Do you want to optimize your landing page? You’re probably very eager to get started testing as soon as possible.

But, I will stop you there.

Soon you’ll see that testing options are limitless.

If you are overwhelmed by a long list of ideas for testing, you will waste time and end up comparing the wrong options.

Instead, you should focus on prioritization.

How? We use The ICE method.

We will now show you how to use this framework to optimize landing pages.

How can you use ICE scoring (i.e. which ones will bring the highest return to your business) to determine which tests to invest in?

Let’s go!

What is the ICE Method?

The ICE Scoring Method (a simple framework to rank items in a project’s backlog, a list of features and tests, optimizations, and other work items) is used to rank them.

ICE stands for Importance. Confidence. and Ease. These are the three criteria that determine an item’s score. The top-ranked items in the backlog are those with higher scores.

The ICE method is used because it allows us to ask the most important questions. Particularly:

• Impact How much will this increase our overall conversion goal?
• Confidence: Based on the data I have collected, will this test pass?
• Easy: How easy (and cheap) is this test?

While not perfect, prioritization frameworks such as the ICE can be used to help teams make better decisions, avoid brainstorming traps, and make incremental business improvements that are grounded in real business goals.

The Goal of ICE Scoring

Prioritization frameworks provide a consistent and systematic way to organize ideas according to business or conversion value.

Particularly, the goal is to score landing pages with ICE scores quickly and create momentum for testing, experimentation, and further development.

You’ll have a lot more data if you audit your website, landing pages, or other page elements. Many of these data will contain potential improvements and test ideas. The ICE method can help to get you through each stage of CRO.

ICE eliminates the guesswork about which tests should be run first.

With research backing your hypothesis and momentum to achieve a conversion goal, experimentation can be done faster when you have a solid hypothesis.

Here are some examples of ICE Scoring

Let’s now look at some context.

What is the ICE method in use in the wild?

• Growth Hacking –using ICE for testing (or growth hacking) was popularized by growth hackers and it is the most commonly used way that marketers use ICE. ICE allows teams to quickly identify and prioritize growth-focused ideas for testing each week.

• Features Prioritization – project managers may also use the ICE method to make sure items in a product’s backlog are worthwhile. You can push features or steps that rank lower to a later stage or eliminate them.

• Kobe Digital Landing Page Optimization – At Kobe, we use ICE Scoring to prioritize optimizations for our clients. ICE quickly identifies the most promising ideas from research and hypotheses.

How to use the ICE Method for Landing Page Optimization

Now that you have a list of test options, how do they calculate ICE scores?

You can start by giving each ICE element a score between 1-5: confidence, impact, ease, and ease.

A 1 indicates that the idea is low in impact, low confidence, or difficult.

If a test has high impact, high confidence, or high ease, assign a 5. These 3 criteria can be combined to give you a total score of 3-15 points.

After you have calculated your priority score for each test you can quickly organize your thoughts. Add these together to get the test’s ICE score

The highest-scoring tests on pages with the highest value represent the greatest potential value to your business.

What do the Impact, Confidence, and Ease ratings mean? How do you determine what score to give them each?

Three essential questions are asked before you assign Impact, Confidence, and Ease scores.

1. Impact

First, impact, or, in other words: how much “reward”, after spending time and money on a test, is it?

The question to ask: How far can this help us or move us toward our overall conversion goal?

An estimate of the impact on a large scale could be used to measure how well an experiment is moving users through the marketing funnel.

We look closer at landing page optimization. You can use landing pages, ads, and thank-you pages to help users move through each stage of the funnel.

Every element of the proposed landing page changes will include elements that drive growth and a metric that we believe will be affected by it.

A headline or form that we believe can increase conversion rates could be an example. Perhaps a new section on landing pages that increases scroll depth.

High-impact experiments could be the best way to achieve this goal.

2. Confidence

Confidence, the next ICE score, measures how supportive your hypothesis is.

The question to ask: Based on the data, am I certain that this test will pass?

A high confidence score indicates that you have lots of evidence to support your hypothesis about the test’s outcome.

Multiple data sources can be used to support a hypothesis. A higher confidence score is warranted for more complex data types and testing histories.

This is an example of a confidence meter.

You will be able to see that formal tests result in higher scores than anecdotal or “hunches”.

Although the example is for product design, you can also use this principle to create your own “confidence meters.”

Are you testing landing pages? You can go back to the research and hypothesis steps of the CRO loop.

These data can be used to assess the support for your hypothesis.

• Test results from the past and suggestions for this page
• Check out similar landing page elements and pages for results
• CRO audits
• Heuristic analysis
• Web analytics data
• Voice-of-customer and surveys
• Technical audits

here are some tips from Nick Gibson, a CRO expert:

A confidence rating of 5 doesn’t necessarily mean you have 100 percent certainty. This would be absurd. Sometimes, the results don’t support your hypothesis.

