Off-the-shelf vs. custom software: when each one wins

“Should we build it custom or buy off-the-shelf?” The most common question at the start of a project. And the most common mistake: deciding based on which one is cheaper in year one.

Custom software development is almost always more expensive in upfront investment than a SaaS subscription. But the total cost of ownership (TCO) over 24-36 months can land in completely different places — depending on how many workarounds the boxed product demands, how many integrations you have to glue together, and how poorly your processes fit its built-in logic.

Here is the framework we use to decide — without ideology, just on numbers and context.

When off-the-shelf wins

Boxed solutions (Shopify, HubSpot, Salesforce, NetSuite, Xero...) cover the typical 80% use case. For a company that behaves according to that 80% scenario, they are unbeatable. Five situations where it does not make sense to replace them with custom development:

• Standard processes without special workflow. A classic B2C e-shop with a cart, payment, and shipping. CRM for a sales pipeline without specifics. Standard accounting. Boxed delivers in weeks, custom in months — with no upside.
• Small volume, limited budget. Below roughly $5k monthly revenue or $50k annual run rate for the system in question — custom development ROI never pays off. SaaS for $50-$500/mo is the right call.
• Team without technical ownership. If you have no one who can run a custom system (or no budget for a long-term support contract), boxed gives you predictable pricing and SLA.
• Fast time-to-market. MVP in 4-6 weeks with off-the-shelf plus 2-3 plug-ins. Custom MVP at minimum 8-12 weeks. If you are testing a business hypothesis, speed has a price tag.
• The ecosystem is more than the software itself. Shopify has thousands of themes, integrations, dropshipping partners. Salesforce has AppExchange. Custom development does not deliver these plug-and-play modules.

When custom software development wins

Custom is the right call where boxed starts swelling with workarounds and integrations. Five signals it is now economical to build your own:

1. Specific processes no boxed product covers

When you find yourself describing your processes with phrases like “almost like Shopify, but...” or “Salesforce would work if it could...” — those are symptoms. Your business logic does not fit the box the boxed software defines. Either you bend it (and lose competitive advantage in the process), or you pay for workarounds.

Real-world example: client in MLM. No boxed software handles their commission structure (downline bonuses, override commissions, monthly qualification criteria). Spreadsheet plus manual recalculation = 40 admin hours a month. Custom system with versioned business logic (how we modelled it) = 0 admin hours, plus auditability for accounting reviews.

2. Deep integration with existing systems

Boxed software typically offers an API, but its integration with your ERP/CRM/accounting is limited to whatever a plug-in handles. When you need:

• Real-time stock sync, not batch
• Custom field mapping between systems
• Two-way synchronisation with conflict resolution
• Logic that spans multiple systems (e-shop + ERP + carrier)

Boxed software gives you these via plug-ins — and each plug-in is another dependency that can break independently. Enterprise integration is one of the main reasons companies move to custom.

3. Boxed is starting to cost more than custom would

The classic signal: monthly fixed costs of boxed software + plug-ins + integrations + manual labour around it > the monthly amortised cost of a custom solution (development cost spread over 24-36 months + operations).

Concrete formula:

Boxed TCO 24m =
  24 × (subscription + plug-ins + payment fees + manual work)

Custom TCO 24m =
  one-off development +
  24 × (cloud hosting + maintenance + 1 minor release/quarter)

For MessyPlay.cz, boxed Shopify + 9 plug-ins cost roughly €240/month. Custom development was €8,000 one-off + €50/month cloud hosting. Break-even at 18 months. After 24 months, custom saves 80% versus boxed. Detail in the 24-month TCO breakdown.

4. Competitive advantage you cannot buy

If your process / data / algorithm is part of your value proposition — boxed software is actually taking your competitive advantage. Same Shopify + same plug-ins is what your competitor runs too.

Custom software development in this case is not a cost, it is a moat. Classic examples:

• Recommendation engine tuned to your specific SKU portfolio
• Pricing logic that reacts to specific signals
• Workflow that cuts your fulfilment time by 20%
• Dashboard your management uses for operational decisions

5. Growth or M&A on the horizon

Boxed software scales until you hit a size where it cracks — either in price (Salesforce Enterprise licences scale into tens of thousands annually), in performance (Shopify Plus at $2,000/mo), or in flexibility (multi-entity setup post-acquisition).

When you are in growth mode or planning acquisitions, your own platform gives you control over how fast and which way you expand. Boxed software constrains you to whatever its roadmap delivers.

Hybrid approach: usually the real answer

In practice, this is not a binary choice. Most enterprise architectures are hybrid — boxed for standard functions, custom for unique logic.

• Accounting in Xero, e-shop custom. Xero handles standard accounting well, custom would not add value. The e-shop has unique logic → custom.
• Salesforce CRM, custom commission backend. Salesforce handles the standard pipeline, custom handles MLM commissions that do not fit any CRM.
• Shopify storefront, custom integration layer. Shopify for frontend/checkout (where the ecosystem wins), custom for ERP/carrier/supplier integrations (where Shopify plug-ins are not enough).

The hybrid approach requires architecture that connects systems without point-to-point chaos. That is the most common scenario where we help clients.

Decision framework

When a client confronts us with the boxed-vs-custom question at the start of a project, we walk through this:

1. Process fit: Does boxed software cover 80%+ of your processes without compromise? If yes → boxed.
2. 24-month TCO: Add up fixed + variable + hidden manual labour. Boxed wins year one; custom usually wins year two.
3. Strategic value: Is your competitive advantage in a process that boxed treats as standard? If yes → custom (or at least a custom layer over boxed).
4. Team capacity: Do you have someone who can run and evolve the system? If not, custom needs a support contract.
5. Growth horizon: Where will you be in 3 years? Boxed scales smoothly until it suddenly does not. Custom scales based on what you invest.

What to do when you cannot decide

Most common situation: it partly fits, partly does not. The progressive approach:

1. Step 1: Build an MVP on boxed software (speed, low risk). Test the business hypothesis.
2. Step 2: Once the hypothesis is validated and the limits of boxed become clear, identify 1-3 processes that most constrain growth.
3. Step 3: Build a custom layer around those processes (often API gateway or middleware), leave the rest in boxed.
4. Step 4: When the economics line up, gradually move more processes to custom — or leave boxed where it works.

This path does not eliminate risk, but it spreads it over time. And it gives you optionality — you can always step back or move forward.

Who we build custom software for

At Codedock we specialise in enterprise systems and integrations — where boxed software hits its limits. Specifically:

• E-commerce platforms with unusual fulfilment workflow (MessyPlay)
• SaaS platforms with an IoT component (Fotopast.cloud)
• Enterprise systems with versioned business logic (JUST)
• Performance-critical systems processing GB/s of data (Netigo)
• Data warehouses for distributed enterprises (Magistra)

If you are weighing whether custom software development makes sense for your project, or whether boxed is enough — book a 30-min consultation and we will walk through your specific situation. No sales process, no pressure. If boxed is enough, we will tell you straight away.