stepbase

By job

Four reviewers, one rubric, one decision

Review work isn't a yes/no gate. It's several people assessing the same submission against the same criteria, then someone weighing what came back and signing off. stepbase runs the scoring, holds the panel together, and records what the decision rested on.

The rubric exists. It's in a document nobody scores in.

Everyone agrees on the criteria. What breaks is the collecting: a folder of applications, a scoring sheet emailed to five reviewers, three versions back, one reviewer who used a 1-5 scale on a 1-10 rubric, and someone averaging it by hand the night before the committee meets. By the time you decide, nobody can say why applicant 14 beat applicant 9, and if a rejected applicant asks, you're reconstructing it from your inbox.

Today

  • Submissions live in a shared folder, scoring sheets live in email
  • Reviewers score in their own copy, on their own interpretation of the scale
  • Someone chases the two reviewers who haven't sent anything back
  • Totals are added up by hand the day before the committee sits
  • Comparing candidates means opening six files side by side
  • The decision is recorded as an outcome, not as a reason

On stepbase

  • Each submission arrives as its own tracked file, ready to review
  • Every reviewer opens the same scorecard and scores independently
  • The stage waits for the panel: all of them, or the number you set
  • Scores land in one place and route the file automatically by band
  • Competing entries are ranked side by side in one screen, not six
  • The sign-off, its rationale, and every score behind it are on the record

How it works

Screen

Screen before you spend panel time

A first pass checks the file is complete and eligible. Anything that fails closes on a recorded reason instead of consuming five reviewers' evenings.

Grant application #GR-118
Northlight Labs
Eligibility screen
Required fields complete
Budget page attached
Within eligible region
Under funding cap
Score

Score on one rubric

A multi-review stage gives every reviewer the same form: the same criteria, the same scale, their own independent submission. You choose whether the stage moves on when everyone has scored, when a set number have, or on the first response.

Scorecard · same for all1–10 scale
ImpactFeasibilityTeam
ALA. Linscoring…
MKM. Kohscoring…
RSR. Silvascoring…
JDJ. Diazscoring…
TWT. Wuscoring…
0 of 5 scoredwaiting on all
Sort

Let the scores do the sorting

Conditions read the scores and route on them: shortlist above a threshold, waitlist the middle band, decline the rest. Conditions can also check across reviewers, like routing any file where a single reviewer scored above 8.

Total score #GR-118
0/30
5 reviewers
Route by score band
Shortlist≥ 24
Waitlist15–23
Decline< 15
Compare

Compare the shortlist

When entries are competing for the same slot (bids, proposals, cohort places), a compare and select stage ranks them side by side so the panel picks a winner rather than approving each one in isolation.

Compare & select · one slot3 proposals
MD
Meridian
82
VG
Vantage
91
HB
Harbor
74
Sign off

Sign off, with the reason attached

The final decision is a named person's action on a specific outcome, with a required comment. Who decided, when, on what evidence, and where it went next are all captured automatically.

Final sign-off #GR-118
DOD. Okafor· Committee chair
Outcome
Award
Decline
Rationalerequired
Enter rationale…
Sign off

FAQ

Common questions

Yes. In a multi-review stage each reviewer fills in the same scorecard on their own. You can also hide specific fields from a given reviewer using the visibility settings on that stage.
You set that per stage: every assigned reviewer, a specific count like any 3 of 5, a percentage, or the first person to act. Reviews that need unanimity and reviews that need a quorum both work.
Yes. A condition block reads the scores and branches on them, for example shortlist 24 and above for the committee, waitlist 15 to 23, and decline below 15 without it reaching the committee at all.
That's a compare and select stage. It puts competing entries next to each other and the reviewer picks an overall winner, or the best entry per line item.
The full trail: every reviewer's scores and comments, who took each action and when, which criteria applied, the final decision and the rationale entered with it. It's captured on every run without anyone turning it on.

What process would you turn into an app?

Build it free, or start from a template and change what you like. No credit card, no setup fees.

Start from a template