A Shortfall in Year One Costs More Than One Year of Value

← Back · ACT II · Chapter 12: Compounded Value

The industry has decided that value needs tending after go-live. Prosci writes about governing benefits beyond go-live, ERP practitioners hold post-implementation reviews, and consultancies now sell value-realization offices. The instinct is right, but the usual reason given for it, that benefits leak, undersells the problem. Value compounds, and compounding is unforgiving about the years you skip.

A deployment is supposed to bend a curve. In the first year the organization realizes some of what it promised. If it measures, learns and adjusts, the second year builds on the first and the third builds on the second, and the distance between the company that governs and the one that does not widens every cycle. The return gets built over time rather than banked at purchase, and time is the term nobody puts in the model. The same thing happens inside a single year when improvements act on one another, which is the subject of Business Cases Add Up Gains That Actually Multiply.

What a Skipped Year Costs

Now skip a year. Leave the first year ungoverned, the outcome unmeasured, the adjustment unmade. The organization loses more than that year’s increment; it loses the base the second and third years would have compounded on, and the loss keeps growing from the point it was allowed.

The books cannot see this. Financial statements record one period at a time, so a shortfall appears as a single-period variance, explained, absorbed and closed. The compounding loss it set in motion never shows up as a line, because no statement in the enterprise carries value across years the way it carries a fixed asset. Most organizations keep a running record of what they spent and reconcile it every quarter, while nothing comparable exists for what they realized.

Ongoing Attention Is Not the Same as a Record

This is where the market’s answer falls short of its own diagnosis, and FinOps is the clearest case. In February 2026 the FinOps Foundation widened its mission from managing the value of cloud to managing the value of technology, and reported that 98 percent of the 1,192 practitioners it surveyed now manage AI spend, up from 31 percent two years earlier. The discipline is reaching, correctly, for ongoing value management. But it reaches with the instruments it has, which read cost and usage telemetry period by period. Those instruments can say what was spent this month. Whether the value is compounding only shows up on a multi-year record of realized value, and almost no organization keeps one.

A value-realization office that meets quarterly and counts activity does not fix this, and a governance forum does not either. What closes the gap is a record: a figure for the value each material investment has realized to date, against a denominator set when the money moved, owned by the buyer and carried forward year over year. On that record a missed year shows up when it happens instead of hiding inside a period variance. Without it, the next round of spending tends to get funded on the last round’s unverified savings, which is a problem of its own.

A vendor whose deployment does keep building value has the most to gain from that record, since it is the only document that shows the second and third years paying off rather than leaving them to be argued about at renewal.

Most organizations now say they govern their AI, their ERP or their transformation after go-live. A practical test of that claim is whether, for the largest deployment, someone can produce a single number for the value realized to date on a denominator set at purchase. Where nobody can, what is being governed is the spend.

Reference: FinOps Foundation, State of FinOps 2026 survey release (February 19, 2026).

Keep reading: ACT II · Chapter 12: Compounded Value