What Investors Actually Check in Diligence

The real question is whether the story reconciles.

Every business presents a version of itself in a deck. Diligence exists to check that version against source data: the accounting system, the contracts, the cap table, the pipeline, the churn.

Nothing damages a process faster than a number that cannot be reproduced from source. It rarely matters that the number was slightly wrong. What matters is that the investor now has to verify everything else, and the tone of the conversation changes.

The single most useful preparation is running the reconciliation yourself first, before anyone external asks.

What gets checked, and where it goes wrong

The areas are predictable. The gaps are predictable too, which is the useful part, because it means most of this can be closed in advance rather than discovered live.

Definitions cause more trouble than numbers

A surprising share of diligence friction comes from the same word meaning different things in different places. Churn calculated monthly in one report and annually in another. Revenue that includes services in the deck and excludes it in the accounts. A pipeline number that counts opportunities nobody has qualified.

None of this is dishonest, and all of it looks bad when discovered by someone else. The fix is to write the definitions down once, apply them everywhere, and be able to say plainly how each number is calculated.

If two people in the business would produce different figures for the same metric, that is the thing to fix first.

Key-person risk gets assessed either way

Investors are pricing what happens if a founder or a critical employee leaves. They will form a view on this regardless of whether it appears on a checklist, and the evidence they use is how much of the business is written down.

The practical test is whether someone competent could pick up a core process from documentation alone. Sales qualification, onboarding, delivery, renewal, incident handling. An honest answer of having to ask someone is a risk being priced.

This is also the cheapest thing to fix in advance, because writing down how the business works has obvious value even if no raise ever happens.

The legal housekeeping

IP assignment, contractor agreements, and the cap table are where early informality is most expensive. A contractor who wrote significant code without an assignment clause, an early adviser who was promised equity in an email, or an option grant that was agreed verbally are all genuinely difficult to resolve under deadline.

In Australia, contractor classification deserves particular attention, because a contractor who functions as an employee creates a liability that diligence will find and that is uncomfortable to price.

None of this is urgent until it is extremely urgent. It is worth an afternoon with a lawyer well before a term sheet exists.

Prepare as if you will be audited

The practical preparation is to assemble the evidence before it is requested: a data room with the contracts, the financials, the cap table, the key policies, and a short document explaining how each headline metric is calculated.

Doing this early has a second benefit. The process of assembling it is what surfaces the gaps, and gaps found by you are a task, while gaps found by an investor are a negotiation.

It also shortens the process considerably, and a short clean process is itself a signal about how the business runs.

Start earlier than feels necessary

Most of what diligence checks takes weeks to fix and cannot be fixed under time pressure. Documenting a process, assigning IP properly, or cleaning up a cap table are not things to attempt while also running a raise.

Two quarters ahead is a reasonable horizon. That is enough time to close the structural gaps without them competing with the raise itself.

The businesses that find diligence straightforward are generally not the ones with the best numbers. They are the ones that could already evidence how they run.

What to check in your business

• Headline metrics reconcile to source systems.
• Each metric has one written definition used everywhere.
• Revenue by customer is known, including concentration.
• IP is formally assigned by every contributor, including contractors.
• The cap table matches every promise ever made, including informal ones.
• Core processes are documented well enough for someone else to run them.
• The data room exists before anyone asks for it.