sheets”, “We have no reliable reports”. The knee-jerk reaction: let’s order a custom business app! More often than not, a well-optimised spreadsheet or an existing off-the-shelf tool solves the problem. How to decide without gambling?
The warning signal: the process became a productivity sinkhole
Month-end reporting takes three days to consolidate. Data lives in silos: one file on one desktop, another in a shared cloud folder, five slightly different versions emailed around every week. When the overhead of managing the process exceeds the value it produces, it’s time to question.
But this signal alone isn’t enough. Many processes are chaotic by design, not by accident. A custom app doesn’t cure disorganisation; it automates it.
Two decisive questions every stakeholder must answer
Do not commission a custom app until you have answered “no” to the first question and quantified the second. Otherwise you buy an expensive tool for a simple problem.
1. Is the process standardisable?
A custom app earns its keep only when the business process is unique, repeatable, and business-critical. Example: multi-shift maintenance scheduling under strict compliance rules. If your process is merely a collage of local practices (e.g., time-sheet collection), an off-the-shelf solution suffices.
Test: try writing your process in five crisp steps. If you hesitate by step three, the process hasn’t stabilised yet.
2. Is the Return on Investment quantifiable?
A spreadsheet has hidden costs: lost hours, errors, maintenance. A custom app has visible costs: quotes, delays, ongoing maintenance. The only valid decision criterion is ROI computed with data, not gut feel.
How?
- Compute monthly labour hours spent
- Multiply by loaded hourly cost (salary + overhead)
- Estimate separately the cost of errors and delays actually observed
- Compare to a realistic custom-app quote including 24 months of maintenance
- If predicted savings exceed total cost, explore custom build. Otherwise, optimise first.
What a custom app doesn’t solve
1. An organisational problem
If the process is chaotic because the team lacks methodology, a custom app won’t fix it. It will merely automate the chaos. Start with process formalisation. Automate later.
2. A one-off need
A custom app has high upfront cost. It pays off only for a durable, recurring need — not a one-off project. Otherwise pick an off-the-shelf tool or a service contract.
3. A process that changes weekly
If requirements drift faster than you can document them, a custom app turns into a technical-debt nightmare within months. A process that still changes every week will make the tool partly obsolete before it even ships — stabilise the way of working first.
Low-risk alternatives to a full custom build
Option 1: Optimise the existing spreadsheet ecosystem
Modern low-code tools (Airtable, Notion, Google Sheets with Apps Script) let you build clean UIs, automated exports, alerts, and consolidated views. Cost: hours of configuration. ROI: immediate.
Option 2: Configure a mainstream ERP/CRM
Solutions like Odoo, Zoho, or Salesforce have configurable modules covering many standard processes. Cost: subscription + a few days of setup.
Option 3: Outsource via a service contract
Complex processes (payroll, logistics) often run better as externalised services. Advantage: no technical debt, no maintenance, and you do not own the tool or its upkeep. Cost: monthly fee. ROI: responsive and zero heavy investment.
Three hard indicators that push toward custom development
Before signing any quote, measure these KPIs. If any is true, explore custom development.
- The process incurs a significant hidden cost, measured rather than estimated (labour + errors + data loss). The order of magnitude depends on your context, not on a general rule.
- The process blocks a critical downstream process (e.g., customer delivery delayed by upstream manual steps).
- The process is a competitive differentiator (e.g., proprietary dynamic pricing algorithm).
A three-stage roadmap
- Audit: quantify current process cost (labour, errors, impact). Identify bottlenecks.
- Explore: test alternatives (optimised spreadsheet, configured ERP, service contract). Compare total costs over 24 months.
- Decide: if ROI > 1, commission a small custom MVP. Otherwise, improve existing tooling or outsource.
Golden rule: never automate an organisational problem
A custom app cannot replace competent staff or clear methodology. If your pain is lack of standardisation or documentation, fix that first. Automating a chaotic process is like pouring concrete on a pile of sand: it doesn’t stick.
TL;DR: Decide by tomorrow morning
Open a blank spreadsheet and fill in:
- A five-step outline of your current process
- Monthly hours spent × loaded cost → economic loss
- Recurring errors and their quantified cost
- Compare to a realistic custom-app quote including 24 months of maintenance
- Decide: optimised spreadsheet / configured ERP / outsourced service / custom MVP
That spreadsheet is the only tool you need. Anything else is guesswork.