RICE stands for Reach, Impact, Confidence, and Effort, and it’s the most common prioritization framework used by product-led growth teams. RICE helps you compare the relative value of one project to another so you can run experiments in the right order.

Many growth product managers decide on RICE scores during a meeting with their engineering team, since engineering hours are typically limited. RICE scoring should be conversational, and colleagues from data science and product marketing can provide helpful input, too.

Panobi will apply a RICE score of 1-100, with 100 being the highest-scoring project, based on your choices. Click on the calculator icon to open the RICE calculator, and then select values from the dropdown lists.

Reach refers to the magnitude of the audience exposed to the project. An A/B test on a landing page or a billboard over a busy freeway could both have “massive” reach for your business.

Impact refers to the potential impact to the business if your project is successful. A project that you think could drastically change the paid conversion rate of your app should probably have either a “massive” or “high” impact rating — since adding new revenue is so important to a business.

Confidence refers to how well-vetted a specific area of work is for your team. For example, if a project in the backlog is something like “Get iOS users to connect their address book on the invite screen”, you can be reasonably confident in running that test. Even if your team hasn’t previously run an experiment using the address book, there’s a lot of prior art available online that suggests connecting an address book will increase the number of invites your users send.

Effort refers to how long it will take your team to create the project. Opportunity cost can be a real drain on growth teams, so this part of the score helps lower-effort ideas float to the top. Growth teams tend to project iteratively and focus on compounding small wins, over time. This doesn’t mean that you should never take on bigger projects. But remember to start with smaller experiments on new product areas or in new channels before committing a lot of engineering time to it.

Suggested next reads: