Sphere Partners
Deal team member standing by a glass board in a financial district office while mapping a software acquisition checklist

Software Due Diligence: The Complete Checklist for Buying a Software Company

A sectioned software due diligence checklist for buyers of software companies, with what to request in the data room and how to decide which findings change the price, the terms or the post-close plan.

9 min read
In this article

Software due diligence is the review a buyer runs on a software company's product, code, architecture, security, infrastructure, intellectual property and engineering team before acquiring it. A good checklist turns that review into specific questions, each tied to evidence the seller must provide and to a decision the buyer will make: walk away, adjust the price, add contract protection, or fix the issue after close.

Below is a sectioned checklist you can hand to a deal team or a diligence provider, followed by what to request in the data room and how to sort the findings. If you were looking for due diligence software, such as a virtual data room, note that this article is about the review itself rather than the tools. For an independent assessment run against this scope, Sphere's technical due diligence team delivers costed findings in four weeks or less.

How Should You Use a Software Due Diligence Checklist?

Not every line deserves the same depth. Start with the investment thesis and work backward. If the plan depends on doubling the customer base, scalability and cost to serve get the deepest review. If it depends on selling into regulated enterprises, security and compliance do. If the product is marketed on AI, the AI questions move to the top. A checklist used without a thesis produces a long report that nobody can act on.

Rate every item red, amber or green, and attach a remediation cost and timeline to anything that is not green. That turns the checklist from a compliance exercise into an input to the deal model. This list assumes a software or SaaS target. For an early-stage company, our startup technical due diligence checklist is a better fit. For a company that runs on technology but does not sell it, use an IT due diligence checklist instead.

The Software Due Diligence Checklist

The eight sections below cover the scope of a full review of a software company. Each question should be answered with evidence, not assurances.

1. Product and roadmap

  • Does the product do what sales materials and customer contracts say it does?
  • How much of the roadmap is already promised to specific customers, and what does it depend on technically?
  • How often do roadmap dates slip, and why?
  • Which features drive most usage and revenue, and which are rarely used but costly to maintain?
  • Is the product multi-tenant, single-tenant or customized per customer?

2. Architecture and scalability

  • Is there a current architecture diagram that matches what is actually deployed?
  • Where are the single points of failure in the core transaction path?
  • Can the platform handle the growth in the deal model with infrastructure changes alone, or does it need re-architecture?
  • How tightly coupled are the components, and how risky are database schema changes?
  • Which third-party integrations are critical, and how are they monitored?

Make the scalability question specific. In Sphere's assessment for Betr, a real-time sports micro-betting platform, the team built a 5x growth benchmark into the review from the start: could the target serve five times its current load without major architectural rework? The report gave a clear verdict alongside a prioritized technical debt register and a CI/CD maturity score. The Betr scalability due diligence case study explains why that framing matters for deal modeling.

3. Code quality and technical debt

  • What is automated test coverage on revenue-critical paths, and do tests run on every change?
  • How old are the main frameworks and dependencies, and what is the upgrade path?
  • Is there a documented technical debt register, and is debt work funded?
  • Is code review mandatory, and does it actually happen?
  • How much code was written with AI assistance, and was it reviewed with the same rigor?

Debt is normal; unpriced debt is the problem. Our guide to technical debt for business leaders explains how to express it in terms a deal model can use. If a large share of the code was AI-generated, add the checks in our guide to reviewing AI-generated code in due diligence.

4. Security and compliance

  • When was the last penetration test, what did it find, and what was fixed?
  • Are there known critical vulnerabilities in production today?
  • How are access, secrets and credentials managed, including for former employees and contractors?
  • Is there an incident response plan, and has it been used or tested?
  • Which frameworks apply, such as SOC 2, ISO 27001, HIPAA, PCI DSS or GDPR, and are the reports current?
  • Has the company had breaches, ransom events or regulator inquiries?

5. Open source and intellectual property

  • Is there a current software bill of materials (SBOM)?
  • Which licences are in use, and do copyleft obligations apply to anything the company distributes?
  • Has every employee and contractor assigned their IP to the company?
  • Are any components used with restrictive, customized or missing licences?
  • Is source code held in escrow for any customer, and under what release conditions?

This section is easy to skip and expensive to get wrong. The 2024 Synopsys OSSRA report (opens in new tab) found open source licence conflicts in 53% of the codebases it analyzed, and code with no discernible or a customized licence in 31%.

