

Off-the-shelf or custom software? A decision guide for executives
- Published
- Reading time
- 7 min
Buy or build? Few IT decisions tie up as much budget for as many years as the choice between off-the-shelf and custom software. Yet it is often delegated to the IT department as a technical detail.
That is a mistake. The make-or-buy decision for software helps determine how quickly a company can change and what sets it apart from the competition. This article shows which criteria executives should weigh, where the typical cost traps lie and why the best answer is rarely either-or.
The starting question: what sets your company apart?
Before licenses, budgets and service providers comes a strategic question. Does the software support an activity that sets you apart from the competition? Or one that every market player handles in much the same way? In 1996, Michael Porter summed up the core of good strategy in the Harvard Business Review: “The essence of strategy is choosing what not to do.” For IT, this means that not every function deserves to be developed in-house.
Software developer and author Martin Fowler therefore distinguishes between utility functions and strategic functions. For a utility function, Fowler argues, you buy the package and adapt your business process to the software. For a strategic function, by contrast, “you don't want the same software as your competitors.” Financial accounting, payroll and email are typical utility functions. A proprietary pricing algorithm, distinctive customer service or a digital product can be strategic.
The analyst firm Gartner captures the same principle in three layers in its Pace-Layered Application Strategy. Systems of record such as ERP and finance systems are meant to remain stable. Systems of differentiation map company-specific processes. Systems of innovation serve experiments with short cycles. Which application belongs in which layer depends on the business model: for an online retailer, the ordering process is strategic; for a machinery manufacturer, probably not.
Off-the-shelf software: proven and quickly available, but not infinitely malleable
Off-the-shelf software spreads development and maintenance costs across many customers. Companies get mature features, regular updates and the process know-how the vendor has gathered from many projects. Cloud solutions (software as a service) are often up and running in weeks rather than months. The vendor takes care of legal changes, for example in tax or regulatory reporting.
The flip side: a company that introduces off-the-shelf software adopts the vendor's process logic. For utility functions, that is precisely the point. It becomes risky when a company wants to keep its own processes and reshapes the software instead.
Discount retailer Lidl shows how expensive that can get. Starting in 2011, it built its merchandise management system eLWIS on SAP for Retail; in 2018, it stopped the project. According to Computerwoche, industry observers cited by the trade paper Lebensmittel Zeitung estimated the cost at around 500 million euros. As RetailDetail reported, the business daily Handelsblatt identified the core of the problem: Lidl values its inventory at purchase prices, whereas the SAP software works with retail prices. Lidl described the move as “not a decision against SAP, but one in favor of a system of our own”* and continued to develop its existing in-house solution. The lesson for others: a company that refuses to adapt a process to the standard is in effect treating it as strategic, and should make that decision deliberately before the project begins.
Custom software: a competitive advantage that needs ongoing care
Bespoke software pays off where a process underpins the business model. It maps workflows exactly instead of forcing them into someone else's template. The company owns the code, sets its own priorities and release cadence, and can test new ideas faster than a vendor that has to serve thousands of customers. That is why Fowler considers cost secondary for strategic initiatives: the opportunity cost of not doing something, he argues, is far greater than the cost of development.
This freedom comes at a price. Custom software is never finished. Security updates, new interfaces, regulatory requirements and technical modernization remain the company's responsibility for good. That requires in-house experts or a reliable development partner, clean documentation and a budget that extends beyond go-live. A common mistake is to plan custom software as a project with an end date rather than as a product with a life cycle.
Total cost instead of license price: the TCO calculation
A comparison that only sets license fees against a development quote is misleading. What counts is the total cost of ownership (TCO) over the planned period of use. For off-the-shelf software, this includes implementation, customization, interfaces, training and rising subscription prices. For custom software, it covers operations, maintenance, further development and the risk that knowledge rests with just a few people.
Both paths create dependencies. With off-the-shelf software, it is vendor lock-in: the vendor decides on prices, product strategy and discontinuations. According to SAP, for example, mainstream maintenance for the core applications of SAP Business Suite 7 ends at the end of 2027, and optional extended maintenance at the end of 2030. This puts many companies under pressure to migrate. With custom software, the dependency is on the development team or service provider. In both cases, therefore, clarify how you would get your data out and what a switch would cost.
Nor does artificial intelligence tip the scales as clearly as is often promised. In a randomized study by the research organization METR, experienced developers using AI tools in 2025 actually took 19 percent longer. METR itself now considers this result outdated. The key point remains: code that is written faster still has to be run, maintained and owned by someone.

The middle way: standard at the core, custom at the edge
In practice, the question is rarely either-or. A hybrid approach has proven its worth. Utility functions run on off-the-shelf software that is left as unchanged as possible. Differentiating functions are built as separate applications alongside it and connected via open interfaces (APIs).
SAP itself promotes this principle under the term clean core. The goal, write Rudolf Hois, Priyadarsana Panda and Steffen Pietsch in the SAP News Center, is to keep the core “clean, standard, and upgrade-stable” while still enabling differentiation “where it counts.” Extensions decoupled from the core, the authors say, simplify upgrades and accelerate innovation.
The advantage for management: updates to the standard remain manageable because no custom development in the core blocks them. At the same time, the development budget flows precisely into the functions that customers perceive as the difference. The prerequisites are a clear integration architecture and the discipline not to build special requests back into the standard after all.
Conclusion: a question of strategy, not technology
The choice between off-the-shelf and custom software belongs on the executive agenda. Four questions help put it in perspective. Is the process a genuine differentiator from the competition? Are we prepared to adapt our processes to the standard? Can we run and evolve our own solution over many years? And what would switching later cost?
If the answer to the first question is no, almost everything speaks for off-the-shelf software, used as close to standard as possible. If it is yes, custom software is an investment in competitive advantage and must be managed as such. For most companies, the solution lies in a deliberate combination: buy what everyone needs, and build what sets you apart.
Sources
- Michael E. Porter: What Is Strategy?, Harvard Business Review, November 1996
- Martin Fowler: UtilityVsStrategicDichotomy, July 29, 2010, updated 2016
- Frederick Vanbrabant: Pace layering an application portfolio, August 10, 2025 (on Gartner's Pace-Layered Application Strategy)
- Martin Bayer: Lidl stoppt millionenschweres SAP-Projekt für Warenwirtschaft, Computerwoche, July 18, 2018 (in German)
- RetailDetail: Lidl's failed IT-project cost half a billion, September 27, 2018 (citing Handelsblatt)
- SAP News Center: SAP Extends Its Innovation Commitment for SAP S/4HANA, Provides Clarity and Choice on SAP Business Suite 7, February 4, 2020
- Rudolf Hois, Priyadarsana Panda, Steffen Pietsch: Discover How to Extend SAP S/4HANA Cloud the Right Way, SAP News Center, August 12, 2025
- METR: Measuring the Impact of Early-2025 AI on Experienced Open-Source Developer Productivity, July 10, 2025
* Our translation of a quote from a German-language source. The images in this article are AI-generated.
We'll work with you to determine which processes off-the-shelf software covers well and where a custom solution pays off – free of charge and without obligation.
Off-the-shelf, custom or both?
More articles

