What is the ICE score in Growth Hacking?


In this article we will answer the question: What is ICE score? And how it’s used in Growth Hacking to prioritize growth experiments.

Before you start execution, it’s important to prioritize. Prioritization further cuts down on cost and time for you and makes it easier and faster for you to find your unicorn growth hacks. The ICE score is useful during this process.

There are many prioritization frameworks out there but the most common one which is called the ICE framework which uses the ICE score as a “Ranking” number that helps you understand which growth hacking experiments you should run first.

What is ICE score?

what is ICE score

I is the Impact on your OMTM.

Some experiments will affect your OMTM directly and some indirectly. For example, If I’m running a campaign to have people opt-in for product demos and there’s a direct impact on the number of demos through my campaign, I will score I for this campaign at around 8 or 9; if it’s an indirect impact, I would score it around 3 or 4. 

C is the Confidence that this experiment will work.

If I’ve run a similar experiment successfully before, I would score this 9 or 10. If someone else has run it with success before, 7 or 8. And if this is the first time I’m running this experiment (unless I have a very strong gut feeling about it) I would rank it 3 or 4. 

E is the ease of implementation.

If the experiment requires a number of APIs and Scripts to be integrated or created, I would score this 3 or 4. If it doesn’t need any tech implementation to start with and can be executed by interns for instance, I would score it 8 or 9. 

Above scores are just examples to give you an idea of the range I use. You can decide your score as per your data and best judgement.

How to Calculate the ICE Score?

These 3 factors are rated individually and then used to calculate ICE as per the following formula:

What’s a good ICE score?

Remember, the ICE Score is for prioritization and usually not for elimination.

So don’t think if you’ve got a good or a bad ICE score. Instead, Focus on the Highest ICE score experiments and prioritize those first.

I would include the top 2 experiments based on ICE scores in my first execution cycle. You can decide to do 3 or 4 in parallel based on your bandwidth and resources.

The goal of doing this is to make sure you WIN faster and ELIMINATE waste. In the true spirit of a growth hacker!

UPDATE: Jan 2021

While the ICE score sounds good on paper, we do not use the ICE score anymore.

And here’s why:

In reality, the ICE score is just a fancy term growth hackers use to look all scientific and process-oriented.

It’s just a way of putting a number of things we feel will work well. It’s just a way of sounding confident even when we know it’s an experiment and it might very well (in very high probability) fail.

While not all reasons for using the ICE score and invalid.It’s also a way to show clients that we can create an impact using experimentation.

It’s also a way to showcase and development seriousness in what we do and not just come across as a team executing small tricks and hacks.It’s also a way to appear more data-oriented.

At the end of the day, it’s all marketing.

And the terms like “growth hacking” and “ICE score” are no different.

Their essential meaning is valid and useful but the math isn’t required to run experiments. No matter what your ICE score is, you need to constantly come up with experiment designs and execute them.

As to what you should execute first, think about what you come with first and the team’s bandwidth and just get started.

All the best!