March ran hot. You released a free build template and the shares kept coming; collaboration inquiries arrived several times a day; your follower count climbed by 4,000 in a single month. May was the mirror image. New posts drew a lukewarm response, and your inbox sat empty for days at a stretch. All through May you turned over the same question: how to recapture March's momentum.
When you closed the books at the end of the month, you discovered the question had been backward. Measured by dollars collected, May's revenue was 1.6 times March's. March's commotion was traffic pooling around a free template, and the share that converted to payment was negligible. May drew no outside reaction, but a product-update note sent to your email subscribers pulled existing customers back into renewing. For two straight months you had been reading the loudest signal as the state of your business. While your gut chased the noise, the payments were happening on a quiet email list.
Measurement Has Growth Stages of Its Own
For someone working alone, this illusion is close to an occupational hazard. A company has a finance team to close the books and colleagues who will puncture a wrong hunch in a meeting. A one-person business has neither, so in the absence of measurement, that day's mood serves as the standard of judgment. To pull the mood down to earth in the moment you need numbers — but grab just any number and it becomes a standard more dangerous than the mood itself. The first task is to sort out which numbers to stand up, and in what order.
This piece divides that order into a measurement clock with four notches, M0 through M3. M0 is the no-measurement state, where the only raw material for judgment is memory and impression. M1 is the traffic stage, counting visible quantities like visits, views, and subscribers. M2 is the conversion stage, counting how many of those quantities — and what share — turned into an action such as a payment or a sign-up. M3 is the revenue-attribution stage, answering which activity and which entry point each sale this month came from.
Each notch up changes the question you can answer. M1 answers 'Are people showing up?' M2 answers 'Do the people who show up buy?' M3 answers 'What made them buy?' Because you were watching only the M1 numbers, you read March — the month traffic crowded in — as the good one. The M2 numbers raise May's hand instead. The same two months flip their verdict depending on the measurement notch, which makes the question of where your business currently sits the starting point for reading any number at all.
Vanity Metrics and Verdict Metrics Are Not the Same
The prescription to raise your measurement stage is no new invention. Eric Ries (2011) proposed innovation accounting precisely because traditional financial metrics like revenue and profit don't work for an early-stage business. The method is to set a baseline, compare retention across cohorts of customers who arrived in the same window, and judge direction by how that retention shifts. What he warned against was the vanity metric. Cumulative sign-ups and total views never go down, so they make anything you do look good and can't serve as raw material for a verdict. The follower growth you stared at all month is exactly that kind of metric.
From the same concern, Amplitude's North Star framework recommends that, instead of revenue — a lagging indicator — you pick a single leading indicator that stands in for the moment a customer actually experiences value, then narrow to the three or four input metrics that move it. A business selling tool subscriptions and build courses might make 'a first payment within 90 days of an email sign-up' its star, with publishing frequency, subscription conversion rate, and email open rate as the inputs. Revenue follows as the result of those inputs moving. Robert Kaplan and David Norton's balanced scorecard was originally a tool for aligning departments, but the solo operator can shrink it to a tool for aligning the builder, marketer, and bookkeeper inside one person onto a single dashboard. The three theories start in different eras and handle different units, yet they converge on the same place: don't multiply your metrics; narrow to the few that produce a verdict.
Only when measurement reaches M2 can you finally compute unit economics — a calculation of whether money is left over not across the whole business but at the level of a single customer, a single transaction. For a subscription product, the ratio between the cost of acquiring one customer (CAC) and the money that customer leaves behind over their lifetime (LTV) plays that role, and in practice the line of health is whether LTV exceeds three times CAC. If a 10,000-won-a-month subscription keeps a customer for ten months on average, LTV is 100,000 won; if acquiring one cost 40,000 won, the ratio is 2.5, below the line. The prescription becomes whichever the numbers point to — a price increase, a longer retention period, or a lower acquisition cost. Every one of these calculations shares a single precondition: conversion has to be captured as a number. At M0 and M1, neither retention period nor conversion rate is captured, so the calculation simply doesn't stand up.
