ICE Score Framework Complete Guide (2024)

Prioritization is super important in product management. You can only spend your time once, so you need to make sure you're doing the right things.

In my 9 years of consulting at Accenture and Ernst & Young, I've seen a lot of companies struggle with this.

Often, product managers prioritize based on gut feeling or who shouts the loudest (usually the boss).

But that's not the best way to do business. You want a more objective way to set priorities.

There are several prioritization frameworks that can help you with this.

I'll walk you through the ICE scoring model, sharing my tips and tricks along the way.

Table of contents

1. What is the ICE Scoring Model?
2. Components of ICE
3. The ICE Prioritization Process
4. ICE in Action: Real-World Examples
5. ICE Prioritization with AI
6. Benefits and Drawbacks of ICE
7. ICE vs. RICE: Which is Better?
8. Do's and dont's of ICE
9. Conclusion

What is the ICE Scoring Model?

This framework is great for product managers who are just starting out with prioritization, but even the pros can benefit from it.

Sean Ellis, famous for coining the term "growth hacking", invented this prioritization model.

The ICE framework helps you make a "good enough" guess at priority. It's not perfect, but it helps you figure out which features are awesome and which ones you should probably skip.

ICE stands for three factors:

• Impact – How much does this help us reach our goals?

• Confidence – How sure are we that this will actually work?

• Ease – How hard is it to make this happen?

The basic formula for calculating the ICE score is:

ICE Score = Impact x Confidence x Ease

It's like the classic impact/effort analysis, but with confidence thrown in for good measure.

Let's break down each component and then we'll walk through the process of using ICE for prioritization.

Components of ICE

Impact

Impact measures how much a feature contributes to your goals. Before scoring impact, ensure your goals are clearly defined. A great tool for this is the Product Vision Board, which helps you articulate your vision, target group, needs, product, and business goals.

When scoring impact, use a 1-10 scale. Here's an example scale:

Examples

• Feature A: Add dark mode (Impact score: 6) - This will improve user experience for some users, but it's not transformative.

• Feature B: Implement AI-powered recommendations (Impact score: 9) - This could significantly increase user engagement and retention.

Confidence

Confidence helps distinguish between data-backed ideas and mere opinions. It's crucial to involve your team when determining confidence scores, as different perspectives can reveal potential concerns.

Here's a scale you can use for Confidence:

(Source: Itamar Gilad)

Example

• Feature E: Fix a minor UI bug (Ease score: 9) - This is a quick fix that can be done in a day.

• Feature F: Integrate a new payment gateway (Ease score: 3) - This requires significant backend work and testing.

Ease

Ease is all about how hard it is to implement something. Think of this as the traditional effort factor. How long will it take to build?

Here's an example scale for Ease:

(Source: Itamar Gilad)

Examples

• Feature E: Fix a minor UI bug (Ease score: 9) - This is a quick fix that can be done in a day.

• Feature F: Integrate a new payment gateway (Ease score: 3) - This requires significant backend work and testing.

The ICE Prioritization Process

Now that we understand the components, let's walk through the process of using ICE for prioritization.

1. List features, enablers, regulatory items

First things first, make a list of all the tasks and features you're thinking about.

Try to keep things MECE (Mutually Exclusive, Collectively Exhaustive). This means your items should not overlap (mutually exclusive) and should cover all possibilities (collectively exhaustive).

For example, "SSO login" and "Tagging" are on the same level and don't overlap. But "Change button color" and "Build a whole new app" are definitely not on the same level.

Take your time to make a good list. It's a pain when you're halfway through prioritizing and someone throws in a new idea. I usually ask customers, partners, and coworkers for their input using surveys or a feedback tool.

Let's add the list in the Score-based Prioritization module:

2. Score each component

Now, go through your list and score each item on Impact, Confidence, and Ease using the scales we discussed earlier.

In our tool, we make decisions based on how many votes a feature gets. We also look at conversion, which is likes divided by views. Even though the SSO login feature has fewer votes, it has the highest conversion rate, showing it's important.

