How to evaluate grant applications fairly, step by step

The grant review process in seven steps

A grant application review process turns submitted applications into funding decisions you can explain. A fair process applies the same rules to every applicant, keeps evaluators independent and leaves a written record. Describe the steps in the call, so applicants know how they will be judged.

1. Eligibility check: formal screening with a reason for every rejection.
2. Assignment: at least two independent evaluators per application, with even workloads.
3. Blind review: hide the applicant's identity where the criteria do not need it.
4. Scoring: each evaluator scores alone against the published criteria.
5. Differences: resolve large gaps between evaluators under one rule.
6. Panel meeting: agree the decisions, starting from the ranking.
7. Decision record and feedback: write down what was decided and why. Tell every applicant the result.

Step 1: check eligibility before anyone scores

The eligibility check is a formal screening. It asks whether the application meets the rules of the call, not whether it is good. Typical checks are the applicant type, the deadline, the amount requested, the number of applications per applicant and the required attachments.

Record a reason for every application you mark as ineligible, in the words of the rule it breaks. Let a second person check borderline cases. If applicants may add a missing document later, for example within 5 working days, say so in the call.

In Juryline, you mark each application eligible, not eligible or duplicate and add a reason. The reason stays in the application's history and is not shown to the applicant. Only eligible applications go into the first jury round.

Step 2: assign independent evaluators evenly

Two evaluators per application is a sound minimum. One opinion is not a review, while two show where judgements differ. Use three for large grants.

Spread the work evenly. With 120 eligible applications, 2 evaluators each and 8 evaluators, each evaluator reads 30 applications. Mix the pairs so that no two evaluators always score the same applications. Mixed pairs make strict and lenient evaluators visible.

Every evaluator declares conflicts of interest before reading anything. An application with a conflict goes to another evaluator.

In Juryline, you choose a set number of jurors per application and Juryline shares the applications out evenly, keeping the pairs mixed. The draw uses a recorded seed, so the same input always gives the same assignment. A conflicted application goes to another juror automatically.

Step 3: use blind review where it makes sense

Blind review means evaluators do not see who the applicant is. It reduces the weight of reputation, names and personal connections in the score.

It makes sense when the criteria concern the project: the idea, the plan, the budget and the expected results. When the applicant's track record or capacity is a criterion, evaluators need to know who applies. Even then, hide what the criteria do not need, such as contact details and registration numbers.

In Juryline, fields you mark as hidden stay hidden from jurors on every application, including those already submitted. The grant template hides contact details and the registration number by default.

Step 4: score against the published criteria

Evaluators score against the criteria and weights published in the call, not against criteria of their own. A scale of 1 to 5 per criterion, with a short description of what 1, 3 and 5 mean, gives everyone the same standard.

Each evaluator scores alone, without seeing the other scores. A short written reason helps at the panel meeting and becomes the basis of the feedback. An evaluator who lacks the expertise for an application says so rather than guesses.

In Juryline, each criterion has a scale and descriptions of what the scores mean. The weights turn each evaluator's scores into a total out of 100. A juror who cannot judge an application can abstain with a reason.

Step 5: resolve large differences between evaluators

A large difference between two evaluators is a signal: one of them read the application or a criterion differently. Define in advance what counts as large, for example a gap of 20 or more points on a 100-point total. One evaluator recommending funding while the other does not also counts.

Do not average a large difference away unseen. Choose one rule and apply it to every case:

• A third evaluator scores the application independently. All three scores count.
• The two evaluators discuss the application and may change their scores, with a written reason.
• The panel discusses the application at its meeting.

In Juryline, the results show the difference between jurors for each application next to the average. A normalised ranking evens out jurors who score more strictly or more generously than the rest. You can add a third juror to a single application by hand.

Step 6: decide at the panel meeting

The panel meeting turns the ranking into decisions. Hold it after scoring has closed, so no score changes during the discussion. A chair keeps time and records the decisions.

1. The chair confirms the declared conflicts. A member with a conflict leaves the discussion of that application.
2. The panel reviews the ranking and the available budget.
3. It discusses the applications near the funding line and those with large differences.
4. It agrees the decisions: fund, fund with a reduced budget, reserve list or reject.
5. It records every departure from the ranking with a reason.

Step 7: record the decision and give feedback

The decision record shows how each decision was made: the rules, the evaluators, the declared conflicts, the scores, the ranking, the decisions and the reasons for any departure from the ranking. Keep it with the programme's files for your board or funder.

Every applicant gets the result, including those found ineligible. Useful feedback follows the criteria: what was strong, what was weak and what would make the application stronger. It does not name the evaluators.

Before the call opens
[ ] Eligibility rules written as conditions anyone can check
[ ] Criteria, weights and scale published in the call
[ ] Evaluators per application decided (at least 2)
[ ] Rule for large differences decided

Screening and evaluation
[ ] A reason recorded for every ineligible application
[ ] Conflicts of interest declared before reading
[ ] Workloads even and pairs mixed
[ ] Large differences resolved under the agreed rule

Panel meeting
[ ] Scoring closed before the meeting
[ ] Members with a conflict left the relevant discussions
[ ] Departures from the ranking recorded with reasons

After the decision
[ ] Decision record filed
[ ] Result and feedback sent to every applicant

In Juryline, you lock the round and create a jury protocol in PDF. It holds the settings, the assignment and its seed, the declared conflicts, every score with comments and the decisions, sealed with a verification code anyone can check. If you allow feedback, applicants read it in their account without the jurors' names.

Questions

How many evaluators should score each grant application?

At least two. Two independent scores show where judgements differ. For large grants or strong disagreement, add a third.

Should grant evaluators see who the applicant is?

Only if the criteria need it. When the criteria concern the project, hide names and contact details. When capacity is a criterion, show who applies but keep contact details hidden.

What counts as a large difference between evaluators?

Set the threshold before scoring starts. A workable one is 20 or more points on a 100-point total. Resolve every such case under the same rule.

Is the ranking the final decision?

No. The ranking is the panel's starting point. The panel can depart from it, for example to stay within the budget, but it records the reason.

What should feedback to unsuccessful applicants include?

The result and two or three points on what was strong or weak, following the criteria. Keep the evaluators anonymous and the tone factual.