Automation due diligence: what I check before capital touches a company

Before capital touches a company, I want to know one thing: does this business run on systems, or does it run on a person? The answer lives in the automation layer — the repeatable work, the handoff points, the tools that hold the org chart together. Here is exactly what I check.

What does the repeatable work actually look like?

Every business has a core loop — the sequence of actions that turns a stranger into a paying customer and keeps them there. My first move is to map that loop end to end and mark every step that a human executes manually on a recurring basis.

I am looking for:

• Intake and qualification — How does a new lead enter the system? Is there a voice agent, a form, a CRM trigger, or does someone manually log it? A business where leads arrive by phone and get written on a notepad is a different risk profile than one using an AI phone agent like Business Runner to capture and qualify every inquiry automatically.
• Follow-up cadence — Is follow-up scheduled and triggered by the CRM, or does it happen when a salesperson remembers? Manual follow-up is one of the most common hidden costs I find. See speed to lead: why the first minutes decide who wins the customer for why the timing gap matters.
• Scheduling and dispatch — Can a customer book, reschedule, or confirm without a human in the loop? In service businesses especially, this is often the single biggest labor sink.
• Invoicing and collections — Are these triggered automatically on job completion, or does someone have to remember to send them?

Anywhere a human is doing the same thing more than twice a week on a predictable schedule, that is a system waiting to be built — or a liability waiting to surface.

Which systems are load-bearing?

Load-bearing systems are the ones that, if they went down or the person running them left, would immediately stop revenue. I distinguish them from nice-to-have automations by asking: what is the blast radius if this breaks?

System	Load-bearing signal	Risk if manual
Lead intake	Touches every new customer	Missed leads, slow response
Appointment booking	Directly gates revenue	Scheduling gaps, no-shows
CRM follow-up	Drives conversion and retention	Leads go cold silently
Invoicing	Controls cash flow timing	Delayed collections
Reporting	Informs every operator decision	Blind spots in the P&L

For each load-bearing system, I want to know: Is it documented? Is it owned by a tool or a person? Can it run over a holiday weekend without intervention? The owner’s dashboard post covers what good automated reporting looks like — I use that as a benchmark when I evaluate whether a company’s numbers are even visible in real time.

How do I test for founder-dependence?

Founder-dependence is the single biggest valuation discount I apply. It is also the most underreported risk in a standard financial audit because it does not show up on the balance sheet.

The test is simple: I ask the founder to walk me through their last week, hour by hour. Then I flag every task that required their personal login, their personal judgment, or their personal relationship to execute. If that list is long, the business has not been systematized — it has been personalized.

Specific red flags:

• Single-login tools — If the founder is the only admin on the CRM, the phone system, or the scheduling platform, access disappears with them.
• Undocumented SOPs — If the answer to “how does X get done” is “I just know,” that knowledge is not transferable.
• Relationship-gated vendor terms — Pricing or service levels that exist because of a personal relationship, not a contract.
• No delegation history — A founder who has never successfully handed off a repeatable task to a system or a person cannot scale.

The human in the loop post is a useful frame here: there are tasks that should stay human, and tasks that should not. The problem is when founders have not made that distinction deliberately.

What does a healthy automation layer look like?

A well-automated business has a few consistent characteristics. Repeatable work is handled by triggered systems, not scheduled humans. The founder can describe their role in terms of decisions and strategy, not execution and follow-up. New volume — more leads, more customers, more locations — can be absorbed without proportional headcount growth.

When I find this, I treat it as a genuine asset. Systems are scalable in a way that people are not. A business where the core loop runs without daily founder input is a business where capital can actually do work.

In the businesses I run and the companies I evaluate as of September 2026, the clearest predictor of post-investment scalability is not revenue growth rate — it is the ratio of load-bearing processes that run on documented, triggered systems versus ones that run on a specific person’s daily action. A company with strong revenue but a fully manual core loop is a job, not a business. A company with moderate revenue and a systematized core loop is a platform. The automation audit is how I tell the difference before capital is committed, not after.

If you are preparing a company for investment — or evaluating one — this is the work that belongs in the data room alongside the financials. I do this formally as a Fractional Chief Automation Officer, and the findings change deal terms more often than most financial metrics do.

For a deeper look at how I approach this across a portfolio, the investment partner page covers how I engage with operators and investors directly.

Want to talk through what an automation audit would surface in a specific company? Start a conversation with the voice agent on this site — it is the fastest way to get time on my calendar.