3. Calculate scores

Now it's time to do some math! The ICE score calculation is simple:

Priority = Impact x Confidence x Ease

After entering all the data, the tool automatically calculates the ICE score. With the ICE scores calculated, we can reorder the list to view features from highest to lowest priority. This helps in identifying which features should be addressed first.

4. Review with team and stakeholders

After calculating the scores, it's important to review the results with your team and key stakeholders. This step helps catch any oversights and ensures everyone is aligned on the priorities.

5. Finalize and make roadmap

Finally, with all the data entered, we can use this input to determine the roadmap.

We review each item on the list. In ProductLift, we can change the status of the items we have reviewed. This step immediatly updates users that votes for the feature request.

As we update the statuses of the items, they are automatically added to our roadmap. This ensures that our roadmap is always up-to-date with the latest priorities and statuses.

ICE in Action: Real-World Examples

Many successful companies have used ICE to drive their product development. Here are a couple of examples:

Airbnb used the ICE product management to choose which features to build that would make people trust their platform more and book more easily. By focusing on the ideas that scored highest, they made their website easier to use and got more hosts and guests to join.

Dropbox used ICE to figure out which features would make people use their app more often and stick around longer. By building the high-scoring features first, Dropbox grew really fast and became super successful.

ICE Prioritization with AI

While the ICE framework is a powerful tool, it can be time-consuming to implement manually, especially for a large number of features. This is where AI-powered prioritization can come in handy.

The AI prioritization feature helps us decide which features to focus on in 30 seconds.

We start by setting up our product vision and adding features we want to consider. As users vote on their favorites, the AI analyzes this data and our vision to find the top 5 features.

Benefits and Drawbacks of ICE

Benefits:

1. It helps you avoid making decisions based on gut feeling or emotions.

2. It focuses your time and resources on the most impactful tasks or features.

3. It aligns your work with your business goals.

4. It improves communication and collaboration in your team by providing a standard way to evaluate ideas.

Drawbacks:

1. The factors can be subjective. Who decides what's a 1 or a 2?

2. All elements of the score are treated equally. One low factor can really drag down the score.

3. Scores can change over time. You'll need to regularly review and update your factors.

ICE vs. RICE: Which is Better?

RICE is another popular prioritization framework that stands for Reach, Impact, Confidence, and Effort. The main difference is the addition of the Reach factor, which considers how many people the feature will affect in a given time period.

So which is better? It depends on your needs:

• The ICE method is simpler and faster to use, making it great for quick decisions or when you don't have detailed user data.

• RICE is more comprehensive and can be more accurate, especially for consumer products where reach is a critical factor.

In general, if you're just starting out or need to make quick decisions, ICE is a great choice. As your product matures and you have more data about your users, you might want to consider switching to RICE prioritization for more nuanced version.

Do's and dont's of ICE

The ICE prioritisation framework can be a powerful tool for product managers, but it's important to use it properly. Here are some do's and don'ts to keep in mind when applying ICE:

Do:

• Start by clearly defining your objectives and goals.

• Involve the right stakeholders in the prioritization process.

• Use the ICE score to prioritize features or ideas, but also consider other factors such as resources, technical feasibility, and market demand.

• Revisit and update your ICE priority regularly as new information becomes available or priorities change.

• Communicate your prioritization decisions transparently and explain the rationale behind them.

Don't:

• Rely solely on the ICE methodology to make decisions. Use it as one of several inputs in your prioritization process.

• Ignore feedback from users or other stakeholders just because a feature or idea scored low on the ICE scale.

• Rush through the prioritization process without giving it enough time and attention.

• Base your ICE scores solely on your own opinions or assumptions. Use data and insights from multiple sources to inform your scores.

• Use ICE as a one-size-fits-all solution. Different projects or products may require different prioritization frameworks or approaches.

Conclusion

Prioritizing features is tough and takes time and discipline. The ICE scoring model helps you make a "good enough" estimate of priority. Now you have the tools to do this for your product.

Remember to revisit your ICE scores regularly, as your experience, goals, and confidence will change over time.

Finally, share your priority outcomes with your customers and stakeholders using your product roadmap or kanban board. This keeps everyone aligned and excited about what's coming next!