Digital due diligence is the pre-close assessment of what a target company runs its business on and how it acquires customers: the technology stack, the quality of the data and analytics beneath the reporting, the digital assets and who legally owns them, and the demand engine that produces revenue. It is not one workstream. It is four, usually sold by four different kinds of provider, and the most common failure in a mid-market deal is not that the work was done badly. It is that the buyer commissioned one of the four and believed they had bought all of them.
The phrase spread because buyers needed a word for everything that was not financial, legal, or market diligence but still lived on a screen. Providers filled the space from whichever direction they came from. Engineering firms sell code review, architecture assessment, and technical debt. Screening firms sell digital footprint and reputation checks. IT consultancies sell infrastructure, cybersecurity, and licensing. Agencies sell a search and paid-media read. All four describe the product as digital due diligence, and all four are describing roughly a quarter of it.
That makes it a scoping problem rather than a quality problem. A sponsor can commission a thorough technology assessment, receive a clean report, and still have no idea whether the customer acquisition assumed in the model works. Both things are true at the same time because nobody was asked the second question.
What the company runs on: core systems, hosting, integrations, security posture, licensing, vendor contracts, and the internal capability to keep it all running. This is the workstream the search results mostly return, and it is well served by engineering-led providers. A good output is a picture of what breaks under the growth in the model, and what integration costs if the thesis is a platform.
Whether the numbers the company reports can be reproduced from its own systems. Not whether it has a dashboard. Whether pipeline, revenue by source, and customer counts reconcile between the CRM, the finance system, and the board pack. Most mid-market targets fail this quietly. In one PE-backed services company we audited, 38 percent of revenue arrived through word of mouth that no system recorded. The business was healthy. The reporting was fiction, and every forward number built on it inherited the error.
Who legally holds the domains, the advertising accounts, the analytics properties, the review profiles, the code repositories, and the customer data. In owner-operated companies these are routinely registered to a founder’s personal email address or sitting inside an agency’s account. This is the cheapest workstream to run and the one most often left out, and it produces the sharpest findings, because an asset the company does not control is an asset the buyer is not acquiring. We have seen a platform lose years of performance history in an agency separation mid-hold.
How customers actually arrive and what they cost: acquisition cost by channel reconciled to invoices rather than to platform reporting, channel concentration, the split between demand the company owns and demand it rents, and whether attribution supports any of it. This workstream has the most direct line to the equity case and is the least likely to be in scope. The findings side of it is covered in Marketing Due Diligence. This guide is about making sure it gets commissioned in the first place.
Findings from the four behave differently in a model, and treating them alike is how good diligence output ends up ignored.
Technology findings are usually costs. A system that will not carry the plan has a replacement number and a date attached to it. That belongs in the funding requirement, not in an argument about price.
Data findings are confidence adjustments. They rarely produce a line item. They change how much of the growth case a committee should credit, because a plan built on numbers that cannot be reproduced is a plan built on an estimate nobody can size.
Ownership findings are conditions. An advertising account in an agency’s name, or a domain registered to a departing founder, belongs on the closing checklist rather than in the model. Cheap to fix before signing, expensive after.
Demand findings reach the multiple. A business acquiring most of its customers through one rented channel carries a risk the price should reflect, because concentration in acquisition is customer concentration one step upstream. The reverse holds too. A company that has already moved its mix toward demand it owns is buying growth more cheaply than its peers. In one portfolio company we rebuilt, acquisition cost fell 60 percent while the share of demand the business owned outright tripled.
Sequence matters when the window is short, and the right order is not the order providers pitch. Ownership first, because it takes days and can be established from registrar records and a short access inventory. Data second, because everything downstream depends on whether the reported numbers are real. Technology and demand then run in parallel, since they draw on different systems and different people.
Two access questions decide the depth of the entire exercise: read-only access to advertising accounts and analytics, and a genuine export from the CRM rather than a report generated by management. Without them the work is outside-in and the findings are directional. With them the numbers are checkable. Ask for both in the first scoping call, and treat a refusal as information rather than an inconvenience. The KPI list a deal team should be asking for is a reasonable opening request.
The seam between commercial and technology diligence. Commercial diligence sizes the market and stress-tests the thesis. Technology diligence assesses the stack. Neither opens the advertising accounts or tests whether the pipeline data is real, and in mid-market processes that gap is nobody’s explicit job.
Durability, as distinct from recognition. Quality of earnings tests how revenue is recognized. It does not test whether the revenue survives the campaigns being switched off, and bought revenue recognizes perfectly well. Quality of Revenue covers where the two questions separate.
The handoff to post-close. Findings delivered as a report rather than as a costed, sequenced fix list get read once and filed. The same questions answered from inside the systems after close become a commercial audit, and the 8-week version of that work should start from the diligence file rather than a blank page. Sponsors who connect the two get paid twice, once in entry price and once in execution speed.
This has stopped being an optional add-on. In S&P Global Market Intelligence’s 2026 Private Equity Survey, 71 percent of general partners and 53 percent of limited partners said they prioritize operational value creation over financial engineering. Pre-close is the earliest point at which that priority either costs something or buys something, and scoping the work properly is what decides which. Our own digital due diligence engagements are built to cover the four workstreams together, on the deal’s clock.
Digital due diligence is the pre-close assessment of what a target company runs its business on and how it acquires customers. In practice it is four separate workstreams sold under one name: technology and infrastructure, data and analytics quality, ownership of digital assets such as domains and advertising accounts, and the demand engine that produces revenue. Most providers do one or two of the four well. The common failure in a mid-market deal is not weak work, it is a buyer commissioning one workstream and believing the whole engine has been checked.
Technology due diligence assesses what the company runs on: architecture, code quality, hosting, security, licensing, integrations, and the team keeping it alive. It answers whether the stack survives the growth in the model. Digital due diligence, used properly, is wider. It includes the technology assessment but adds whether the reported numbers reconcile across systems, who legally owns the domains and advertising accounts, and what customer acquisition actually costs by channel. A clean technology report and an unexamined demand engine can coexist in the same deal.
Two to three weeks inside a normal mid-market process, run in a deliberate order. Asset ownership comes first because it takes days and can be established from registrar records and a short access inventory. Data and analytics reconciliation comes second, because everything downstream depends on whether the reported numbers can be reproduced. Technology and demand-engine work then run in parallel, since they draw on different systems and different people. Depth is set by access, not by calendar time.
Findings behave differently and should not be treated alike. Technology findings are usually costs, with a replacement number and a date that belong in the funding requirement. Data findings are confidence adjustments that change how much of the growth case a committee should credit rather than producing a line item. Ownership findings are conditions to close, not haircuts, and are cheap to fix before signing. Demand findings are the ones that reach the multiple, because concentration in customer acquisition is customer concentration one step upstream.
Have a revenue problem the board is asking about? Start a conversation.