Invest · Technology & AI Due Diligence
How will technology change the economics of this business?
Technology diligence built around the holding period rather than the current system landscape — for software targets and, just as often, for companies that make physical things.
Conventional diligence examines what exists today. The return depends on what technology does to this business over the next five years.
The question the work is organised around
A standard technical due diligence answers whether the architecture is sound, the code maintainable, the security acceptable and the technical debt survivable. Those answers are necessary, and no buyer should proceed without them.
They are also answers about the present. Each one describes the asset as it stands. None of them describes what happens to the economics of the business when engineering, analysis, administration and parts of operations become substantially cheaper to perform — for this company and for everyone it competes with.
What can technology compress here, who captures the compression, and what breaks?
Three questions follow from it, and they are the ones that decide a holding period: which parts of the cost base can genuinely be compressed, which parts of the current advantage are becoming reproducible, and which operational dependencies would turn a cheap tool into an expensive failure.
Not only software targets
Technology diligence is often scoped as though it were only relevant when the target sells software. That was defensible when technology mainly was software. It is not defensible in a manufacturer whose production planning runs on spreadsheets, whose engineering throughput is the constraint on revenue, and whose quality process depends on a person who retires next year.
In an industrial business, the technology question is rarely about the product. It is about how the company actually runs, what that costs, and how much of it is about to become automatable — by this company, or by a competitor.
Where the specialised method fits
Where the investment case rests on an advantage that cheaper software and automation could erode, this work extends into AI Disruption Due Diligence: the rebuild test, platform absorption, and the three-to-five year case for what remains defensible. It is the same engagement, weighted differently, and which weighting a target needs is usually clear within the first conversation.
01
Where it applies
The method does not change by sector. The weighting does.
- Industrial and manufacturing businesses
- Machinery, engineering and automation companies
- Robotics and industrial technology
- Software and SaaS targets
- Technology-enabled services
- Professional and business services
- Logistics and multi-site operations
- Portfolio companies preparing a value-creation plan
For a software target the question is what remains defensible. For an industrial target it is usually what can be compressed, and who gets to keep the gain.
02
What gets examined
Not a checklist to be ticked. A structure for the argument, weighted to the target and the thesis.
Operations and process
- Which critical processes depend on manual work, email or spreadsheets
- Undocumented knowledge and single-person dependencies
- Disconnected systems and repeated administrative work
- Where the process is bad and technology would only digitise the fault
AI and automation potential
- Which knowledge-work processes can economically compress
- Whether AI can raise engineering throughput, and what that is worth
- Service and customer-interface automation
- What cannot realistically be automated, and why that matters
Production and physical systems
- Machine vision, robotics and physical automation potential
- Digital work instructions, process control and predictive systems
- Data integration between shop floor and business systems
- Where physical integration creates defensibility rather than cost
Technology estate and resilience
- ERP, MES, CRM, PLM, custom applications, low-code tools and interfaces
- Data platforms, cloud, on-premise and OT systems
- Which systems are business-critical, and what a day without them costs
- Cyber exposure read as an operational risk, not as a checklist
Data and defensibility
- Whether proprietary operational data exists — and whether it is usable
- Reproducibility of the data advantage by a funded competitor
- Workflow ownership and operational switching costs
- Platform absorption risk in the product or service category
Value creation and economics
- Build, buy, platform or stop — for each capability that matters
- Who captures the compression: company, customer, competitor or platform
- Technology investment actually required during the holding period
- Whether the operating model can execute the value-creation thesis at all
03
How the work runs
Scope, duration and depth follow the target and the stage of the transaction. The sequence does not change.
The thesis in the buyer’s own words
Before any technical work: what is being underwritten, at what price, on what assumptions. The assumptions are what gets tested — the diligence is only useful if it can contradict them.
How the company actually runs
Systems, processes, data flows, engineering and production economics, the estate and its interfaces. Management sessions, and where access allows, the people who operate the process rather than the ones who present it.
Compression and dependency mapping
Where cost and time can genuinely be taken out, what that requires, and which processes are carried by tools whose failure cost nobody has priced. The two maps rarely overlap, and both change the case.
Adversarial reconstruction
What would it take for a well-funded competitor with frontier tools to reproduce what customers actually pay for? The points at which the reconstruction fails are the moat. The points at which it succeeds are the exposure.
Read-out
Findings, red flags, the opportunities the plan has not priced in, and the questions that remain open. Delivered as an argument that can be attacked, not as a rating.
04
What you receive
Written work, presented and defended in person to the deal team or investment committee.
- Technology and AI compression map — where cost and time can be taken out, and by whom
- Operational dependency and criticality assessment
- Technology moat assessment, separated from technology quality
- Platform absorption and competitive-response view
- Automation and robotics potential, with the cases that do not pay
- Technology investment required during the holding period
- Direct challenges to the value-creation thesis
- Red flags, and the opportunities the plan does not capture
That is the shape of the work, not a fixed package. Which parts matter depends on the target, and saying so in advance is part of the job.
05
Where this is worth doing
A good fit
- Industrial or manufacturing targets where the value-creation plan assumes operational improvement
- Software or technology-enabled targets whose moat is asserted rather than demonstrated
- Theses that rely on productivity gains reaching EBITDA
- Businesses whose critical processes appear to run on tools nobody formally owns
- Boards being asked to approve an acquisition on a technology argument they cannot independently assess
A poor fit
- A code audit or a penetration test — necessary work, but not this work
- Confirmatory diligence intended to support a decision already made
- Situations where no independent judgment can change the outcome
Challenge an investment thesis
Describe the target, the stage of the transaction and the part of the thesis that would hurt most if it were wrong.
For private equity, growth equity, family offices, strategic acquirers and boards.