5 steps to conversion optimisation | Weboptim

Conversion means different things for every website: some want more visitors, some want more newsletter subscriptions and some want more sales. The goal of conversion optimisation is to ensure that your website achieves as many conversions as possible. Often with just a few small changes: for example, rearranging the landing page, providing more contact options, using simpler forms; you can achieve spectacular results.

It seems simple, but it always turns out not to be. There are several factors that can prevent visitors from becoming customers.

Some common problems:

lack of trust
poor communication at the point of sale
(missing customer reviews)

How do we know what the problem is?

Test the website and monitor changes in conversions.

Step I - data collection

3 main areas: business - website - customers

1. The business

a. Why does the business exist?

This is one of the central questions, because if we can answer this question, we can say what is it that makes us different from other companies. People vote for a company by what they buy, by the message it sends.

b. What are the objectives of the company?

Data collection and subsequent testing should be structured around our goals. There are several types of objectives: some are obvious (macro-conversion) but there are also some that are just small steps towards a bigger goal (micro-conversion). For example, if our goal is to sell a product, the macro-conversion is the purchase itself, while the micro-conversion may be, among other things, a subscription to a newsletter. After all, if someone signs up, there is a greater chance that he or she will become a customer in the near future.

c. What is special about the company?

What differentiates our business from other companies offering similar products/services? For example, free delivery, a multi-year warranty, immediate service, etc... It is a special extra if we can offer something that our competitors cannot.

2. The website

a. What does the sales process look like?

To identify problem areas, the sales process needs to be mapped out step by step. Identify the points where users get stuck most often (this could be caused by a technical problem), where most exits occur.

Are there steps in the conversion channel that don't perform well on a mobile device or desktop computer? Do visitors start the process from the first step, or do they only enter somewhere in the middle?

b. What is the current turnover?

By reviewing a few key points, you can get a relatively accurate picture of the website:

What is the average traffic on your website?
What are the demographic characteristics of the visitors (age, gender)?
What technology do they typically use (mobile, tablet, laptop)?

A Google Analytics to collect this data:

Once you have the data, you need to look for the pattern that converting users (customers) follow on the website and when they go to the website. Where they are coming from is sometimes difficult to determine, but if you do a detailed analysis, you can find out which routes are the most common.

3. The buyers

One of the most popular ways to collect information about customers is to use a Google custom survey making. We can choose the target audience and the question we want to answer. The topic is then displayed on various websites, embedded. After a few hours, you can see the results of the survey.

Step II - hypothesising

We have 3 main questions to answer:

1. What will we test?

The first step is to look for common buying trends based on the information gathered. For example, if you have found that visitors like the free delivery option, then making this feature more prominent on your product pages can increase the number of customers. Once we have a hypothesis, it is important to know what counts as a success and how to decide whether a particular test result is positive or not.

2. Who will we test?

Websites are visited by different types of people, not only in terms of demographics, but also in terms of the buying process. To achieve reliable results, new and returning visitors need to be tested separately.

The importance of segmenting users: whether by browser, location or other criteria. If, for example, the page you want to test does not load properly in one type of browser, it is worth filtering out visitors from that browser to get reliable results.

3. Where will we test?

Here you have to choose which parts of the website you want to test: you can choose just one sub-page or several pages of the same type (but in the latter case you have to take into account that the customer cycles may be different).

Step III - wireframe design test

A wireframe is the screen plan of the website that the user sees when they arrive at the website. However, the focus here is not on colours or font, but on functionality and usability.
During conversion optimisation, tests may show that you can get customers with minor or major changes. In such cases, it may be worth taking this step to see if this is the case.

Some questions to be answered here:

Do the changes directly test my hypothesis? (Then it's good if they do. )
Are they having an impact on the brand? (An important question for large websites.)
Are they technically feasible?(If we don't have enough development time to make the changes, it may limit testing.)

Step IV - implementing the design

If the previous step shows that the new design is more convincing to visitors, we can test it in real life. The more complex the design, the more time it takes to implement. Before going live, it is important to check that the site works in all browser types.

You can specify what percentage of traffic should go to the page you want to test. Sending only a small proportion of the high number of visitors to a given page can somewhat offset the risk of a lower conversion rate. For example, if only 15% of visitors are directed to the page, the remaining 85% of visitors will normally provide the daily visitor/conversion number.

Step V - establishing the validity of the hypothesis

1. Have we reached the statistically relevant data?

Statistical significance is reached if we have enough conversions and the conversion rate is high enough. On average, 200-1000 conversions are needed. However, it is difficult to predict how long it will take to achieve this.

2. Was the hypothesis correct?

If the test is successful and the conversion rate has increased, what is the next step? In the short term, we can try sending 100% of visitors to the site, but we need to monitor the numbers on an ongoing basis.

3. If the hypothesis was not correct..

In this case, you should not give up either. There are lessons to be learned from failed tests, and lessons learned can help future testing. When testing again, you should always look back at the data collected during previous tests, even if the test was unsuccessful.

You can see that conversion optimization is a constant process, as there will always be points that can be improved. You have to look for opportunities!

Source: moz.com