The first year-end after going independent, you open your books to prepare your income tax return. Under assets there's a single laptop, a monitor, and a secondhand desk. After depreciation, total assets come to a little over two million won — roughly $1,500. Yet when you think about what you actually built over the past year, the list looks nothing like that: a newsletter with 2,800 subscribers documenting the build, a bundle of work procedures and prompts refined more than a hundred times, the rates and quality records of six freelance developers you've worked with, the contacts and transaction history of the client managers who commissioned your tools. The most valuable things don't appear on the books at all.
At first you assume the books are wrong. But this blank isn't an error — it's exactly what the accounting standards intend. A laptop can be replaced in a day. The 2,800 subscribers, the record of freelance evaluations, the build procedures all take another year. The inversion in which the thing that takes longest to rebuild is recorded most cheaply is a baseline condition of working alone. So what belongs next to the empty ledger? That question is where today's piece begins.
Your accounting books can't record the assets you made yourself
IAS 38, the international accounting standard governing intangible assets, nails down one principle: intangibles a company generates internally — brands, customer lists, mastheads — cannot be recognized as assets. They're hard to delineate, and the cost of creating them can't be measured reliably.
An odd asymmetry follows. The same brand, if you acquire it by buying another company, goes onto the books at fair value. A brand I spent ten years building is worth zero, while one someone else built becomes an asset the moment I buy it. The people who wrote the standard accepted this asymmetry to protect reliability.
For large corporations the principle is a safeguard, since it stops executives from writing up the value of their own brand. Applied to a solo business, though, it turns into a paradox. Almost everything a person working alone owns is an internally created intangible, so most of the substance of the business sits off the books. AI widens the gap further. Training data, refined prompts, and fine-tuned models have no treatment at all under current standards, so each time they're made they vanish as an expense. The more AI-made the holding, the less the books can record it — and what they do record, they erase faster.
Build an asset register with five labels of different dimensions
Accounting scholars weren't blind to this gap. The Integrated Reporting Framework that the International Integrated Reporting Council (IIRC) released in 2013 holds that corporate value flows not from financial capital alone but from six capitals — financial, manufactured, intellectual, human, social and relationship, and natural — and recommends reporting the changes in off-balance-sheet capital alongside the financials. Empirical studies followed, showing that firms with higher-quality integrated reporting enjoy a lower cost of capital. The direction converges on a single line: assets you don't record drop out of management.
So the document a solo manager needs is a labeled asset register. Separate from the financial statements, it gathers every holding I have — whether it lands on the books or not — and records each one with a verdict on its dimension. The point lies less in gathering exhaustively than in sorting by dimension. Holdings come in five kinds, defined not by category but by how their value grows.
A Cost-Saving asset (CS) is a holding that reduces my time and cost; build procedures and prompt bundles belong here. Its value equals my hourly value times the hours saved, but since my total hours are fixed, it has a ceiling. A Revenue-Making asset (RM) is a holding the market pays for, like a tool-build service sold to client companies; its value is revenue times margin, and because revenue can be decoupled from my hours, it has no ceiling. A Potential-Revenue asset (PR) is internal for now but becomes a component of a revenue asset once externalized; a freelance developer's quality record is just a management memo to me, but to a builder in the same position it skips months of trial and error. A Funnel asset (FN) is a holding I don't sell directly but that channels customers and trust downstream, like a free newsletter. An Engine asset (EN) is the infrastructure that sets the pace of the whole; when publishing automation improves, both funnel and revenue speed up with it, so instead of scoring it on its own I treat it as a multiplier on the other four labels.
Whatever can be copied gets marked down
Once you've gathered your holdings into the register, you have to run them through two sieves before pricing them. The first is the moat concept from investing. The expression Warren Buffett used in his shareholder letters was organized by Pat Dorsey and Morningstar into five sources: intangible assets, switching costs, network effects, cost advantage, and efficient scale. At least one of the five has to be working to stop the regression in which competition eats away excess returns. That theory was built for companies with factories and patents, so a solo business reads it with the materials swapped out. Trust built on your own name stands in for intangible assets; an owned audience you can contact directly stands in for network effects; data and relationships accumulated through transactions stand in for switching costs; a workflow no one else has stands in for cost advantage. This is where funnel and potential assets get their value.
The swapped-in materials aren't equally strong. The hardest is the owned audience. A hundred thousand followers on a platform can be cut off the moment the algorithm changes, but an email list survives even when the platform shakes. The cost advantage of a workflow, by contrast, wears out fastest, because a way of working that leans on a tool's features becomes most people's default in the next version. Only a workflow that leans on my own data and judgment resists the erosion slowly.
