• Subscribe
    1. Discussions
    2. How do you Prioritize Features: "RICE" or "Rattouie"?

    How do you Prioritize Features: "RICE" or "Rattouie"?

    Mike Kyriacou
    5 replies
    How does your team decide what to build next? --------------------------------------------- Context: We're launching Kopernicus AI soon, which generates Strategic Market Intelligence as-a-Service. Makers have been asking us how we cut through the noise to prioritize what to build. Conventional wisdom suggests the RICE model: RICE Score = Reach X Impact X Confidence X (1/Effort). But we found some flaws: * REACH - is great - but this is an exponential quantity so anything of low priority reaching 1,000,000 users, say, will dominate. * IMPACT - is critical, but because its just one of the multipliers something of massive, critical impact to a subset of your users can easily be weighted out by a nice-to-have with global reach. * CONFIDENCE - useful but hard to cause and again something critical of low confidence gets weighted away. * EFFORT - vital, but again a small task which takes a fraction of a time unit (part of a week) creates a massive multiplier effect, which can force the team to spend their entire time on tiny trivial tasks. --------------------------------------------- Our Solution? The RI2E "Rattouie" Score: --------------------------------------------- RI2E = Reach * (Impact)^2 / (1+Effort) Here’s how it works: REACH - we grade this out of 10 where 10/10 is our entire user base or pipeline. This avoids orders of magnitude warping the score. IMPACT - the most critical quantity, and rated out of 10. We originally squared this (which is there the "2" comes from) but more recently we've been cubing it. This places absolute primacy on customer impact. If there is a customer niche that has something critical, Rattouie will correctly weight it. EFFORT - is as before scored in weeks. We add 1 to the divisor to blunt the effect of tiny fractions. This gives weight again to critical, high impact issues which otherwise take a long time. CONFIDENCE - goes away as a parameter - we just update our other assumptions every week as new information comes to hand --------------------------------------------- Could this work for you? Do you see any issues or even tweaks which might be necessary? Find out what we built using Rattouie right here: https://www.producthunt.com/products/kopernicus-ai So what method do YOU use to prioritize? How well is it working for your team? Looking forward to your thoughts and methods! Mike Kyriacou CEO & Head of Product Kopernicus AI Launch Coming Soon! https://www.producthunt.com/products/kopernicus-ai

    Replies

    Mike Kyriacou
    @thisiskp_ - would love to learn what methods you have seen work for prioritizing tasks. What have you seen work well?
    Share
    Mike Kyriacou
    @thisiskp_ thanks for weighing in! RICE was invented and promoted by Intercom, and has become something of a standard on Product Management Platforms like Product Plan & Product Board. ProdCamp and Roadmunk base their product around the RICE system and the clincher for me was I read somewhere that Tesla was using it (or at least trialing it). We used it to quite good effect in the early iterations of Kopernicus AI but we ran into those issues which maybe 30% of the time thew out our priorities. You're absolutely right about needing to weight tasks by customer input. Here's how we incorporate customer feedback: REACH (/10) - external, from customers - what % are requesting this (this can come directly from uses requesting it - like our "Market Projections" feature, or starts with a guess on how many would use/appreciate the feature, like our LLM based company search) IMPACT (/10) - external, from customers - of those this will reach, to what extent will it positively impact their insights, relevance or experience? EFFORT (/10) - estimated internally by Engineering, and we regularly revisit as we get a better sense of what is involved (supplanting the misleading "Confidence" score) Curious about what approach Paddle uses.
    Share
    KP
    Dappier
    Dappier
    @mike_kyriacou Havent tried RICE yet but I am really curious to learn more. Any of your fav resource on RICE? In the past, most of the feedback iterations in my prev startups came from just weighing based on customer requests and mapping it with the vision of what I wanted the product to be
    Share
    Mike Kyriacou
    What method do you use right now?
    Share
    2 Answers
    Mike Kyriacou
    @busmark_w_nika - looks like you've been across a huge number of products. Have you seen limitations with RICE? I'd be fascinated to learn what you've seen work.
    Share