Topic 63 of 64· Software engineering and digital delivery connection

Benefits Realization

Benefits realization management is the discipline of identifying, baselining, tracking, and evidencing that the benefits promised in a business case actually materialized after go-live. In UK public investment it lives inside HM Treasury's Green Book Five Case Model and the Infrastructure and Projects Authority's dedicated benefits management guidance; without it, "the system saved caseworkers thirty minutes a claim" remains an unaudited assertion forever.

Why it matters

Business cases are promises; benefits realization is the audit. The Green Book requires every spending case to pass five tests — strategic, economic, commercial, financial, and management — and the management case must set out how benefits will be realized before approval: owners named, baselines captured, and measurement dates fixed. The Infrastructure and Projects Authority's guide, Benefits Management: A Guide to Realizing Benefits for Government Major Projects (https://www.gov.uk/government/publications/benefits-management-a-guide-to-realizing-benefits-for-government-major-projects), exists because the IPA's own portfolio reporting on the Government Major Projects Portfolio has repeatedly found delivery confidence and benefits realization cited as recurring weaknesses across major programmes. A project can close "on time and on budget" against its delivery milestones while still failing to realize the benefits that justified spending the money in the first place — a distinction the IPA's guidance treats as the entire point of the discipline.

The maths

Realization rate = benefits realized / benefits forecast   (per benefit, per period)

Mechanics that make it computable:
  baseline captured BEFORE go-live (else the delta is unmeasurable)
  each benefit: named owner, metric, data source, measurement schedule
  forecast adjusted for optimism bias at appraisal (Green Book mandate)
  benefits classed cash-releasing / capacity-freed / qualitative,
  tracked and reported separately

Worked example

Local authority: a digital planning-application portal business case promised, per year: £300,000 in printing and postage overhead reduction (cash), 4,500 officer hours freed (capacity), and improved applicant satisfaction (qualitative). Twelve months post go-live:

Benefit           Forecast   Realized   Rate   Evidence
Cash savings      £300,000   £210,000   70%    finance ledger vs baseline year
Officer hours     4,500      3,200      71%    time-motion sample
Satisfaction      +8pp       +11pp      138%   applicant survey data

Actions from the review (the point of benefits realization):
cash shortfall traced to two service areas still processing paper
applications by exception → close the exception route;
next business case's optimism-bias correction raised from 10% to 25%
based on this case's forecasting error.

A 70% realization rate is not a failure — it is knowledge that lets the next forecast be better calibrated. An unmeasured case would have claimed 100% forever, and the finance team would have had no basis to challenge it.

Software engineering connection

Engineering organizations routinely approve platform and tooling investments on forecast benefit and almost never audit them afterwards — exactly the pathology benefits realization management exists to fix. The lightweight port: every proposal above a materiality threshold names a benefit owner, a baseline metric, and a fixed review date (typically six months post go-live), and realization rates from past proposals should discount how much the organization trusts a team's or vendor's next forecast. This closes the loop back to Green Book appraisal, which sets the forecast this discipline audits, and it is the same logic behind the widely reported finding that a large majority of generative-AI pilots show no measurable return — see AI productivity in the public sector — because the pilots that did return value were, almost without exception, the ones with a named, trackable benefit line from the start. It also depends on distinguishing what was actually delivered from what was actually realized — see outcomes vs outputs.

Pitfalls

  • No pre-go-live baseline: the fatal, unfixable omission — without it, no realization rate can ever be computed, only asserted.
  • Benefit orphanhood: a benefit with no named owner has no one collecting the data, and every portfolio review reports it as "broadly on track" by default.
  • Double-counted benefits across a programme portfolio: two projects both claiming the same freed caseworker capacity as their benefit — keep a single benefit register across the portfolio to catch this.
  • Realization theatre: measuring and reporting the easy qualitative wins prominently while the cash and capacity lines go quietly unexamined.
  • Confusing delivery with realization: a project closing its milestones "on time and on budget" says nothing about whether the forecast benefit ever actually occurred — the IPA's guidance treats these as two separate questions with two separate evidence trails.

Sources