Website Optimisation, 10 Best Practices You Can Use

Website Optimisation, 10 Best Practices you can use

Website optimisation is an on-going process of testing. In this blog, our experts will share with you the top tips to enable you to not only effectively manage the process, but also to achieve the objectives you set up, in a systematic manner.

The diagram below maps out a simple strategy, showing what’s involved in each experiment stage.

The Steps in the Website Optimisation Process

1. Test the idea. Understand your objective, and come up with a hypothesis and idea you want to test on.

2. Test prioritisation. You have come up with a list of test ideas and you want to prioritise the more important ideas.

3. Implementation. Understand and manage the factors involved in the implementation.

4. Report and Analysis. Once you’ve run the tests, there will be reports and analyses.

5. Action. Lastly, it’s about making decisions based on the reports and taking actions.

With this as the background, we can now further explore what it takes to launch a bulletproof workflow for website optimisation and get the result.

Website optimisation is a collective effort. The first and foremost task is to get relevant teams such as Marketing, Development, Design and Product Teams involved from the start to align the objectives.

It’s important to have each stakeholder understand the experiment plan, status and prioritisation queue and report dashboard. So they know where their responsibilities lie and what results are expected.

With that in place, the 10 tips below will work magic for your website optimisation.

With this as the background, we can now further explore what it takes to launch a bulletproof workflow for website optimisation and get the result.

1. Start high

Start testing on the highest traffic pages which have the potential to bring great improvement on conversion. Calculate the required sample size to get a statistical result.

2. Don’t blind guess

Use either qualitative or quantitative data to build your testing hypothesis.

3. Set clear objectives

Make sure the objectives are measurable across the original and its variants. Also, the objectives should also be aligned with the business macro or micro goals set out from the beginning.

4. Prioritize test ideas in PIE

The PIE Framework is made up of the three criteria you should consider in prioritising which pages to test in which order: Potential, Importance, and Ease.

Potential

How much improvement can be made on this page(s)?

Prioritise your worst performers, taking into account factors such as customer and web analytics data, and expert heuristic analysis of user scenarios.

Importance

How valuable is the traffic to this page(s)?

The most important pages are those with the highest volume and the most expensive traffic. You may have identified a number of non-performing pages, but they’re not necessarily testing priorities if they don’t have significant volume of costly traffic.

Ease

How difficult is it to implement a test on this page(s)?

The final consideration is the degree of difficulty of actually running a test on this page, which includes technical implementation, and organisational or political issues.

5. Always verify

Before launch, doublecheck experiment setting, preview links and the optimise installation. After launch, check on the reporting in the next few days to see whether the data is properly tracked.

6. Experiment

Let the experiment run for a minimum of two weeks to account for cyclical variations in web traffic. Another consideration is to keep the experiment running until at least one variant has a 95 percent probability to beat the original. As we run the experiment, make sure we have specific target segments such as by geographics, user types, particular behaviours, etc.

7. Custom dimension

Set up a custom dimension to store the exp value. This allows for segmenting experiment participants for the selected date range. This works well with Google Analytics’ as the provided Experiment Name and Experiment ID dimensions are user-scoped for the duration of the experiment only.

8. Google Analytics

Segment and further analyse experiment participants behaviour on Google Analytics.

9. Free up experimentation slots

Once the experiment is no longer in use, end the experiments to free up experiment slots. As there is a limit on the number of experiments, make good use of them.

10. Work with experienced partners

If website optimisation is still quite a new territory for you and your team, work with a consultant who specialises in this kind of service. Experience is as important as, if not more so than, the technology itself.

To find out more about how the experienced team at FiveStones can assist your organisation in optimisation of your website, get in touch with the team today.

Case Study - website optimisation experiments in the Financial Services industry

Towards the end 2019 and in early 2020, we had the opportunity to work with one of our longstanding clients, who provide a range of loan and credit card services in the Financial Services industry. This opportunity included creating and testing website optimisation experiments within Google’s advanced experimentation platform, Google Optimize.

Experiments - Planning & Objectives

We worked with the client for a number of months in planning the following experiments on their website and of the recommendations we provided to the client, we started with the following experiments:

Updating the layout of the Credit Card Latest Offer Page (23 days)
- Drive conversions (new applications)
- Reduce bounce rate upon click-through

Leveraging different Product Icons (tracking destination pageviews) (48 days)
- Drive more conversions (new loan applications)

Different text for a Card Orientation Reward VS Slider (40 days)
- Drive more Credit Card product ‘Learn more’ button clicks

Implementation

These 5 experiments ran from a range of 22-48 days in which our team continually monitored progress against the objectives and our predictions throughout the implementation period inGoogle Optimize.

Results & Analysis

We had mixed results for the 5 experiments run:

1. Credit Card Latest Offer Page Layout – Variant 1 achieved a 27% increase in conversions compared to the original version meaning Variant 1 was the winner. As an OFI, we could potentially look at continuing the experiment for a longer period

2. Leveraging different Product Icons Destination Pageviews – the original product icons resulted in 6% higher conversion rate compared to Variant 1 so no change was made on-site

3. Different text for a Card Orientation Reward VS Slider – there was little evidence to suggest that the different text affected button clicks so the original was kept on-site

It is also important to consider the time frame in which an organisation runs its experiments as additional learnings to apply in the future.

Next steps - upcoming experiments and website optimisations

With a number of completed experiments and optimisations in 2019 and early 2020, we recently began looking at a number of new experiments to run through Google Optimize to continue driving a data-driven mindset and ‘test and learn’ mentality.

The next stage that is currently being tested is the addition of an “Apply now” button on the mobile navigation menu, so stay tuned to find out the results of this experiment and for more insights you can apply for your own site optimisations.

Conclusion

In conclusion, by using the steps and tips from this article, it will allow you and your organisation to optimise your website and provide the best opportunity to convert your traffic to leads as well as convert returning leads to customers.

To find out more about running experiments in Google Optimize or how the experienced team at FiveStones can assist your organisation in optimisation of your website, get in touch with the team today.