Failing projects get folded into a larger, successful program shortly before review, so their losses disappear into the host's numbers and their sponsors avoid both a kill decision and accountability.
An example
Losses on a paid puzzles-and-games app at a digital media company reached $11 million in its first year, against a plan to break even. Six weeks before the annual portfolio review, the app was moved into the streaming segment, which had grown 30 percent and carried enough margin to absorb the loss inside normal variance. The memo called it a natural consolidation of subscription products. A finance director preparing the segment results saw that the app's cost center had been merged into streaming's and its separate reporting closed. The portfolio review discussed streaming as a success. No one asked about the app. The finance director's question about restating the pre-merger baseline was routed to the segment's own controller, who reported to the sponsor, and it was answered with a note that the segment was now evaluated as a whole.
Why it works
Once the project sits inside a winner, continuing it costs the sponsor nothing: there is no line item to defend and no kill decision to face, which makes the dilution easy. A clean post-mortem on the standalone project would have produced learning, accountability, and capital reallocation, and dilution destroys that by removing the object to review. Anyone who wants the failure visible must separate shared budgets, staff, and milestones, which is costly and easy to call a witch hunt. Only someone who controls reporting boundaries can do it, and it also requires a host large enough to absorb the loss and auditors who look only at program level.
The thinking behind it
George A. Akerlof George Akerlof's The Market for "Lemons": Quality Uncertainty and the Market Mechanism (1970) shows that when one side knows the quality of what is offered and the other cannot tell, the uninformed side cannot judge it properly. After the merger, the sponsor knows how the absorbed project performs and the reviewers see only the combined numbers. Preserving the pre-merger baseline and having an auditor disaggregate the results restore the information the review needs.
Richard H. Thaler Richard Thaler's Mental Accounting and Consumer Choice (1985) shows that people judge a sum against the account it falls in, so the same amount can feel trivial in one account and unacceptable in another. An $11 million loss reported on its own is a failure. Inside a growing segment it becomes variance. Keeping evaluation units fixed below the level where dilution happens keeps the loss in an account where it is still visible as a loss.
Warning signs
Program or segment boundaries are redrawn shortly before a portfolio review, and an underperformer ends up inside a winner. The move is described as a natural synergy or a consolidation for efficiency. The absorbed project's separate reporting stops, its pre-merger baseline is not carried forward, and questions about it go to people who report to its sponsor.
How to defend yourself
Get the baseline before the merger. If you have any role in finance, audit, or the review process, ask for the absorbed project's pre-merger plan and actuals to be preserved and reported separately for at least the next two review cycles. Ask an independent auditor to disaggregate the project's costs and results from the host's. Frame the request as normal evaluation practice and keep it about the numbers, since the sponsor will try to make it about motives. Dilution trades a visible loss for a hidden one, and a hidden loss tends to come back larger; a dated request on file protects you when it does.
How a leader shuts it down
Pre-register evaluation units and lock baselines before any reorganization of programs or segments, and require independent disaggregation when an underperformer is absorbed. Evaluate below the level where dilution happens. Take away the sponsor's unilateral control over reporting boundaries, since the tactic is unavailable to anyone without it.