An Airtable employee suggestion workflow can connect an improvement idea to a review decision, a small experiment and a recorded outcome. Use miniExtensions to collect practical suggestions and let contributors follow their own submissions. Keep reviewers responsible for deciding what to try, who owns it and whether the result justifies a wider change.
This design is for everyday operational improvements, not confidential complaints, whistleblowing, personnel allegations or emergency reporting. Tell employees which established channel handles those matters. Do not describe this identified submission workflow as anonymous or promise confidentiality that your actual implementation has not established.
Start with a problem someone can explain
At fictional Cedar Studio, Alex notices that staff repeatedly interrupt each other to locate spare display adapters. Alex suggests a labeled checkout shelf with a simple return checklist. The operations team wants to understand the recurring problem before deciding whether a shelf, a purchasing change or a different process would help.
A useful suggestion describes what happens now, who encounters it and one proposed improvement. Avoid asking the submitter to produce a polished business case. A short example such as “We searched three cupboards before Tuesday’s client call” is more useful than an unsupported claim that the idea will save hundreds of hours.
Separate ideas, reviews and experiments
Use these proposed tables in Airtable:
- Contributors: Name as primary single line text, Email as email and Team as single line text.
- Ideas: Idea title as primary single line text, Contributor as a linked record to Contributors, Current problem as long text, Suggested change as long text, Example as long text, Submitted at as created time and Contributor follow-up as long text.
- Staff fields on Ideas: Review stage as single select, Reviewer as collaborator, Public explanation as long text and Internal review notes as long text.
- Experiments: Experiment name as primary text, Source idea as linked record, Owner as collaborator, Start date and Review date as date fields, Success question as long text, Observations as long text and Outcome as single select.
Possible idea stages are Received, Clarifying, Considering, Experiment planned, Not proceeding and Reviewed. Experiment outcomes might be Continue, Revise or Stop. These are team-defined labels, not automatic miniExtensions decisions. Keep a declined idea distinct from a failed experiment: one was reviewed without a trial, while the other produced useful evidence.
Give contributors a focused submission and follow-up area
Create a Portal with Ideas as the data table and Contributor as its linked user field. Prepare the direct relationship described in the users-table guide. A team name or email copied into a text field does not replace that link.
For each contributor-facing view, apply the personal-record restriction. Configure email verification for the contributor’s login address. This example shows each person their own submissions; it is not a company-wide voting board.
Allow idea creation through the portal table’s form using the documented creation controls. Expose the problem, suggestion, example and follow-up fields. Display the staff explanation and review stage as read-only, and omit internal notes. Check every available view and expansion route before inviting employees.
Make review decisions understandable
A coordinator reviews new ideas at an agreed interval. For Alex’s shelf proposal, the reviewer asks whether adapters are missing or simply hard to locate. Alex adds that most are returned, but different teams put them in different places. That clarification changes the experiment from buying more equipment to improving the return point.
The reviewer records a plain explanation: “We will try one labeled shelf near the meeting rooms for two weeks.” If the idea is not being pursued, explain the relevant constraint without disclosing another employee’s private information. A status of Not proceeding should not be the only feedback a contributor receives.
Similar submissions can share an experiment without being silently deleted. Link the experiment to the relevant ideas and acknowledge their different examples. Do not turn submission counts into employee performance scores or treat the loudest suggestion as the best one.
Run a bounded experiment and record what happened
Assign a real owner, a review date and a practical success question before starting. For the shelf trial, the question could be whether staff can find a working adapter at the start of a meeting. Record observations using the team’s chosen method rather than inventing a numerical benefit in advance.
At review, the owner notes that adapters were easier to find but cables became tangled. The team chooses Revise and adds compartments. Preserve the first trial’s observations when starting a second round. An accepted suggestion is not automatically implemented, and a trial is not evidence that the change works everywhere.
Dates and labels do not send reminders or assign work by themselves. This baseline uses manual review and communication. Any notification or routing automation needs its own configuration, ownership and checks.
Check the complete feedback loop
Use two fictional contributors to check that their ideas remain separate, that a saved follow-up appears on the intended record, and that neither can edit staff decisions. Walk through one idea needing clarification, one declined idea and one completed experiment. Confirm that an owner can still understand the original problem after the review stage changes.
This approach fits a small, regularly reviewed improvement queue. It is less useful if nobody owns decisions or explains what happened. Its outcome is not a pile of suggestions: it is a traceable path from an observed problem to a considered decision and, where appropriate, a measured experiment.


