Innovation accounting

Innovation accounting measures a venture’s progress with leading indicators while its revenue is still zero. It is the scoreboard, not the plan: an innovation strategy picks the arena, an innovation portfolio holds the bets, and this counts what each one learns.

Eric Ries introduced the term in The Lean Startup in 2011. The problem it answers is blunt. A venture with no revenue line produces a finance report that says nothing useful for a year or more.

So you swap the lagging measures for leading ones — activation, repeat use, referral and the cost of reaching one more user, each read month by month and cohort by cohort.

Key takeaways

• Replaces revenue and profit with leading indicators while a venture still has no sales.
• Eric Ries introduced the term in The Lean Startup in 2011 to make learning auditable.
• Runs in three steps: set a baseline, tune the engine, then pivot or persevere.
• A sound measure moves inside one cohort cycle and cannot be bought with more spending.

How it works

Innovation accounting runs in three steps. Build a baseline from a crude first version, tune it with experiments that each move one number, then decide to pivot or persevere. Results are read by cohort, never as a running total.

A baseline is the awkward part. You have to put something real, however crude, in front of a real group of people, because a measure taken against a prototype nobody used is a guess with a decimal point.

Then you tune. Each experiment is designed to move one number, and the result stays attached to the cohort that saw it — not folded into the total.

The third step is a scheduled meeting, not a mood. Pivot or persevere gets a date in the diary, an agreed definition of what movement would count as progress, and a team that holds itself to the answer.

Pick the measures before the experiment runs. Choosing them afterwards is how a team ends up reporting whichever number happened to rise that month.

Vanity numbers are the standing risk. Cumulative signups, page views and press mentions only ever rise, so they cannot tell you whether this month’s work made the venture any more likely to succeed.

Totals hide the truth. A rising signup count can mask a falling conversion rate, which is why the internal metrics of a new venture are always read cohort by cohort.

Performance metrics built for a mature line report on something that already happened. These report on something that might, so each number is tied to a decision before anyone collects it.

Efficiency metrics still matter, but they arrive later — tightening a process that nobody wants is the most expensive mistake in the whole discipline.

The method does not replace finance. A cost-benefit analysis still gates the capital and budget variance analysis still explains the spend.

Scale is why the gap matters. U.S. spending on research and experimental development reached $789 billion in 2021, with 2022 estimated at $886 billion by the National Center for Science and Engineering Statistics.

Examples

Defence and aerospace programmes score maturity instead of sales. The U.S. Government Accountability Office built its 2020 technology readiness guide around the finding that immature technologies have led to delays and cost increases.

Pharmaceutical firms have done this for decades under another name. A phase II trial earns no revenue, so the scoreboard is enrolment rate, dropout rate and effect size, with a go or kill decision at the end.

A software venture inside a Philippine outsourcing provider measures the same way. Billing has not started, so the only figures reported upward are weekly active users by cohort and the hours of manual work removed.

A grocery chain in the United Kingdom testing a new delivery format reports basket repeat rate per launch city, because national revenue would drown the signal coming from three stores.

A bank in the Netherlands funding an internal payments tool reports adoption inside its own branches first. An internal pilot has users long before it has a price, so usage is the only honest early number.

A university spinout in Sweden reports signed pilot agreements and repeat laboratory visits. Grant income says nothing about whether anyone wants the product, so it stays out of the progress report entirely.

Related terms

• Performance metrics: the lagging measures innovation accounting temporarily stands in for.
• Internal metrics: the cohort level numbers a venture team reads every week.
• Cost-benefit analysis: the capital gate that applies once a revenue forecast is credible.
• Finance and accounting outsourcing: where the bookkeeping behind these numbers often sits.
• Forecast accuracy rate: the later check on how well those early projections held up.

FAQ

Who came up with innovation accounting?

Eric Ries set it out in The Lean Startup in 2011, as the measurement half of the build, measure and learn loop. It sits alongside the minimum viable product and the pivot in the same book.

What do you measure when there is no revenue?

Behaviour and cost: activation rate, repeat use, referral, retention by cohort and the cost of acquiring one more user. Each has to move inside a cycle short enough to act on. A useful measure is actionable, accessible and auditable.

How is it different from management accounting?

Management accounting explains money that has already moved. Innovation accounting estimates whether a bet is becoming more likely to work, which is a judgement about learning rather than a record of spending. One looks back at cash, the other forward at behaviour.

Does it replace the business case?

No. The business case still decides whether capital is released, and innovation accounting supplies the evidence that case is built from while the revenue column is empty. Treat it as the feed, not the replacement.

When should a venture stop using it?

Once sales are steady enough that ordinary reporting tells you something — usually after the first full year of real revenue. Keep the cohort view even then, because averages hide your newest customers.

Find providers who can staff the measurement and reporting work in the Outsource Accelerator index of outsourcing partners.