How do we choose between applications that all meet the criteria?

Every competitive round reaches the same point. Eligibility has filtered what it can, the scoring rubric has done its work, and there are still more strong applications than there is budget. Scores alone stop helping when several sit within a point of each other.

The first move is to accept that a tie is real rather than to invent precision. Adding decimal places to a score does not separate two applications that a panel genuinely rates alike; it only makes an arbitrary choice look calculated. Say openly that scoring has done its job and that a second criterion now applies.

That second criterion should be chosen before the round, not during it. In practice funders use four, and they are not interchangeable. Portfolio balance, to avoid funding five versions of the same project or the same region. Reach, favouring the application that serves more people or a less served population. Additionality, favouring the project that would not happen without you. Relationship history, favouring an organisation that has reported reliably before.

Each has a cost worth naming. Portfolio balance penalises an excellent application for the company it keeps. Reach favours large organisations that can count beneficiaries well. Relationship history quietly closes the door on first-time applicants, which is the opposite of what most funders say they want. Choosing one means accepting its bias deliberately rather than discovering it in three years of data.

What does not work is deferring to the room. A selection committee that breaks ties by discussion tends to reward whoever argues best or last, and produces decisions nobody can reconstruct afterwards. Discussion is useful for surfacing information the scores missed; it is a poor tie-breaker on its own.

Two safeguards belong here. Check conflicts of interest again at the tie-break stage, because this is where personal familiarity has the most room to operate. And record which tie-break rule was applied to each decision, not just the outcome.

That record is the whole point. A year later, an unsuccessful applicant or a board member will ask why one project was funded and a comparable one was not. Answering “it scored higher” is not available when the scores were equal. Answering “we apply portfolio balance, and we already funded two projects in that region” is a defensible answer, and it only exists if it was written down at the time.

A grant management platform keeps the scores, the reviewers, the decision and the reason on one record, so the basis of a close call is still available when it is questioned.