A high confidence score means that there is substantial evidence supporting the hypothesis in similar cases, and it may be compatible with the case.

What is the takeaway? Do not worry if your 5/5 confidence test fails. You still have valuable insights.

Take a deep, slow breath, and then go back to the next cycle.

3. It’s easy

Ease simply refers to the number of resources required to complete a test. The higher your ease score, the fewer resources you will need to experiment.

The question to ask: When you look at costs such as time, money, and other resources, how simple is it to implement this test?

It is important to consider the actual cost of testing. Even high-impact changes have a lower ROI due to high costs.

Because it takes into account measurable elements such as:

• Time: How long does this take to implement?
• Expense: Will this test require additional resources or budget?
• Contract Work: Does this test require a developer or an external team member to be implemented?
• Team involvement. How many members of the team will be required to participate in this experiment or manage it? (In % of the department or collective hours).
• Effort How automated this experiment is vs manual (relevant if the test requires manual analysis, or is moderated in-person)

We recommend starting as easily as possible (using budget spending or time estimates).

ICE Scoring Template (Steal It!)

You now have a better idea of how the ICE method works and how you can score.

Next, calculate a few ICE scores for yourself.

We’re sharing the same ICE Template we use for our clients to make things easier for you:

This template contains all of the sections in the previous example, as well as an easy prioritization formula that automatically calculates each test’s priority score.

Grab your ICE Sheet, and get scoring!

Next, the pros and cons of ICE. Next, we will discuss the best ICE methods for optimizing landing pages.

The ICE Framework: Pros and Cons

Let’s compare the pros and cons of using ICE scores for landing page testing.

The advantages of ICE Methodology

ICE scoring has the immediate advantage that it is quick and simple. You won’t have to waste time selecting your next experiment.

Many ideas for testing will surface during the research and hypothesis phases of the CRO loop.

It is easy to get bogged down in your ideas, which can often lead to a loss of momentum.

Instead of confidently moving through the testing and experimentation phases,

However, this doesn’t mean that the ICE Method’s simplicity can’t be applied to all project types.

Landing page optimization is an iterative process. Experiments are not always based on past data and hypotheses. Even when tests are successful, it’s still an improvement.

ICE scoring is a good choice for these landing page optimization projects because it prioritizes momentum over accuracy.

This will help you prioritize tests within the right “ballpark” of your goal while minimizing costs and time spent analyzing your backlog.

Your job is not to prioritize. It’s to analyze, test, and grow!

There are some drawbacks to using ICE Scoring

The main criticism and real concern with the ICE method are that it uses subjective scoring.

Teams could have problems with consistency if there are no quantitative guidelines about what constitutes an impact score of 1 or 5. This is an example:

• Scores may be assigned differently by two team members based on their experience or other random factors.

• One scorer can assign different scores to the test or tests that are similar over several days.

• Potential for bias or score manipulation to force a particular experiment through

In the next section, we will discuss a few methods to ensure consistent scoring.

The larger question is: How can ICE inconsistencies affect your landing page optimization?

We don’t think so, provided you apply ICE in a context other than CRO (more later).

Increasing your testing volume due to obsessive prioritization will hurt your efforts more than small variations of ICE scores.

Try to reduce wild inconsistencies as much as possible and then move on!

Five Tips to Apply the ICE Method for Landing Page Optimization

1. Concentrate on one goal

Identify the most important goal or heuristic for each page.

Also, try not to push change across too many metrics simultaneously.

Begin by identifying the primary goal of each landing page.

Once you have a clear understanding of the page’s purpose, you can easily decide which tests to focus on and how they will impact metrics such as:

• Conversion rate
• Rate of bounce
• Qualified leads
• Scroll depth

Last thought: Don’t forget to include the people that matter! Get to know your marketing agency in Miami, team, and other stakeholders who are involved in the growth project. Get clear about the metrics that are important to you and your business.   

2. Make a backlog of hypothesis-centered testing

How can you produce more successful tests? A better-quality testing backlog.

You need to begin with reliable data sources to do this.

In other words: Poor research will not yield valuable insights from your tests.

Peep Laja, CXL created the ResearchXL model. This model provides a solid foundation for data collection.

You can make really strong hypotheses once you have mastered the art of conducting research.

Nick Gibson, CRO Director, says this about the power and importance of data and hypotheses.

“It is important to remember that the [CRO] process does not end after one round of testing. However, it should provide valuable insights for as many rounds as possible.

Each test should be designed to either support or oppose a hypothesis. The test results can be used to guide your next optimization round.

Nick shared this real-life example as well to illustrate his point.

The Power of Hypothesis-Centered Test

Imagine that you are optimizing a landing page for an offer to an older audience.

