Startup evaluation software with a weighted panel rubric

Two evaluators score an application out of ten. One gives it a seven because the team is strong. The other gives it a seven because the market is large. The average is seven, and the number is meaningless - it has flattened two unrelated judgements into one figure that cannot be compared with anything.

Now do that across eighty applications, five evaluators and two cohorts, and try to explain to a governing body why application forty-one was selected over application nineteen.

The fix is not more discipline from evaluators. It is a rubric that makes the criteria explicit and weights them before anyone starts scoring.

You build a set of evaluation questions. Each question has a weight, reflecting how much it should count towards the total. Each question has scored options, so an evaluator picks a defined level rather than inventing a number.

The result is that every evaluator is answering the same question, on the same scale, with the same weighting - which is what makes their scores comparable to each other, and comparable across cohorts.

The best evaluator for a deep-tech cohort is often someone who does not work at your incubator and should not have access to your funding records.

Panel members are assigned to specific programs, and can only see and evaluate applications inside those programs. They can be added without being granted a full portal role - so bringing in an external domain expert for one program is a small act, not a security decision.

Evaluation scores are only half of a defensible selection process. The other half is the record of what happened to each application and who did it.

Applications move through twelve statuses, and every change is recorded with the person and the timestamp. Requests for changes and requests for documents each carry their own state, so an application that stalled has a visible reason. Combined with the scoring breakdown, you can reconstruct any selection decision without relying on anyone's memory.

Evaluation is not a standalone step. The rubric is defined while you are building the program, alongside the application form, so the questions you ask applicants and the criteria you judge them on are designed together.

After selection, the same records carry forward - the startup that was evaluated becomes a startup in a batch, with its funding, tasks, office allocation and mentorship all hanging off the same profile. Nothing is re-entered.