The blank page
The first marketer starts from nothing.
The method lives in a deck, a few calls and one partner's memory. Each new hire rebuilds it from scratch, and the founder pays for the same lessons the last company already paid for.
AI operating system
Company OSfor venture capital
Run it on a company before you invest, and see where its growth is waiting. Install it after you invest, so every company in the portfolio runs its go-to-market on what is true about its own business, with the method your best operators would write down if they had the time.
One system does both jobs, so nothing from diligence is thrown away. The growth read becomes day one of the install.
The problem
Every company in the portfolio learns go-to-market on its own.
Your platform team knows what works. Each company still hires its first marketer, opens a blank page and works it out again, one company at a time.
The exit interview
What they learn leaves with them.
A year of tested messages, buyer objections and channel results sits in one person's inbox and chat history. When they move on, the next hire starts again.
The average answer
Everyone's AI writes the same thing.
Founders and their teams all use AI, each in their own way. Without the company's own facts, customers and voice underneath it, the output reads like every competitor's, and nobody can check it.
Each company should start from what the portfolio already knows, and keep what it learns itself.
01Before you invest
What growth is this company leaving on the table?
Before the term sheet, we run the first phase of a Company OS install on a company you are assessing, from the outside. You get a growth read: where its buyers' demand is, how much of it the company captures, what its customers actually say, and what would move first with the right system behind it.
From the public web
Six questions answered without asking the founder for anything.
Market
Who do they actually compete with?
The real competitor set, taken from search results, review sites and buyers' own comparisons rather than the pitch deck. Each profile ends in the positioning nobody in the category owns yet.
Customers
What do their customers praise, and what do they complain about?
Public reviews, forums and case studies, sorted into themes with the customers' own words kept. It shows whether the story the company tells matches what its users experience.
Buyers
Who buys, and is the marketing written for them?
A draft buyer profile built from the site, case studies, job ads and reviews, set against who the website and ads are actually speaking to.
Search demand
How much buyer demand exists, and how much do they win?
A keyword map of the category with monthly search volumes by market, set against where the company ranks today. The gap between the two is the size of the organic opportunity.
AI answers
Do AI assistants recommend them?
We put the questions buyers ask to ChatGPT, Gemini, Perplexity and Claude, and score whether the company is named, in what position, and whether what is said about it is right.
Site and presence
Is the site built to be found, and are they listed where buyers look?
A technical and on-page audit with real speed measurements, plus the review sites, directories and comparison articles buyers read before they make a shortlist.
With the founder's access
In the data room, three more answers from the company's own systems.
Paid search Read only
Is the ad budget buying demand, or waste?
The Google Ads account checked against our 89-point registry: conversion tracking first, then the spend going to searches, markets and networks that never turn into buyers.
Search and analytics Read only
What is actually arriving on the site?
Real clicks, queries and conversions replace the public estimates, so the opportunity sizes become a baseline you can measure the company against after you invest.
CRM Read only
Where do leads stall?
Lead sources, stage conversion and time to first meeting, read in aggregate. No contact's details leave the company's system, and nothing in it is changed.
The founder grants this access and can remove it at any time. Every connection is read only, and we change nothing in the company's accounts.
A product its customers love, missing from the searches its buyers make every month.
The read Scored out of 10, from public sources
| Category demand | 8 | About 3,900 buyer searches a month across Australia and New Zealand | |
| Search capture | 2 | Top ten for 4 of 46 buying keywords | |
| AI answers | 3 | Named in 1 of 10 buyer questions. Provisional | |
| Customer voice | 8 | Reviews praise setup speed. The website never mentions it | |
| Positioning | 4 | The site speaks to clinic owners. The reviews come from practice managers | |
| Site foundations | 6 | Fast on mobile. Twelve product pages share one title |
Where growth comes from first Ranked by impact, then effort
- Build the comparison pages buyers already search forLow effort The two competitors winning these searches have thinner products and better pages.
- Write the message for the person who actually buysMedium effort Practice managers choose the software. Their reviews already supply the words.
- Get listed on the review sites AI assistants quoteMedium effort Absent from three of the four sources cited most in AI answers for this category.
What we could not see from outside: paid search performance, real traffic, and how many leads become demos. These come from the founder's access, in the data room or after you invest.
It stays quiet
The company does not see you looking.
