CRO Consultant

The methodology for CRO consultants guarantees that we will be able to approach things in an optimal way. If it doesn’t work, it will teach us to pay attention to it. We must assume that we are going to make a mistake. The good thing is to realize this and to propose a new way to find the correct solution.

1. Analysis

We try to make this phase as objective as possible. As a CRO consultant we need to find things that are “improvable” (trying to figure out why).

In this process, we can find enemies within the company, who go against us by being responsible for everything “improvable”. To avoid that, we focus on relying on data and tools that support each and every thing we try to improve or bring to light. We can work in three areas:

Business Analysis

Convene a workshop to understand each other first (understanding the company’s business model).

The more information about the company, the more things we can detect on how to make money and before that we will be able to detect the specific things that will generate a greater impact on results. We need to know how everything works, even, most of the time better than the people working in the company itself.

We must carry out an analysis of the client’s business to understand their digital model -> What objectives they want to achieve.

Data analysis (quantitative) What happens?

We must have a digital analysis tool correctly configured, since it will be our starting point to obtain information.

Source: cardinalpath.com

From here we will begin to understand what happens on the website, if the investment in traffic they make is profitable. At the same time, we can make segmented conversion funnels where we can detect leakage points.

What does the user do on my website?

Behavioral analysis (qualitative) Why does it happen?

After data analysis…

Tools: hotjar, crazyegg…

Heat maps (where users move, can give us information of areas of greater or lesser interest)

Click maps (where users click)

There are many techniques to improve in CRO consulting.

A complete analysis requires a proper understanding of the business, available quantitative data and the behavior of customers or potential customers.

Everything we get out of this phase must be 100% objective.

2. Hypothesis

This is the first subjective phase. Some things that can be improved have already been detected, and someone must fix them.

Any proposal is welcome. We will then validate whether it works or not, but we should not rule out anything at the outset.

Open mind.

Roadmap of experiments

CRO consultants must list everything that they can think of, indicating:

• Language
• Device it applies to
• Traffic segment
• URL or type of page
• Hypothesis to be validated
• Name of the solution
• Explanation of the solution
• Estimated impact, you have to know the business well
• Estimated effort, it is very important to value it
• Priority, based on impact and effort. Here, getting it right in the first tests is complex, but it is the best way to gain our client’s trust.

From the start in a site, we can get relatively fast 80-90 points to test.

Now let’s see how we establish those priorities in the roadmap.

CRO priority matrix (impact VS effort)

The priority matrix is a tool that allows the selection of options based on the weighting and application of criteria, in our case:

• Impact
• Effort

From this, we extract 4 zones:

• Quick wins: tasks that require little effort and that bring us quite close to achieving the goal
• Majors projects: tasks that require effort, but we know that they work and we should always keep them in mind.
• Fills ins: few things that can be done quickly, that if there are many can add up to impact, and if not, I can discard the ones I consider.
• Hard slogs: with things of high effort and low impact. We will see if we do them or not.

We always start with the Quick wins to gain, in addition, the confidence of the client. At the same time, we have to start up Majors projects. As soon as we start some Majors projects, we must also parallelize some Fills ins, to get some quick things out.

How to measure

Of each solution proposed, we must assess whether it is better or worse than what we have proposed.

Many tests can give neutral or negative results. If most of them are like that, we must know how to identify why (through KPIs) and turn the tests around to propose them again once modified taking into account all that data.

3. Test

There are many projects that recognize these two facts:

1. How many users enter
2. How many users convert

What happens in the middle of this very long process? They don’t know.

This is where we can start testing. Experiment. If at this point, we skip all the previous steps… we will agree that the probability of failure is exponential.

Types of tests

• A/B test: these are tests in which we change only one thing

• Multivariate tests: these are tests in which we change 2 concepts, we must see which combination would work best for us. We must have very clear which variables to test. The more information, the longer the test.

• Split test or redirection test: it is a type of A/B test where we compare several versions of an element. They are very useful to test, for example, two very different landing pages, even with completely different designs.

• Multi page test: test where different pages or conversion funnels change.

CRO tools

These are the tools within our specialty of CRO consultant that we use the most.

4. Results

Let’s see an example of A/B testing, where we can see how much traffic the experiment has got, how many people it has converted, and what the conversion rate is.

The conversion is always relative data, it is not of the complete website, but simply of the area that we are analyzing.

Improving the conversion is always done little by little, it is not usually a matter of very large percentages, but it is noticeable in the accumulated positive tests.

Reliability of a test: depending on the traffic you have had, the time of the experiment and how constant it is. There are experiments that never become reliable, we must decide when to stop them (neutral test). Some payment tools give us a degree of reliability, although it is always better to analyze the trend and decide for ourselves.

Graph: evolution of the original and new conversion %. As we see, the difference remains, so it is clear that it is reliable, since it remains without an overlap from the first moment.

The overlap is not always be due to the fact that the change does not work, but to external factors such as the price of our competition at that particular time. We must know how long to wait and give the test a chance before stopping it hastily.

Types of results

• Negatives: we should always try to get them right, analyze and rethink them to see where we might have gone wrong. Let’s remember the KPIs that we have commented previously.
• Neutral: no conclusive data.
• Positive: we will try to present the data and get started on the web as soon as possible.

We must implement a clear methodology for analyzing results.