From internal knowledge to decisions and tests
Finance, innovation and people, with recognition throughout. Anyone can propose; value does not depend on rank, writing skills or getting more votes.
An application of your own. A complete cycle.
Innovar IN proposes its own application, adapted to each company's size and needs. A challenge, an open ideas inbox or an employee initiative is the beginning: the work ends when an improvement is put into use, measured and its contributors recognised.
Ideas are reviewed, merged without erasing their authors, developed by teams and tested in pilots with an owner, authorised resources and indicators. A better nut matters too: value does not require inventing a new product.
The proposal includes a dashboard of activity, trends and results by site or region, plus employee summaries. Authorship is retained for internal recognition; any external disclosure needs anonymisation and review. Votes do not decide feasibility or assess people.
A common core, two routes
The mailbox and the challenges are entry doors. After that, the work changes according to what we already know and what we need to learn. Not every improvement needs a startup process.
Change something close and measure it
Observable problem, solution close to current work and an operational owner. Set a baseline, review safety and procedures and propose a small reversible change. Compare before and after without attributing the whole result to the intervention.
Fictional example: reusing boxes between two departments. Before changing anything: check materials, permissions and how to measure waste.
Check before building
A service, segment or solution with uncertainty. Identify the riskiest assumption and design a permitted observation or prototype. Seek evidence of behavior and feasibility; do not demand an invented exact return or confuse enthusiasm with demand.
Fictional example: a new exchange service between sites. First test need and use with a prototype; do not put it into operation without permission.
Seven steps, without losing authorship
Understand the problem
Context, unit, objective and internal owner. Before inviting participation: protected time, shift-friendly access, response rules and agreed resources.
Capture and recognize
Open ideas and concrete challenges. Preserve the original, date and authorship. An incomplete idea is welcome; every contribution deserves an answer and thanks.
Enrich together
Ask, comment and document contributions without erasing the original. Invite the author to develop it, without imposing that task on them. Needing help is not a lack of talent.
Choose the route and review risks
Known improvement or uncertain exploration. Identify legal, people, safety, operations and resource limits. Votes show support, not viability. No employee scoring and no automatic rejection.
Prepare a test
Hypothesis, method, indicator and baseline, the result that supports or refutes, owner, deadline, risks and resources. A test proposal is not permission to run it.
Decide on evidence
Decide to move forward, repeat, park with a date or close with a reason. Record who decides, when, conditions and next step. Review conflicts and uncertainties; allow reconsideration with new evidence.
Implement, measure and transfer
Operations receives the change only after the corresponding authorization. Measure results and costs without inventing savings. The internal team must be able to process another idea and explain its decision without depending on the facilitator.
Three distinct permissions
Research
What may be observed or asked, of whom and with what data. Exploration also needs limits.
Test
Environment, owner, controls, deadline and authorized resources for a specific experiment or pilot.
Operate
Implementation decision, maintenance owner, monitoring and stop conditions.
The committee takes part by stages: finance, people, business, operations and specialists according to the risk. One person answers for each decision and date; not everyone vetoes everything. The demo grants none of these permissions.
Governance that avoids silence
Owner and next date
Received, to clarify, exploration, proposed test, results, implementation or closure: every active state needs an owner and a next step. “Later” carries a date; closing carries reasons and an answer to the contributor.
Service targets to be negotiated: immediate acknowledgment, first human reply within 5 working days and initial review within 10. They are not the deadlines of a contracted service nor automations of this demo.
A portfolio that can be managed
Without an owner, time or a testable question, it remains an idea, not an active project. Separate problem, strategy, value, feasibility, risk and effort. Do not reduce them to a score that looks objective.
A complex idea can be broken down. It is not funded indefinitely: each step has the evidence it needs and a stop criterion.
A 12-week start, not a guarantee
Prepare and listen · weeks 1-4
Sponsor, problem, unit, privacy, baseline and agreed time. One challenge and an open channel; guidance and support accessible across shifts.
Test · weeks 5-8
Authorized tests with hypothesis, owner, date and limited resources. Brief review and experts when needed. Change or stop based on evidence.
Decide · weeks 9-10
Contrast results with operations and finance; separate observed savings from estimates. Decide the next phase; do not force a launch in regulated sectors.
Transfer · weeks 11-12
An internal team carries a new review and answers unaided. Hand over the manual, portfolio, metrics and risks. Autonomy is proven, not declared when day 90 arrives.
One unit, 30-60 invitees and 2-3 active tests are initial design hypotheses to be agreed, not proven optimal limits or commitments. If capacity is lacking, reduce scope or agree on continuity. There is no pilot with a real company booked or started.
People are not the ranking
Participation and recognition
Voluntary, open to all roles and shifts. Not participating must not penalize performance or promotion. Working time and conditions are agreed; no leisure-time work is expected. A 5% figure is just a configurable example, not policy.
An answer for everyone; recognition of collaboration and learning; recognition of results when there is evidence and agreed rules. No amounts, formulas or promises of promotion.
Authorship, identity and protection
Original, versions and contributions preserved. A visible pseudonym is not anonymity. For a real service, the design and validation of restricted identity, permissions, privacy, labor and intellectual property are still missing.
The demo accounts are shared and public. They offer no real confidentiality. Do not enter real names, private documents or company secrets.
AI as a possible aid, not a judge
It can support questions, summaries and possible duplicates, always correctable and without erasing authorship. It will not decide viability or evaluate employees. Before connecting it: test fidelity and biases with permitted data, and agree on provider, retention and cost.
Today: a manual guided editor with versions and example human reviews. There is no validated operational AI, enterprise SSO, production isolation or demonstrated scale for 100,000. No demo review is legal advice or spending authorization.
Designed with each company
The demo shows a working base. A real implementation requires studying each company's needs, culture, processes and systems, and agreeing on identity, journeys, roles and integrations. Adaptation goes beyond the logo.
Documentation, configuration and transfer are planned with their IT team so the application can evolve. A program adaptable to each company's needs: scope, maintenance and autonomy conditions are agreed before building.
What you can try today
Open idea, challenge, record and manual review in the intranet; two conceptual routes from employees. Direct implementation and consulting support are modalities to be studied according to scope and capacity, not contracts in force.
Measure activity, learning and results separately. An app, a workshop or many votes do not prove impact. The goal is to learn what improves something and leave the capability to repeat it.