Introducing Launch AI Instant: Find Your Strongest Ideas Before You Invest in Concept Development

Product innovation often starts with dozens or even hundreds of ideas. The challenge is deciding which deserve further investment. Consumer concept testing provides valuable evidence, but testing every early-stage idea is rarely practical.

Launch AI Instant adds an evidence-based idea screening step between ideation and concept testing. It enables teams to evaluate anything from tens to hundreds of ideas based on their potential for stronger Rate of Sales (ROS) performance.

Teams can use it to:

• screen large volumes of ideas quickly
• identify and rank stronger ideas for further investment
• diagnose what may be holding weaker ideas back
• improve and rescreen weaker propositions before making decisions

Why better early-stage idea selection matters

The commercial difference between weaker and stronger products is substantial. Based on Cambri's analysis across FMCG categories in Europe and the U.S., new products in higher-performing ROS (Weighted Rate of Sales) quintiles achieve at least 2x the ROS of products in the lowest ROS quintile.

That's where Launch AI Instant comes in.

From idea to decision in seconds

The Launch AI Instant workflow is simple. Teams can (bulk-)enter 1 to 500 ideas and receive an evaluation of each idea in seconds.

1. (Bulk)-enter your ideas

Each idea is structured around four core elements: Product Type, Benefits, Reasons-to-Believe, and Product Variant. If variant information is not available, you leave it empty. Brand market share in the relevant category or sub-category is used as a proxy for brand strength.

2. Launch AI Instant evaluates each idea

Behind the scenes, Launch AI Instant draws on a model trained on standardized descriptions of launched products and their actual launch performance, measured using Weighted Rate of Sales. For every new idea, the model evaluates two main signals:

Performance patterns: It compares the idea and combination of its elements with patterns associated with stronger and weaker launch ROS performance.

Similar launches: It identifies semantically similar products in the training data and considers their actual Weighted Rate of Sales performance.

These signals allow Launch AI Instant to evaluate a new idea against patterns learned from real in-market launches - in seconds.

3. Get your results

The model combines these signals into two actionable outputs:

• Instant Score (0–100): provides a consistent basis for screening and ranking ideas
• Diagnostic heat map: shows which parts of an idea contribute more positively or negatively to its score

Unlike approaches based primarily on stated consumer purchase intent, Launch AI Instant learns from actual in-market performance.

Turn the Instant Score into a decision

The Instant Score helps teams answer two practical questions.

Should this idea continue?

A score of 20 or above indicates an idea likely to reach at least the fourth ROS quintile and can progress to further development and validation. This is a meaningful threshold: depending on the category, only 10–35% of new products reach the top four ROS quintiles.

A score below 20 indicates an idea likely to land in the bottom ROS quintile and should be reconsidered or screened out. In most categories, the majority of new products fall into the bottom ROS quintile.

Which ideas should we prioritise?

Teams can use Instant Scores to rank ideas within a pipeline according to their relative ROS potential. If a team starts with hundreds of ideas, it can evaluate the entire set in seconds, compare their relative potential, and identify which warrant further development.

Diagnose, improve, and rescreen

The Instant Score tells teams which ideas are stronger. The diagnostic heat map helps them understand what is driving the result. The heat map shows which phrases and elements contribute more positively or negatively to the score.

Teams can use these insights to rethink weaker parts of an idea, revise the proposition, and evaluate it again - in seconds.

Where Launch AI Instant fits

Launch AI Instant and consumer concept testing play different roles. Launch AI Instant is designed for early-stage screening, ranking, and diagnosis, when the pipeline is still large. Launch AI concept testing with human respondents is designed for deeper testing and refinement once teams have selected the ideas worth progressing. Used together, they create a simple innovation sequence:

Generate ideas → screen broadly with Instant → develop the strongest ideas → validate with consumer concept testing.

This allows teams to bring evidence into decisions earlier and invest deeper research where it adds the most value.

Put your next pipeline through Launch AI Instant

The easiest way to understand the value of Launch AI Instant is to apply it to a real innovation pipeline.

Start with an upcoming innovation pipeline, the output of an ideation workshop, or a portfolio of early-stage propositions. Enter the ideas, get the results in seconds, compare their scores, and use the diagnostics to improve weaker propositions.

Then take the strongest ideas forward for deeper development and consumer validation.

The objective is simple: test more possibilities before making expensive decisions about which ideas deserve further investment.

Watch a short Launch AI Instant demo video here or contact us to give a it try!