Once numbers exist, the next trap waits. The sample is small, and the person who designed the experiment and the person who judges the result are the same one person. Digging through past post performance and mining a rule like 'Thursday releases land well' is less a discovery than a fit to noise. Quant investing research showed long ago that repeatedly tuning a strategy on the same historical data accumulates results that look good only by chance. The prescription is to set aside a validation slice of data in advance and fix your hypothesis beforehand. A rule found after the fact should be recorded as a hypothesis only, promoted to a rule only after it holds up against next month's fresh data. On a low-traffic, one-person site, a half-baked imitation of statistics is the worst choice of all. Once you start assigning meaning to a difference that isn't significant, the verdict — numbers in hand — follows the mood just as it did before.
Three Things to Touch This Week
First, split the numbers you currently watch into vanity metrics and verdict metrics and write them down. Push the numbers that never go down — followers, views, downloads — to one side, and move the ones captured as actions — payments, subscription conversions, renewals — to the other. If the verdict column is empty, you are still at M1. The next move is to define your conversion event in a single sentence. 'Payment completed' or 'subscription signed up' is enough.
Second, spend 15 minutes on the same day each week writing the same three or four numbers into the same table. Visits, change in subscribers, conversion count, and dollars collected will do. There's no reason for the format to be fancy; if anything it should be identical week to week. The value of measurement comes not from one precise analysis but from numbers measured by the same yardstick lining up over time. The urge to expand to ten metrics or redraw the graph every week usually doesn't survive the third week. Cutting the numbers you track so the habit never breaks beats starting elaborate and stopping within a month.
Third, if you have more than one revenue source or spend a single cent on ads, build an attribution table. Put the entry points (blog, social, email, search, referral) in the columns and this month's sales in the rows, mark where each sale came from, then list revenue and hours spent per entry point side by side. It's fine if more than half the sales have unknown origins. The 'unknown' column shrinking month over month is itself the measure of progress. Once the table has stacked up for two or three months, you can ask yourself whether to shift hours from the low revenue-per-hour entry points to the high ones. A decision to add an activity can be made on feel, but a decision to cut one can only be made with a table like this.
From Productivity to Profitability
Even with the tools and the market in place, if the value a customer gains isn't captured as a number, the product loses any basis for claiming a price proportional to revenue. The same ability stays trapped under an hourly-rate ceiling. Put the other way, the work of raising your measurement by one notch is the work of lifting the price ceiling on the same product. The bridge that turns the output of faster hands into income is laid precisely here. A judgment that used to swing on feel between 'March was good' and 'free stuff is useless' becomes, on top of measurement, the concrete decision to 'keep releasing the templates but switch the verdict metric from download count to subscription conversions.' The difference between productivity and profitability is the difference between those two sentences.
The cost of building measurement isn't what it used to be. A decade ago, cohort retention or path-level attribution belonged to companies with analysts on staff; now, feed your payment records and subscriber list into an AI tool and a first draft comes out within minutes. What's in short supply is less the analytical ability than the design of what to measure, and the persistence to apply the same yardstick every week. Cheaper tools don't do the designing for you.
Once measurement is in place, the numbers bring a new anxiety along with them. The gap between good months and bad ones — the swing in your revenue — only now comes into view. The next installment takes up the design of that buffer: twelve months of money built not against the average but against the worst month. Once measurement is laid down, the next question is fixed: how much has to sit in your account so that a run of bad months doesn't break you?
Concept Appendix
- Innovation accounting and vanity metrics — proposed by Eric Ries (2011). The argument that an early-stage business needs a separate accounting built from a baseline, cohort retention, and directional judgment, rather than vanity metrics like cumulative sign-ups or total views that never go down. - North Star framework — Amplitude (John Cutler et al., The North Star Playbook). A product-strategy framework that says to pick one leading indicator standing in for the moment a customer experiences value, instead of the lagging indicator of revenue, and to narrow measurement to the three or four input metrics that move it. - Unit economics (LTV/CAC) — the calculation of whether money is left over at the level of a single customer or transaction. For subscription products, the line of health is whether customer lifetime value (LTV) exceeds three times acquisition cost (CAC), and it holds only when conversion is captured as a number (M2).
About the series — Insights based on the manuscript Running a Company by Yourself: A Management Framework for the Solo Founder in the Age of AI. Each installment takes up a single management decision.