Our scripted checks block the company's analytics, ad pixels and CRM tags from the first page load. We never fill in a form, book a demo or accept a cookie banner, so nothing turns up in their dashboards or their sales pipeline.
It knows its scope
Go-to-market, and nothing else.
It is not financial, legal or technical diligence. Every finding drawn from public signals is labelled as such, and scores built from outside are a starting read, confirmed with the founder later.
It carries forward
If you invest, it becomes day one.
The growth read is the first phase of the company's own Company OS. Its draft foundations already exist, so the install begins with the company's own material instead of a blank page.
02After you invest
Put an operating system inside each company.
The same system we install for our own clients, built inside each portfolio company and owned by it. The team keeps the AI they already use. Underneath it sits one source of truth about the business, and the workflows that run its go-to-market.
-
Phase 1
A draft from the public web
Competitors, customer voice, buyer profile and brand voice, drafted from public sources. Nothing at draft goes into live work.
Done in diligence -
Phase 2
The company's own material
Call recordings, documents, emails and CRM exports folded in. Each fact is filed once, and anything contradictory goes to a person to decide.
From the company -
Phase 3
Confirmed by the founder
A discovery call goes through every file. Only what the founder confirms is built on.
Founder sign-off -
Then
The workflows run
Content, SEO, AI visibility, LinkedIn, outbound, Google Ads, lead magnets and reporting, switched on in the order the company needs them.
Ongoing
What the founder keeps
It is their company. The system is built to stay theirs.
Ownership
The company owns its OS.
It lives in the company's own workspace. If the company leaves the program, or the portfolio, the system goes with it intact.
The final say
Nothing goes out without a person.
AI never publishes, sends or spends on its own. Every post, email, page and ad change stops for the company's approver, and is released by a human.
Their voice
Each company sounds like itself.
The brand voice, and the voice of each founder who posts under their own name, is built from their own writing and confirmed by them.
03Across the fund
Method shared. Data never.
Across a portfolio the system has two layers. The method (agents, workflows, checks and report formats) is kept by the fund and improves with every company. Each company's knowledge stays in its own OS and never crosses to another.
Compounding
A lesson learned once is used everywhere.
When a cold email angle wins at one company, or a check catches a broken conversion tag, the lesson goes into the shared method. Every company after it starts from there.
One way of reading
Every company reports the same way.
The same weekly and monthly report formats at every company, so your platform team reads the whole portfolio one way. Each founder decides what is shared with the fund.
Separation
No company sees another.
Names, numbers and customers never cross between companies, in any report, page or example. A comparison is described, never named. Our own client work already runs on this rule.
Questions investors ask
Before you bring it to a founder.
Does the founder need to agree?
For a growth read, no. It uses public information only, and it leaves no trace in the company's systems. For an install, yes. It is their company and their system, and it only works when someone there wants it.
Who can see a portfolio company's data?
The company. We work inside its own OS on its behalf. The fund sees what the founder chooses to share, such as the monthly report, and nothing from one company is ever shown to another.
Does it replace hiring a marketer?
No. It means the first marketer starts on day one with the company's facts, customers and voice already in place, and the method already written. People stay in charge of everything that goes out under the company's name.
Can our platform team run it themselves?
That is where it ends up. We build the method layer with your team, run it alongside them across the first companies, and hand over as they take it on.
Which AI does it use?
Whatever each company already uses. The OS is the knowledge and the method underneath, kept in files the company owns, so it is not tied to one model or one vendor.
Working with us
Three ways in. Start with one deal, one company, or the whole fund.
Most funds start with a growth read on a company they are already looking at. It is a fixed fee, it answers a real question on a live deal, and if you invest, the install has already begun.
For a deal you are assessing
Growth read
One company, from the outside.
A go-to-market read on a company you are assessing, from public information only, delivered to the deal team as a confidential report.
What's inside
Competitors · Customer voice · Buyer profile · Search demand · AI answers · Site audit
Fixed fee · public data only
For one portfolio company
Pilot
One company, from the inside.
We install Company OS in one portfolio company and run its first workflows, so you and the founder see it working before anyone commits further.
What's inside
Foundations build · First workflows · Approvals · Reporting
Scoped to the company · they own it
For the platform team
Fund-wide
Every company, one method.
The shared method layer for your platform team, plus an OS in each company that wants one. We run it with your team until they run it themselves.
What's inside
Method layer · Growth reads on new deals · An OS per company · Portfolio reporting
Scoped to the fundBook a conversation