Every successful food product on the shelf in front of you went through a process.

Not a straight line. A funnel.

It started as one idea among many, and most of those ideas never made it to the next stage. That is not a flaw in the process. That is the process working correctly.

Here are the six stages that funnel actually runs through.

Where this framework comes from

This is not a new idea.

The consulting firm Booz, Allen and Hamilton documented a seven stage new product process back in 1982. It became the foundation most modern frameworks still build on.

Robert Cooper's Stage-Gate model, developed in the 1990s, refined that thinking into a structure separated by clear go or kill decision points, called gates. By the year 2000, roughly three out of four North American product developers had adopted some version of it.

The version below adapts that thinking specifically for food products, where formulation, sensory testing, and scale-up all carry their own distinct challenges.

The six stages

Stage 1: Idea generation

Every product funnel starts wide. Food product development programs typically start with somewhere around 12 to 15 raw ideas.

Most of those ideas will not survive. That is expected, not a sign anything went wrong.

Stage 2: Concept screening and market research

This is where ideas get tested against reality. Is there an actual unmet need? Does the category already have five brands making the same claim? Is this genuinely different, or does it just feel different to the person who thought of it?

This stage is where the gap between the top 20 food companies and everyone else usually opens up. The companies that do this work rigorously fail far less often later on.

Stage 3: Formulation

Around 3 to 5 ideas typically make it to formulation.

This is where the recipe gets built and stabilized so it performs the same way every single time it is made, not just the one time it happened to taste perfect.

Stage 4: Sensory and consumer testing

This is one of the most commonly skipped stages, and one of the most costly to skip.

Blind panel testing across multiple formula versions matters here, because internal teams lose objectivity fast. Testing against competitor products, not just against your own previous version, is what actually tells you whether the product can compete on a real shelf.

Stage 5: Commercialization

Typically 2 to 3 products reach this stage, where the recipe that worked in a test kitchen has to prove it can survive a real production run at real scale.

Shelf life testing usually happens at the end of this stage, because you need a product that is actually stable at scale before you can honestly document how long it lasts.

Stage 6: Launch

Out of the original 12 to 15 ideas, usually just 1 or 2 products actually reach this stage.

A serious process includes a post-launch review too. What did the market actually tell you, and what does that mean for the next idea already sitting in the funnel behind this one?

Why the narrowing is the point

It can feel discouraging to watch a list of 15 promising ideas shrink down to one or two.

It should not.

A funnel that never narrows is not a rigorous process. It is a company launching everything and hoping something sticks, which is a far more expensive way to find out what does not work.

Each stage exists to kill weak ideas cheaply, before they become expensive ones. That is what separates a company with a 24 percent failure rate from a company with an 88 percent failure rate. Not better ideas. A better filter.

The bottom line

You do not need a research and development department to run a real process.

You need a willingness to let weak ideas fail early, at stage two or three, instead of discovering they were weak after you have already paid for a production run.

Greg Aronoff is the Communications Manager for Oregon State University's Professional and Continuing Education program.