6. Infrastructure, cloud and operations

  • What does hosting cost per customer, and how has it trended against revenue?
  • Is infrastructure defined as code and reproducible in a new account?
  • When was disaster recovery last tested, and how long did recovery take?
  • What is the uptime and incident history against customer SLAs?
  • How often does the team deploy, and how often do deployments fail or roll back?

7. Team and delivery process

  • Who holds critical knowledge, and what happens if they leave?
  • What is the mix of employees, contractors and outsourced teams, and who owns those relationships?
  • How does work move from planning to release, and where does it stall?
  • What are engineering attrition and hiring trends over the last two years?
  • Which engineers need retention arrangements for the deal to work?

It helps to compare the target's organization with a healthy reference point. Our piece on the ideal software development team structure sets out the roles and ratios worth checking against.

8. Data and AI

  • What customer data is stored, where, and under what consent and contract terms?
  • Is the data model clean enough to support the analytics or AI features on the roadmap?
  • If the product uses AI, is it proprietary or built on third-party models, and what does inference cost per customer?

If AI is central to the valuation, this section needs its own workstream. Our guide to evaluating an AI company before you invest covers data rights, evaluation evidence, inference economics and vendor dependency.

What Should You Request in the Data Room?

The checklist only works if the evidence arrives early. Ask for these items in the first request list, because each one answers several checklist questions at once.

RequestWhat it reveals
Read-only repository accessCode quality, test coverage, commit history and how concentrated contributions are
Architecture diagrams and runbooksWhether documentation matches what is deployed, and how operations really work
SBOM or dependency scanOpen source vulnerabilities, outdated components and licence obligations
Recent penetration tests and fix ticketsWhether security findings actually get resolved, and how quickly
Twelve months of cloud billsCost to serve, cost trend against revenue, and waste
Incident and on-call historyReal reliability, as opposed to the SLA on paper
Org chart, tenure and contractor listKey-person risk and gaps in IP assignment
Compliance reports and policiesCertification status, exceptions and open audit findings

Which Findings Should Change the Deal?

Once the checklist is complete, sort every red and amber item by what the deal team should do about it. Most findings fall into one of four groups:

  • Walk away or re-scope: the company cannot show it owns its core code, the platform cannot support the growth plan without a rewrite the buyer will not fund, or there is undisclosed breach or regulatory exposure.
  • Price adjustment: remediation that can be quantified, such as framework upgrades, security fixes, infrastructure rework or replacing critical contractors.
  • Contract protection: risks that are real but hard to size, such as licence exposure or data rights, handled through specific warranties, indemnities, escrow or holdbacks.
  • Post-close plan: accepted debt and process gaps, sequenced into the first 100 days and funded in the investment plan.

How those groups flow into valuation and the 100-day plan is covered in what technology due diligence covers and what investors check. Sellers can use the same list in reverse: running it on their own company before a sale process starts gives them time to fix the costliest items before a buyer finds them. For the full methodology, including how a code audit fits alongside this checklist, read our technical due diligence and code audit guide.

Frequently asked questions

It is a buyer's review of a software company's product, code, architecture, security, infrastructure, IP and engineering team before an acquisition or investment. The goal is to confirm the technology supports the deal thesis and to price whatever does not.

A structured list of questions and evidence requests used to make sure no area of a target is missed. For software companies it covers product, architecture, code quality, security, open source and IP, infrastructure, team, and data and AI.

No. Software due diligence is the review of a software company. Due diligence software refers to tools, such as virtual data rooms and document review platforms, that help deal teams manage any kind of diligence.

Typically an executive summary for the investment committee, findings rated by severity with remediation cost and timeline, a technical debt register, security and licence risks, team and key-person observations, and a recommended post-close plan.

Most reviews of a single software company take two to four weeks once data room and repository access are in place. Narrow reviews finish faster; carve-outs or multi-product targets take longer.

Often, yes. A sell-side review a few months before going to market lets founders fix the most expensive issues, prepare evidence for the data room and answer buyer questions faster, which tends to reduce price chips and shorten the process.

Buying a software company?

Sphere's technical due diligence verifies the checklist against code and systems and returns costed findings and a 100-day plan in four weeks or less.

Related posts

We'd love to hear from you!

Please provide your contact details, and our team will get back to you promptly.