sodcalculator

10 Tech How-To

Build vs Buy Software: Deciding Guide

Guide 10-51160 · Updated · Step by step

This guide walks you through a structured decision process for choosing between building custom software and buying an off-the-shelf product. You will learn how to define requirements, evaluate costs, assess risks, and make a decision that fits your organization’s strategic goals. Before you start, gather input from key stakeholders and document your current workflows.

What You’ll Need

  • List of business requirements and pain points
  • Budget estimate (initial and ongoing)
  • Timeline expectations
  • Access to stakeholders for interviews
  • IT team availability and skill set
  • Existing software inventory

Step-by-Step

  1. Define the problem clearly. Write down the specific business problem you need to solve. Avoid vague goals like “improve efficiency”; instead, say “reduce manual data entry by 50%” or “automate invoice approvals.” This becomes your north star.
  2. List must-have vs. nice-to-have features. Create two columns. Must-haves are non-negotiable for the software to work. Nice-to-haves are enhancements that can wait. This helps you compare products objectively.
  3. Research existing solutions. Search for commercial off-the-shelf (COTS) software, open-source options, and SaaS products that address your problem. Use review sites, vendor websites, and industry forums. Shortlist 3-5 candidates.
  4. Evaluate each candidate against your requirements. Create a scorecard with your must-haves and nice-to-haves. Rate each product on a scale (e.g., 1-5) for each feature. Include non-functional requirements like security, scalability, and ease of use.
  5. Estimate total cost of ownership (TCO) for buying. Include license fees, subscription costs, implementation, customization, training, maintenance, and hardware. Don’t forget hidden costs like data migration and integration with existing systems.
  6. Estimate TCO for building. Include salaries for developers, project managers, QA, and designers. Add infrastructure costs, development tools, testing, deployment, and ongoing maintenance. Plan for a 10-20% buffer for unexpected delays.
  7. Compare time-to-value. Buying typically gets you up and running in weeks; building can take months or years. Consider if speed matters for your competitive advantage.
  8. Assess strategic importance. If the software gives you a unique competitive edge (e.g., a proprietary algorithm), building may be justified. If it’s a commodity function (e.g., payroll), buying is usually smarter.
  9. Check vendor viability. For buy options, research the vendor’s financial health, customer reviews, and roadmap. You don’t want to invest in a product that will be discontinued.
  10. Run a proof of concept (POC). For the top 1-2 candidates, run a small pilot with real data. For build, create a prototype or MVP. This validates assumptions and reduces risk.
  11. Make a decision with a weighted score. Assign weights to factors like cost, time, features, and risk. Score each option and compare totals. Involve stakeholders in the final vote.
  12. Plan for post-decision. If buying, negotiate the contract and plan implementation. If building, set up agile sprints and assign a product owner. In both cases, define success metrics and review them after launch.

Common Mistakes

  • Ignoring hidden costs: Customization, integration, and training often exceed the initial license fee.
  • Overestimating in-house capabilities: Developers may lack domain expertise, leading to a product that doesn’t meet business needs.
  • Underestimating maintenance: Building is not a one-time cost; you’ll need ongoing support, updates, and bug fixes.
  • Choosing based on features alone: A product with more features may be harder to use or slower to implement.
  • Not involving end-users: Software that IT likes but users hate will fail adoption.
  • Ignoring vendor lock-in: Proprietary formats can make it hard to switch later.

Tips

  • Use a decision matrix with weighted criteria to avoid bias.
  • Talk to peers in your industry about their experiences with specific vendors.
  • Consider a hybrid approach: buy a base product and customize it lightly, or build a thin layer on top of an open-source solution.
  • Think about the total lifecycle: what happens in 5 years? Will the vendor still support it? Will your team still maintain it?
  • Prioritize integrations: software that doesn’t play well with your existing stack will cause headaches.
  • Get a written SLA from vendors for uptime and support response times.

Troubleshooting

If the decision gets stalled or the chosen path fails, here’s what to do:

  • Decision paralysis: Set a deadline and involve an impartial facilitator. If you can’t decide, run a POC for both options and let data decide.
  • Vendor doesn’t meet expectations: Revisit your requirements and see if they were realistic. If not, renegotiate scope. If the vendor is truly failing, have an exit plan with data export options.
  • Build project is over budget: Cut nice-to-have features, extend the timeline, or consider buying a component instead of building everything from scratch.
  • User adoption is low: Invest in training and change management. Sometimes the best software fails because people don’t use it.
  • Integration issues: Use middleware or APIs. If the vendor won’t support integration, reconsider.

Frequently Asked Questions

What is the biggest factor in the build vs buy decision?

Strategic differentiation. If the software is core to your business and gives you a unique advantage, building may be worth it. If it’s a support function, buying is usually more cost-effective.

How do I estimate the true cost of building?

Include developer salaries (fully loaded with benefits), project management, QA, infrastructure, tools, training, and maintenance (typically 15-25% of initial cost per year). Add a 20% buffer for scope creep.

Can I start with buying and later build?

Yes, many companies buy a product to get to market quickly, then build custom features as they grow. Ensure the vendor allows data export and API access.

What is a reasonable timeline for building custom software?

For a simple MVP, 2-4 months. For a full-featured product, 6-12 months or more. Buying can be done in weeks, including implementation.

How do I handle vendor lock-in?

Choose vendors that offer open APIs and data export in standard formats. In the contract, include a clause for data portability.

What if we don’t have internal technical expertise?

Then buying is almost always the better choice. Building without a strong technical team is high risk. Consider outsourcing if building is essential, but that adds complexity.

Related

General steps for current mainstream versions; menus can differ by device and version. See the disclaimer or report an error.

More in this bed

The whole Tech How-To bed →