From Idea to Implementation: How Companies Develop New Ideas

Companies receive ideas from many directions: customers, employees, sales teams, support cases, suppliers, competitors, research and changes in technology or regulation. The challenge is rarely a complete lack of ideas. It is deciding which observations deserve attention and moving them forward without investing too much too early.

There is no single innovation process that fits every company. Developing a new service feature is different from creating a regulated medical product or changing an internal way of working. Still, most effective processes contain the same basic logic: understand the problem, define the opportunity, explore alternatives, test assumptions and implement the strongest option.

1. Capture the observation

Ideas disappear when there is no clear place to record them. A useful submission should include more than a proposed feature. It should describe:

• who experiences the situation;
• what they are trying to achieve;
• what currently prevents or complicates progress;
• what consequence the problem creates;
• what evidence or examples are already available.

The proposed solution can be recorded too, but it should not replace the problem description. Several different ideas may address the same need.

Companies also need to make participation feel worthwhile. If employees and customers submit observations but never hear what happened to them, the flow of useful input will quickly decline.

2. Clarify the problem before evaluating the solution

The first review should ask whether the need is understood clearly enough—not whether management immediately likes the proposed answer.

Useful questions include:

• Does the problem occur repeatedly?
• Who is affected, and how strongly?
• What do people do today?
• What does the current situation cost in time, money, risk or lost opportunity?
• Is the issue aligned with the company’s customers, capabilities and strategy?
• What important information is still missing?

The Design Council’s Double Diamond describes this early work as Discover and Define: research user needs, make sense of the findings and frame the challenge before developing solutions.

Some submissions will be combined because they describe the same underlying need. Others will be redirected to an existing product, process or team. A transparent explanation is valuable even when an idea does not progress.

3. Prioritize the opportunity

Companies cannot pursue every idea at once. Prioritization should balance several dimensions rather than relying only on popularity or the seniority of the person who suggested it.

Common criteria include:

• Customer value: Would solving the problem create a meaningful improvement?
• Strategic fit: Does the opportunity support the company’s direction?
• Evidence: What shows that the need is real and relevant?
• Reach: How many customers, users or processes could benefit?
• Feasibility: Does the company have the necessary technology, expertise and partners?
• Viability: Can the solution be delivered and sustained economically?
• Risk: What legal, safety, privacy, operational or reputational constraints apply?

A score can support discussion, but it should not create false precision. The evidence behind the score matters more than the number itself.

4. Generate more than one possible solution

Once the problem is sufficiently clear, teams can explore alternatives. The original submission may contain a strong idea, but treating it as the only option too early can limit creativity.

Possible responses may include:

• changing an existing product feature;
• improving instructions or communication;
• redesigning a service process;
• integrating with another provider;
• automating a manual task;
• creating an entirely new product or service;
• deciding that no new solution is currently justified.

The aim is not to generate the largest possible number of ideas. It is to compare meaningfully different ways to address the need.

5. Turn assumptions into tests

Every new idea contains uncertainty. Customers may not care enough to change their behaviour. The technology may not perform as expected. The economics may be unsustainable. A process change may create problems elsewhere.

Instead of debating these uncertainties indefinitely, teams can convert them into explicit assumptions:

We believe that [a defined group] experiences [a specific problem] strongly enough that [a proposed improvement] will lead to [an observable outcome].

The next step is to identify the riskiest assumption and design the smallest credible test. Depending on the situation, this might be:

• customer or employee interviews;
• observation of the current process;
• a clickable prototype;
• a manual pilot;
• a limited release;
• a technical proof of concept;
• a pricing or purchasing experiment.

An experiment should have a learning goal and decision criteria before it begins. Testing is not intended to prove that the team was right. It is intended to improve the decision.

6. Build the business case progressively

At an early stage, precise revenue forecasts and implementation plans may create an illusion of certainty. The business case should develop as evidence improves.

Initially, the case may describe the need, strategic relevance and key assumptions. After research and experiments, it can include stronger information about expected value, cost, risks, required capabilities and implementation options.

Decision-makers can then compare the opportunity with other uses of the same resources. An idea may be valuable but still not be the company’s best next investment.

7. Implement, measure and learn

Approval is not the end of the process. The solution must be delivered, adopted and measured against the original need.

Before launch, define what improvement should be visible. Depending on the idea, useful measures might include time saved, fewer errors, increased use, reduced support contacts, improved conversion, lower environmental impact or better customer satisfaction.

After implementation, compare the outcome with the original assumptions. Unexpected effects and remaining problems should feed back into further development. Innovation is not a straight line; learning may return the team to an earlier stage.

What happens to ideas that are not implemented?

Rejecting an idea does not necessarily mean it was poor. The timing may be wrong, evidence may be insufficient, another solution may already be planned or the opportunity may not fit the company’s current strategy.

A healthy process records the decision and explains it. Some ideas should be revisited when technology, customer behaviour or company capabilities change. Others may help another organization even if they do not fit the company that first received them.

The quality of the input shapes the process

A company can improve its innovation process by encouraging people to submit clear problems, not only finished solutions. This widens participation: a customer or employee does not need to know how to build the answer in order to contribute a valuable observation.

Making needs visible also allows similar signals to accumulate. Companies can recognize patterns across individual submissions, while innovators and specialists can apply their expertise to challenges they might not otherwise encounter.

The journey from idea to implementation is therefore not simply a funnel that eliminates suggestions. At its best, it is a learning system: observations become shared problems, problems become testable opportunities, and evidence guides which solutions should move forward.

Further reading

• Design Council, History of the Double Diamond
• OECD and Eurostat, Oslo Manual 2018: Business innovation and knowledge flows
• Strategyzer, Testing Business Ideas