According to user research, this audience responds well if the message is clear and concise. The current landing page, however, uses messaging that emphasizes emotional resonance rather than clarity.

This hypothesis is created before you start testing. We will satisfy the target audience’s need to be clear, increasing their trust and increasing their confidence in the offer.

Two outcomes can be expected, regardless of test results:

1. Emotional resonance is more positive for the audience
2. Clarity is more attractive to the audience.

What happens after the results are in?

Record the impact of the test. A 15-20% lift/drop is better than a 3-5% increase/drop.

This is because you’ll notice that the greater the impact, your confidence in using the supporting data for scoring future tests will increase.

Importantly, keep a record of the results.

If you know that your audience responds positively to clear messaging How can you use this data to optimize other elements on the landing page?

The hypothesis can be used to test a single element of a landing page.

3. Document ICE Scores and Test results

It is easy to lose track of team members or get disorganized.

A shared testing sheet will streamline testing and assist you in future CRO research.

Include the following information in your ICE sheet:

1. Backlog for ICE testing
2. A test report is a record of every test you have run and the results.

As we have mentioned, place the most important pages at the top ICE sheet. Keep track of metrics such as page traffic, test descriptions, and I.C.E. Scores in columns before the final “priority score”

Pro Tip: Limit yourself to 2-5 test ideas per webpage to start.

Next, create a test report for every test you conduct. Here is a template that you can use to outline your test plan:

Page: [PAGE NAME]

A page title or page name is hyperlinked directly to the actual page.

Date: [DATE RANGE]

The range that includes start and end dates for your test

TEST NAME

This is the second test of the ICE Sheet, “Removed redirects 1:1.” 

GOAL TYPE

Focus metric that is affected by the specific test/hypothesis. These include conversion rate and bounce rate, scroll depth, and qualified leads.

HYPOTHESIS:

A paragraph describing your prediction, some context about the element being tested, the change made to the element, and primary evidence supporting the hypothesis.

[Champion vs. Challenger Section]

Include the following information for both the champion (usually the variant) as well as the element title: # of visitors during the testing period; metric volume (e.g. # of conversions) and rate (e.g. conversion rate, bounce rate).

[Results Section]

This section includes the lift/drop (in %) of the goal metric, the test’s confidence, and whether the hypothesis is supported (Did the challenger win?).

TAKEAWAY:

A paragraph that summarizes your findings and offers suggestions for further testing

4. Do not obsess over ICE Score Consistency

We understand that inconsistencies in ICE scoring may be concerning to you.

Be aware of these concerns before you allow them to slow down the prioritization process.

Your CRO process won’t be affected by subjectivity in your ICE scores.

Why? Because the CRO loop growth process is flexible enough to allow for imperfect scoring.

Consider…

You may not have enough data to back up your hypotheses if you are completely new to testing. You might be able to assign confidence scores or impact based on your intuition.

However, if you keep your eyes on the next cycle, your team’s testing reliability will naturally improve.

Each test will provide you with actual data and evidence that support your future research and hypotheses.

Comparison of pre-and post-test data will help you understand why some tests fail, and how to improve your confidence.

You can also recall past testing results and automatically calculate potential costs before you assign ease scores.

This means that the CRO process is naturally iterative. You can quickly move past any early scoring mistakes if you are committed to documenting your tests.

Tip: Does subjectivity still bother you? Scoring standards can reduce inconsistencies, and help scale ICE scoring for a larger group. Here are some recommendations:

• A final ICE score review should be assigned to an experienced CRO

• Add up the scores of all team members

• To reduce test lists down to 2-5 ideas, align on a single goal metric for each landing page

• Make sure to update your documentation process so that all test results (and insights!) are easily accessible to all members of the team.

Maximize your scoring consistency wherever you can and then move on!

5. When to delegate

The ICE Method is simple enough to use. It is also easy to set up and run tests.

Remember that prioritization and tests are only two components of the CRO picture.

You’ll also need to know:

• The right tests should be chosen based on sound research and hypotheses

• The right conclusions to draw from the testing data

• Based on the results, make recommendations for future tests

Reach out to agencies or contractors who are experts in the field to assist with the experimentation process.

Take a look at Ice Scoring: Should I Use the ICE Score Method?

Do you think ICE scoring should be used to prioritize landing page testing?

With a few caveats, we’ll give you an enthusiastic “yes !’….”

ICE scoring is not the most rigorous method of prioritizing. It “gets the job done” and efficiently takes you from ideation into actual testing.

You can also settle your concerns about the subjectivity of ICE methods by using techniques such as aggregate scoring to improve scoring consistency across a large group.

Without sound hypotheses and a thorough reporting process, the ICE score can be misused in an unoptimized manner.

You should leave testing to an experienced member of your team or your favorite CRO agency.