📋 PLAYBOOK
Every trace stops at go-live except the one somebody kept measuring.
📍 IN BRIEF
By the end of this you will be able to operate the return on your platform as a rhythm rather than a one-off harvest: a business case re-anchored with finance, a handful of measures the business can actually influence, a Centre of Excellence that owns the loop, a review cadence that makes decisions, and a handover that keeps benefits alive after the project closes.
Every platform owner eventually meets the same question, usually in a budget round and usually without warning. The platform cost this much, so where is the return? The organisations that struggle to answer are rarely the ones that delivered badly. They are the ones that treated ROI as something you harvest once at go-live, collected the early savings, and stopped measuring. A return on an enterprise platform behaves less like a harvest and more like a crop under cultivation, it keeps coming only while someone keeps tending it. Most large organisations have watched at least one confident business case fail to show up in the numbers afterwards, and the analysts have been making the same point for years, a large share of digital programmes never demonstrate the outcomes they were funded on. The difference between the platforms that keep earning and the ones that quietly become a cost line is not the technology. It is whether value is run as a standing discipline with an owner, a baseline and a cadence, or filed as a closing slide in the project deck.
What follows is the operating loop, five moves in sequence, each with the place it usually breaks.
When to use this
This is for platform owners and CoE leads whose ServiceNow platform is live and carrying real workload, and who are being asked to show a return rather than a roadmap. If you are still pre-launch, start with the business case and come back, because a value loop with nothing in production is a forecast, not a measurement. It also assumes you have, or can convene, some form of central platform function, even a small one, since several of the moves need a home that outlives any single project.
Stage 1, re-anchor the business case before you optimise anything
Take the document that won the funding and turn it into the document you measure against.
The original business case was written to win an investment decision, which means its figures were advocacy, tuned to clear an approval bar. That is normal and not a flaw, but it makes those figures useless as baselines until they are re-examined in daylight. Re-anchoring means refreshing the cost, return and time-to-value estimates against what actually got delivered, agreeing with your finance partners exactly how each benefit will be counted, and taking the revised case back to the sponsor for a fresh endorsement. The conversation with finance is the load-bearing part, because a benefit whose counting method finance never agreed will be argued about later instead of counted, and an argued benefit is worth nothing in a budget round.
The gotcha here is the souvenir business case, the one that won the money, was filed the same week, and has not been opened since. When the ROI question lands, the team reverse-engineers a justification from whatever the platform happens to report, and everyone in the room can tell. If your business case has not been touched since approval, that is the first hour of work, before any framework, any dashboard and any committee.
Stage 2, choose measures the business can influence and feel
Pick a small set of metrics that sit close to the business process, and baseline them before you change anything.
The useful measures live where the work happens, not where the platform reports. Map the process steps that produce each outcome you claimed, then choose metrics that reflect an activity the organisation can genuinely move, the rate at which issues are resolved on first contact, the elapsed time from request to fulfilment, the share of demand arriving through the channel that was meant to absorb it. Translate each into money using four plain categories, direct cost savings, cost avoidance, revenue benefit, and intangible value. Then record the baseline from historical data and validate the whole set with finance, using peer benchmarks to sanity-check the targets. A handful is enough, because the discipline of choosing which numbers matter is itself half the value, and the difference between the numbers that predict and the numbers that merely record is one this publication has covered in leading and lagging indicators.
Value category | What it looks like on the ground |
|---|---|
Direct cost savings | Licence, tooling or contract spend actually retired because the platform replaced it |
Cost avoidance | The hires and renewals you did not need as demand grew, priced honestly |
Revenue benefit | A revenue-carrying service reaching customers faster than it did before |
Intangible value | Fewer audit findings, better employee experience, risk you can show has shrunk |
The gotcha is the vanity dashboard, the one built because teams instinctively measure what the platform team can move, tickets closed, stories shipped, workflows launched, and none of it is felt by the business. The test is blunt, if a metric never appears in an operational review outside your own team, it is the wrong metric, however good it looks.
Stage 3, make the Centre of Excellence own the loop
Give the value work a permanent home with a name against it, not a task list spread across projects.
A Centre of Excellence exists to consolidate strategy, design authority, technical governance, operations and data standards into one function, which is precisely the set of vantage points a value loop needs. So mandate it explicitly, the CoE owns the baselines, the measurement, the reporting and the recommendations, and a named person inside it is accountable for each claimed benefit. This is not an invented burden bolted onto the day job. In a full enterprise delivery method, value management and analytics is a workstream in its own right, carrying around 8 - 10% of the overall programmes delivery tasks. Organisations that skip it are not saving effort, they are declining a workstream the method considers structural. Wire the intake side too, every idea and demand that enters the pipeline should carry its expected value at the door, the way a well-run demand board already insists on, so the loop begins before a single thing is built.
