Idea management process, seven stages from submission to implemented

An idea management process is the set of stages every employee idea passes through, from the moment someone writes it down to the moment the company can say whether it made a difference. Without a defined process, ideas pile up in inboxes, the same proposal is reviewed three times, and nobody can explain why one idea was funded and another was not. This guide lays out a seven stage process (capture, enrich, dedupe, evaluate, decide, implement, measure), the roles that run it, and the decision trail that keeps it honest.

It is written for the people who own the outcome: innovation leads, heads of operations, HR teams running suggestion programs, and executives who want employee ideas to feed real priorities. The stages apply whether ideas arrive through focused challenges or an always-open channel, and whether you run the process in idea management software or in a shared spreadsheet.

Why a defined process matters

Most companies do not lack ideas. They lack a reliable way to move ideas forward and a consistent way to say no. When the path is unclear, three things happen. Reviewers spend their time on whatever arrived most recently rather than whatever matters most. Authors never hear back and stop submitting. And leadership sees a list of suggestions with no way to connect them to the goals it has set.

A defined process fixes this by making each step explicit, giving each step an exit criterion, and naming who is responsible. It also produces something companies rarely have, a record of why each decision was made. That record is what lets you learn from past choices, defend them when asked, and revisit them when circumstances change.

The seven stages at a glance

Stage 1, capture

Capture is about lowering the effort to submit while still getting something usable. The most effective way is to ask a question. Ideas submitted in response to a focused challenge, such as "How could we reduce first-response time in support without adding staff?", are easier to compare and far more likely to be acted on than ideas sent to a general inbox. Our guide on how to run an idea challenge covers writing that question well.

At capture, require only two things: the problem and the proposed solution. Everything else can be added later. A form that demands ten fields up front will lose the frontline employee who has one sharp observation and five minutes before the next shift.

Stage 2, enrich

Raw ideas are rarely ready for review. Enrichment turns a sentence into a proposal by adding a concrete first step and a way to tell whether it worked. The success measure matters most, because it is what you will return to at the end of the process. Without it, "implemented" just means "someone did something".

Enrichment can be done by the author, by an admin who follows up with questions, or with AI drafting the missing parts for the author to check. In Brainstormer, AI expands each submission into problem, solution, first step and success measure, and the author or an admin can edit the result before it moves on. The point is not polish. It is that every idea reaches evaluation in the same shape.

Stage 3, dedupe

In any program with more than a few dozen contributors, the same idea will arrive several times in different words. Deduplication merges them into one idea, keeps the clearest wording, credits every author and combines their votes. It then groups ideas into themes so reviewers can see where attention is concentrated.

Skipping this stage is expensive. Reviewers score the same thing repeatedly, votes are split across copies, and the idea with the broadest support looks weaker than it is. Deduplication also often reveals that a theme, taken together, is a stronger proposal than any single idea in it. If you manage a large standing list, idea backlog management covers how to keep it from growing stale.

Stage 4, evaluate

Evaluation should measure ideas against what the company has decided matters, not against a reviewer's taste. Set the strategic goals and their weights before review begins, and score each idea on a small number of criteria, typically impact, cost, risk and strategic fit, with a written reason for each score and a weighted total.

Written reasons are not optional. They make scores comparable across reviewers, they let decision makers challenge a score without redoing the work, and they give authors real feedback. AI can apply the same criteria to every idea as a consistent first pass and write out its rationale, and reviewers then adjust where their judgement differs. For the mechanics of weights and scales, see the weighted scoring model.

Stage 5, decide

This is the stage most programs skip, and it is the one that determines whether people keep contributing. Every evaluated idea needs a recorded outcome. Keep the options few and clear.

• Approved, with a named owner, a first step and a target date.
• Parked, with a reason and a date or condition for revisiting.
• Rejected, with a specific reason the author can understand.

Then tell the author. A two sentence note with the decision and the reason is enough. Decisions are usually taken in a regular review meeting by the program owner or a small group, using a shortlist ranked by weighted score. A board report that summarizes the ranking, themes, stage counts and a recommendation makes that meeting shorter and gives leadership a document to keep.

Stage 6, implement

Approved ideas become work. The idea owner is responsible for the first step, which should be small enough to start within weeks: a pilot in one team, a prototype, a changed procedure in one location. Tracking implementation in the same place as the idea keeps the connection between the original proposal, the decision and the result. When implementation lives in a different system with no link back, programs lose the ability to show what employee ideas actually produced.

Stage 7, measure

Measurement closes the loop. Compare what happened with the success measure written at the enrichment stage and record a short result. Did onboarding time drop, did first-response time improve, did the pilot team adopt the new procedure? Then mark the idea implemented, or close it with what was learned. Across a quarter or a year, these results tell you which kinds of ideas and which challenges produce value, which is the basis for sensible innovation metrics.

Roles, owner, admin and member

A process needs clear responsibilities. Three roles cover most companies.

• Owner. Holds final accountability for the program, sets strategic goals and weights, and makes or signs off on decisions. Usually a senior leader or head of innovation.
• Admin. Runs the program day to day. Opens challenges, seeds starter ideas, merges duplicates, organizes themes, coordinates reviews, moves ideas between stages and sends feedback to authors.
• Member. Any employee who submits ideas, improves them and votes on colleagues' proposals.

Keeping these roles distinct stops the common failure where everyone can comment but nobody can decide. Brainstormer uses the same three roles, with admins running the pipeline and members submitting and voting.

The decision trail

Every change of stage should leave a record: who moved the idea, when, from which stage to which, and why. Over time that trail answers questions that otherwise cause friction. Why was this idea rejected last year? Who approved this pilot? How long do ideas typically wait in review? It also protects the program, because decisions with visible reasons are much harder to dismiss as favoritism.

A good trail is automatic, not a separate log someone has to remember to fill in. Brainstormer records every stage change with the person, time and reason, and the audit trail can be exported as CSV for review or archiving.

Common pitfalls

• No exit criteria. Ideas sit in "under review" for months because nobody defined what finishes the stage.
• Scoring without goals. Criteria that are not tied to strategy reward enthusiasm rather than value.
• Decisions without reasons. Authors cannot learn, and the program cannot defend its choices.
• Approval without ownership. An approved idea with no owner and no first step will not happen.
• Stopping at implementation. Without measurement, you cannot show what the program achieved.

A process like this does not need to be heavy. Seven stages, three roles, clear exit criteria and a recorded reason for every decision are enough to turn scattered suggestions into a pipeline leadership can trust. If you would rather run it in one workspace than across forms, spreadsheets and email, Brainstormer's idea management software covers each stage, from capture to the board report.