The second sieve is Jay Barney's VRIO: for a resource to become a sustained advantage it must be Valuable, Rare, hard to Imitate, and backed by an Organization able to exploit it — four tests. In the AI era, two of them get reread. On the rarity test, anything AI can produce something similar to in an hour drops out. The organization test shifts from "is there an organization to exploit it" to "is there an automated system to exploit it." No matter how good the data, it doesn't become an advantage without a pipeline that runs it every week.
One founder who went independent teaching AI proudly put 400 slides, 200 auto-generated articles, and dozens of prompt templates on his asset list. But when dozens of similar channels sprang up within six months, his views and revenue sank together. Run through the two sieves: the slides and articles can be made much the same way with the same tools, so they fail on rarity; the templates are held by most of the people who bought them, so they fail on inimitability. Meanwhile the three years of student questions and the data on which examples trip people up — things he hadn't counted as assets — pass both sieves. A holding is an asset only when the cost of imitating it is time, not money. Every time AI tools improve, a capability that was rare yesterday becomes commodity today, so a value verdict has to be reissued each time a new tool appears. Put it off, and the register drifts out of true just like the books.
Spend an hour labeling every holding you own
First, clear an hour and start by listing your holdings. Write down only what's on the books and you'll struggle to get past five, so turn over six places: content you've made, people you can reach, records you've accumulated, procedure documents, tools and environments, and credentials and track record. Open your messenger favorites and bookmark folders too, and you'll usually clear fifteen.
Second, for each holding, decide its label and move it into a four-column table: holding name, label, one number as the basis for value, and the next action. The table's weight rests on the value-basis column. For a revenue asset, write the deposits of the last three months; for a cost-saving asset, hours saved per week; for a funnel asset, monthly inflow and the number that led to a transaction. A column where you can't write a number means there's no measurement, and that itself is a finding. A holding you claim earns revenue but can't prove with deposit records gets demoted to a potential asset. Revenue not proven by deposits is still at the hope stage.
Third, once the table is filled, count the Revenue-Making (RM) rows. If there isn't a single RM row, what you're doing now isn't a business but a preparation stage — not something to beat yourself up over, but a coordinate check to accept. Pick the one funnel or cost-saving asset closest to becoming a revenue asset, and spend the next quarter moving only that candidate into the revenue column. Time spent alone runs in a single stream, so start two at once and both will drag.
From productivity to profitability
The books say zero won, but the labeled asset register sees the same company as eleven labeled holdings. The revenue numbers are unchanged, yet only now does the outline of your own company come into focus. As Paul Romer showed in endogenous growth theory, ideas are nonrival, so knowledge made once is copied at almost no added cost. The compounding of a solo business comes not from making more assets but from moving the labels on the same assets. Confirming a cost-saving asset as a potential asset and turning it into a revenue asset — this path is the bridge from productivity to profitability.
This series crosses that bridge one piece at a time. It starts from a single manuscript that reweaves the standard theories of accounting, economics, management, and investing — across disciplinary lines — into the problems of the solo business. The next piece takes up why the formulas for the cost-saving column and the revenue column differ in dimension to begin with, and why, of two holdings made with the same care, only one has a ceiling. If no amount of stacked savings ever turns into revenue, the bridge across that gap is the next piece's question.
Concept appendix
- Recognition criteria for internally generated intangibles — The international accounting standard IAS 38 (revised 1998 and 2004) bars a company from recognizing internally created brands, customer lists, and mastheads as assets, on the grounds that identifiability and reliable cost measurement are difficult — the reason a solo business's core holdings stay at zero won on the books. - Integrated reporting and the six capitals — The IIRC's Integrated Reporting Framework (2013) holds that corporate value flows from six capitals — financial, manufactured, intellectual, human, social and relationship, and natural — and recommends reporting changes in off-balance-sheet capital alongside the financials. The labeled asset register transposes this frame to the solo scale. - The VRIO framework — Jay Barney's resource-based-view test (1991, 1995): for a resource to become a sustained advantage it must satisfy four conditions — value, rarity, inimitability, and an organization to exploit it. In the AI era the rarity and organization tests get reread, filtering out whatever is copied in an hour and any data with no weekly automation behind it.
About the series — Insight ②, drawn from the manuscript 『Running a Company by Yourself — A Management Framework for the Solo Founder in the AI Era』. Each piece takes up a single management judgment.