✅ PRINCIPLE
A benefit is a line with a name against it. If you ask who owns a claimed benefit and the room looks at the PMO, the loop is not running, whatever the dashboard says.
The gotcha is the CoE staffed as an org chart rather than an engine, the one that has the authorities, the forums and the standards, and value belongs to everyone, which day to day means nobody. The fix costs one meeting, put a named owner against every benefit line in the business case, inside the CoE or the business, and publish the list.
Stage 4, run a cadence that decides, not one that reports
Put value on a fixed rhythm, monthly or quarterly, and oblige every review to change something.
The loop earns its keep at the review. On a fixed cadence, the CoE brings the measures against their baselines to senior leadership, alongside a regular platform health review, and governance acts on what the numbers say, because acting is the point. A working cadence retires things, the capability nobody adopted, the duplicated spend the cost review keeps finding, redirects things, investment moving toward the outcomes that are actually responding, and reinforces things, the pattern that worked in one business unit funded to spread to three more. The strongest operators also refine the governance itself on a slower cycle, quarterly or twice a year, so the machinery keeps fitting the organisation it serves.
⚠️ COMMON PITFALL
The reporting ritual. The deck is presented, leadership nods, nothing is retired, redirected or reinforced, and the same deck returns next quarter with the numbers lightly aged. If a value review has no decision log, it is not governance, it is a meeting about a dashboard.
The gotcha is the ritual itself, and the cure is a standing rule, every review must retire, redirect or reinforce at least one thing, and a review that does none of the three gets recorded as having made no decision, visibly, in front of the people who fund the platform.
Stage 5, hand value to the business before the project closes
Transfer each benefit to a business owner while the project team still exists to do the handover properly.
Projects end, and the return does not stop needing owners when they do. Before closure, the value plan and its baselines pass formally to the process and product owners who live with the outcomes, with three things made explicit, where the business case, measures and value roadmap live, who now owns each benefit line, and when the recurring value health checks happen. Keep measuring through the stabilisation period after go-live, resolve the defects that suppress the benefits before declaring business as usual, and keep the leadership reporting running on its cadence after the project team has gone. This is the move that separates a return that compounds from a return that decays.
The gotcha is the orphaned benefit. The project closes clean, the team disperses, and the measurement discipline leaves in the same taxi. Six months later the number has drifted, nobody noticed because nobody owned it, and the next business case inherits a credibility problem it did nothing to earn. A benefit with no owner does not hold still, it erodes quietly until someone asks where the return went.
📋 QUICK REFERENCE
The ROI operating loop, in five moves:
Re-anchor - Turn the funding case into the measuring case, counting rules agreed with finance, sponsor re-endorsed.
Measure - A small set of business-felt metrics across the four value categories, baselined before anything changes.
Own - The CoE runs the loop, and every claimed benefit has a named owner.
Decide - A fixed review cadence where every session retires, redirects or reinforces something.
Hand over - Benefits pass to business owners before project close, with health checks on the calendar.
Common failure modes
The souvenir business case. Filed at approval and never opened again. The fix is to re-anchor it with finance before you claim anything, and to treat the counting rules as the deliverable.
The vanity dashboard. Metrics the platform team can move but the business cannot feel. The fix is to keep only the numbers that appear in someone else's operational review.
The reporting ritual. A cadence that presents and never decides. The fix is the retire, redirect or reinforce rule, with the no-decision outcome recorded where the sponsors can see it.
The orphaned benefit. Ownership evaporates at project close and the return erodes unwatched. The fix is a named business owner per benefit line and scheduled value health checks, agreed before the project team stands down.
The bottom line
If you protect one thing, protect ownership of the loop after go-live. Frameworks, dashboards and committees all decay gracefully, but a benefit nobody owns decays silently, and silent decay is what turns a well-delivered platform into next year's cost-cutting target. The platform will keep producing the raw material of a return either way. Whether that turns into ROI your organisation can point to depends on whether anyone is still weighing it.
A return you stopped measuring is a return you stopped earning.
P.S. One drill before your next budget round. Open last year's business case and count how many benefit lines have a named owner who could tell you the current number without looking it up. That fraction is your real value maturity, and it is better to learn it from your own drill than from the finance director's question.