The Numbers We Stopped Reporting (2026)
The metrics that survived contact with real decisions and the ones that did not. A pattern from many engagements: which numbers change decisions and which…
Across enough engagements, a pattern emerges: certain numbers show up on every founder’s dashboard and change no decision they ever make, while a smaller set of numbers quietly drive everything. Over time we stopped reporting the first kind, not out of principle but because they failed the only test that matters, whether a decision ever changed because of them. This is the graveyard of numbers that did not survive contact with real decisions, and the short list of ones that did.
Published: July 2026
Some numbers die because they flatter without informing. Total registered users is the classic: a big, growing, satisfying number that says nothing about whether the business works, because a registered user who never pays or returns is not a customer, just a signup. Cumulative anything, cumulative revenue, cumulative downloads, cumulative sign-ups, belongs beside it, because a cumulative number can only ever go up, which means it can never signal a problem, which means it cannot inform a decision. A metric that cannot fall cannot warn you.
Blended CAC in isolation dies too, because averaging the cost of your cheap organic customers with your expensive paid ones hides the truth about each channel, and the blended figure flatters the channels that are actually losing money. Revenue without margin is a persistent zombie: a revenue number quoted with no reference to what it costs to earn, which tells you the business is busy but not whether it is profitable. And GMV quoted as revenue is the most misleading of all, presenting the total value flowing through a platform as though it were the business’s own income, when the business only ever keeps a small take of it. Each of these appears on dashboards constantly, and each has changed approximately no decisions, which is why they left our reporting.
The common cause of death is the same: no decision ever changed because of the number. That is the whole diagnosis. Total registered users never made a founder do anything differently, because the number went up regardless of whether the business was healthy. Cumulative figures never triggered an action, because they could not fall. Blended CAC actively misled channel decisions by hiding the per-channel truth. Revenue without margin prompted celebration that the margin numbers would have tempered. GMV-as-revenue created a sense of scale that the actual take-rate did not support.
A number earns its place in reporting by changing decisions, and these did not. They survived on dashboards for other reasons, they were easy to measure, they felt good to report, they made the business look bigger, but none of those reasons is a decision changed. Once you apply the test, the graveyard fills up quickly, and the reporting gets shorter and more useful in equal measure. This is the same anti-sprawl discipline as most dashboards die in a month and a well-built KPI tree.
The numbers that survived did so because a decision hangs on each. The 13-week cashflow low point survives because it drives real decisions constantly: whether to hire, when to pay, whether a stock buy is affordable. Contribution per unit, order, or seat survives because it tells the founder which products, orders, or customers actually make money, which drives pricing and mix decisions. Payback months survives because it measures how long cash is locked up before a customer turns profitable, which gates acquisition spend. Net revenue retention survives because it says whether the existing base grows or shrinks on its own, which is the truest signal of durability. And pipeline coverage survives because it tells a services or sales-led business whether the work supports the capacity, which drives hiring.
The pattern is exact: every survivor has a decision attached, and every casualty did not. The survivors are less flattering than the casualties, a 13-week low point is a sobering number, contribution is often lower than founders hope, but they are the numbers that actually run a business, precisely because they can deliver bad news and therefore inform action. Detail on several of these sits in net revenue retention and unit economics for ecommerce.
The test that sorts the survivors from the graveyard is one question: what would you do differently if this number halved? If a metric halved and you would change a decision, it is a real number, keep it. If a metric halved and you would shrug, because it does not connect to anything you control or decide, it is a vanity number, and reporting it is costing you attention that the real numbers deserve. Total registered users halving would alarm a founder emotionally but change no decision; a 13-week low point halving would change several immediately. That difference is the whole test.
The discipline is in the stopping, which is harder than the starting. Adding a metric feels like progress; removing one feels like losing information, even when the information never did anything. But a reporting set that keeps every number that ever felt interesting buries the few that drive decisions under the many that do not, which is exactly the dashboard death that this pattern is drawn from. The numbers we stopped reporting were not wrong; they just never changed a decision, and a number that never changes a decision has no business taking up the space where the ones that do should be. Stopping is the discipline, and a shorter, sharper reporting set is the reward.
What is a vanity metric?
A number that flatters without informing: it looks good, tends only to rise, and changes no decision. Total registered users, cumulative anything, blended CAC in isolation, revenue without margin, and GMV quoted as revenue are the common examples. They survive on dashboards because they are easy to measure and feel good to report, not because any decision ever changed because of them.
Why is a cumulative number a vanity metric?
Because a cumulative figure can only ever go up, which means it can never signal a problem, which means it cannot inform a decision. A metric that cannot fall cannot warn you of anything. Cumulative revenue, downloads, or sign-ups always look like progress regardless of whether the business is healthy, so they change no decision and earn no place in operating reporting.
Which numbers actually matter?
The ones with a decision attached: the 13-week cashflow low point (drives hiring, payment timing, affordability), contribution per unit, order, or seat (drives pricing and mix), payback months (gates acquisition spend), net revenue retention (signals durability), and pipeline coverage (drives capacity decisions). Each survives because a decision hangs on it, and each can deliver bad news, which is precisely why it informs action.
Why is GMV quoted as revenue misleading?
Because it presents the total value flowing through a platform as though it were the business’s own income, when the business only keeps a small take of it. GMV-as-revenue creates a sense of scale the actual take-rate does not support, which is why it is the most misleading of the common vanity metrics. The number that matters is the revenue the business actually keeps, not the value that flows past it.
How do I test whether a metric is worth keeping?
Ask what you would do differently if the number halved. If you would change a decision, it is a real number worth keeping. If you would shrug because it does not connect to anything you control or decide, it is a vanity number costing you attention the real numbers deserve. Total registered users halving changes no decision; a 13-week low point halving changes several.
Why is it so hard to stop reporting a number?
Because adding a metric feels like progress and removing one feels like losing information, even when the information never did anything. The discipline is in the stopping. A reporting set that keeps every number that ever felt interesting buries the few that drive decisions under the many that do not, which is exactly how dashboards die. A shorter, sharper set is the reward for the harder discipline of stopping.
Sydney Virtual CFO is a Sydney-based virtual CFO service for founders running $2M to $15M businesses across SaaS, ecommerce, professional services, construction, and other low-volume, high-value industries. We deliver fixed-scope CFO engagements with a named deliverable on day 90: a 13-week cashflow forecast, a fundraise-ready financial model, a unit economics build, or a board reporting pack you can run on your own.
Our front-door product, the 90-Day Number, is fixed scope at $17,850 plus GST. We are one of the few project-based virtual CFOs in Australia, in a market built almost entirely on monthly retainers. No retainers without a deliverable. No 80-page reports. No theatre.
Visit Sydney Virtual CFO | The 90-Day Number | Book a Call
This content is general information only, written for Australian founders running businesses in the $2M to $15M revenue range. It does not constitute tax, financial product, investment, or legal advice and should not be relied on as such. The work referenced is led by a Chartered Accountant (CA ANZ), but Sydney Virtual CFO is not a licensed tax agent, not a licensed financial adviser, and not authorised to provide personal financial advice. Tax obligations, accounting treatments, fundraise terms, and statutory requirements depend on your individual circumstances. For advice specific to your business, contact the team directly or consult a registered tax agent, licensed financial adviser, or qualified lawyer. Information was current at the time of publication and may change without notice. We review and update guides periodically.
