August 13, 2018

What is conversion rate optimisation and how do you do it?

August 13, 2018

What is conversion rate optimisation and how do you do it?

What is conversion rate optimisation and how do you do it?

For those unfamiliar with the concept of conversion rate optimisation (CRO), in its simplest form, it is any effort to increase the percentage of customers who complete a specific action on a business' website or application.The most basic example of CRO is making changes to a website in order to get more visitors to buy a product. On-page changes to your site are not the only elements that contribute to conversion rate optimisation - it is about making every digital touch point your customers have as easy and enjoyable as possible.

The benefits of conversion rate optimisation

There are huge benefits to conversion rate optimisation that goes beyond just getting more people to complete a specific action on your website. These include:

  1. Increase profits (not just revenue) - as conversion rate optimisation is strictly focussed on maximising the performance of your current website or app, there are no additional fixed costs to your business or increase in marketing budget required. This means your profit increases disproportionately
  2. Maximise marketing spend - if your on-page conversions aren't working as they should be that can reflect poorly in your marketing performance. CRO makes the most of your marketing budget without having to increase the amount spent.
  3. Understand customers better - a huge part of CRO is gathering data on customer behaviours and testing. As a result, you will develop a clearer picture of who your customers really are and what they are trying to achieve.
  4. Improve brand reputation - conversion rate optimisation means that customers have a more enjoyable and more seamless experience with your brand. This translates into how you are perceived more broadly with positive customer reviews, testimonials and Net Promoter Scores (NPS)
  5. Decisions are based on data - decisions are made scientifically based on testing and data, so there is a rock-solid rationale behind every change that is made, and every decision is completely objective.
  6. Minimise development time and cost - CRO not only narrows down the changes that should be made in order to positively impact the conversion rate, but it also weeds out the negative. This eliminates unnecessary development time and prioritises work that will have the biggest impact.
  7. Make changes fast - By making specific, incremental changes to your website, changes can be implemented quickly and easily without a drawn-out development cycle
  8. Improved search performance - By optimising your website you will be rewarded by Google with improved search results. Improved content, functionality and site performance all contribute to your search rankings which will, in turn, drive more traffic to your highly converting site.

However, just making changes to your website or application and hoping for an improvement is a hazardous adventure at the best of times. If you are lucky the changes cause an improvement, but what if the issue you are seeking to address stays the same or even gets worse? How can we pinpoint that the cause of the change in sales can be attributed to website or application change and not a marketing or seasonal change?With that in mind, a strategic, process-based approach to conversion rate optimisation is critical to ensure that the improvements we make are validated effectively.Before we get to the step-by-step process of conversion rate optimisation there are two ideologies that underpin CRO, and should also be foundational to your business more broadly - customer centricity and learning.

Customer centricity

Customer centricity refers to the idea that all decisions that a business or brand makes are based on what's in the best interest of their customers, and to enhance the customer experience. This may seem like an obvious concept, but the reality is that outside of tech companies and startups this is rarely done well.The impact that customer experience has on a business is profound both when done well and when done poorly. Shaping your brand to be customer-centric is critical in order to remain competitive and profitable. You can read more about customer centricity and the impact of the customer experience here where I dive deeper into the subject.


Saying that learning is important is about as obvious a statement as the sky is blue, but when I speak of learning more broadly in the context of conversion rate optimisation I’m talking about a culture of learning.That is, to see every activity as an opportunity to gather more information and improve on it the next time around.It is imperative to develop a culture of learning and improving your business in order to continue to evolve and grow. This means that mistakes aren't seen as failures, rather as opportunities to identify what went wrong and how a process can be improved going forward.This is the foundation of conversion rate optimisation. Asking questions, testing and learning what works best for the customers in order for them to achieve their goals. Not every test or hypothesis will ‘win’ in terms of increasing the conversion rate but it is critical that we are able to learn from each test no matter the outcome and adjust accordingly.

Developing a conversion rate optimisation strategy

There is no hard and fast rule on how businesses should approach conversion rate optimisation. Some will identify key tasks or goals they want the customer to achieve and focus on that in a cyclical process. Others will implement the top level concepts and run tests as the ideas are generated, not dependant on the objective they are trying to achieve.Either way, there are five key elements that must be included in any CRO strategy:

Clearly defining your objectives

As with any digital activity, it is critical to clearly define what you are trying to achieve by starting conversion rate optimisation efforts. Having the goals and specific metrics mapped out from the get-go allows you to anchor all tests and activities back to the objective and maintain a laser-sharp focus.When talking about defining objectives there are two different types that we look at in the context of conversion rate optimisation - macro and micro. Your macro objectives are the big picture KPI’s like:

  • Increase revenue
  • Increase profit
  • Grow lifetime customer value
  • Lower churn rate
  • Improve Net Promoter Score (NPS)

These are still the main KPI’s that we aim to improve and are the defining metrics that measure the success of the business. The micro-objectives, on the other hand, are metrics that in and of themselves may have little or no value, however, they may contribute to the greater macro objectives. Examples of micro-objectives are:

  • Increase click-through rate on a button
  • Increase number of site searches
  • Decrease bounce rate
  • Increase add to carts
  • Increase email open rate

It is important to note that for conversion rate optimisation it is necessary to take into account both the micro and macro objectives at all times.Any ideas and hypotheses to be tested must always be tethered back to a macro goal. An increase in a button click-through rate on its own may, or may not, be beneficial to the overall website revenue.Most website tests should have an initial hypothesis of an increase or decrease in the micro objective (ie button clicks) leading to an increase or decrease in macro objective (ie sales). In some instances, the tested variation might ‘win’ in that there are more button clicks, however, that has actually caused a decrease in revenue. By ensuring both the micro and macro objectives are clearly identified (and tracked) from the outset means that this becomes a validated learning.This is a great example of an opportunity to learn from what on the surface may seem like a failure.

Idea mining

While there are often individuals within a business or agency whose job it is to be responsible for analysis and optimisation, it is important to be open-minded with regards to where ideas for improvement can come from.There are a variety of different data sources that can be utilised to generate ideas on how to improve the conversion rate of a particular task:

  • Web analytics (ie Google Analytics, Adobe Analytics)
  • Heatmaps and customer recordings (HotJar, CrazyEgg)
  • Customer surveys

Additionally, it is important to personally experience the process or task that you are trying to optimise, from the perspective of a customer, in order to understand what the journey entails and identify any pain points that a customer may have. This doesn't necessarily require a full UX audit, but that would certainly be beneficial.In addition, it is important to open up the idea generation to your team more broadly. Empowering other team members to generate ideas on how to improve the customer experience will not only create an alternative source of ideas and potential tests but has the added benefit of staff feeling a sense of ownership of the optimisation process. This may be as simple as having a quick workshop to brainstorm ideas and concepts, or having an in-depth systematic approach to gathering ideas for defined journeys from across the organisation.Not all ideas are made equal though and it is important to go through a validation process. This includes:Ensure the idea aligns with a macro goalMake sure that the idea is clearly aligned with one of the business' macro goals.Are we able to effectively measure it?What is the macro objective and is it measurable? Is extra tracking required?What is the specific hypothesis?Ensure you are able to effectively articulate the hypothesis of the idea.How long would the test run for?Depending on your site traffic and what you are testing some tests will need to run for so long there is virtually no point. The VWO Test Duration Calculator is a great tool to use to help determine how long a test might run, and if it's viable.How difficult is it to implement?Possibly the most important element of the validation is determining the potential impact vs the effort involved. In order to make sure we are prioritising the ideas and tests that will have the most impact on our macro objectives, we need to go through the process of estimating the impact vs effort.For most ideas and tests this is a relatively simple process that involves an educated guess of the impact based on existing data/feedback and a conversation with those implementing the test or development.For more complex tests involving multiple elements or significant development, it is recommended to develop a prioritisation framework. There can be a number of different inputs to take into account but the output can look like the example below. Optimizely has a terrific in-depth article about how to create a prioritisation framework.

Prioritisation matrix | | What is conversion rate optimisation and how do you do it?


Testing is the heart and soul of conversion rate optimisation. From A/B and multivariate testing on websites to email subject line testing, UX and UI testing and everything in between it is critical that our decision making is driven by data and validated by customers.At this point, we should have a list of validated ideas or hypotheses that we want to test that have clear micro objectives and align with our macro business goals. With these, we have a clear direction as to what we want to change in order to test the idea.

An example of AB testing

In order to run a valid test, we need to ensure that we have a control variable (the current website/page/app) and at least one independent variable (the changes that we are trying to test based on our hypothesis). Generally, it is best practice to avoid having too many independent variables as it will slow down how quickly you can get a result.It's important to remember that when we are testing, the idea is not to fix everything, or bundle ideas or hypotheses into one test, but rather to create individual tests for each hypothesis. By doing this not only will be able to clearly identify which changes have made a positive impact but also how much of an impact it has made.The complexity of the test generally dictates how the implementation should be approached. If it is a basic change to your website in terms of page layout, styling, content or call to actions it can generally be completed within one of the testing applications on the market like:

  • Google OptimizeGoogle recently launched Optimize as a standalone product consolidating some of its testing features. For most basic tests this is a great option as its free, integrates seamlessly with Google Analytics and is quite simple to use, however, there is a cap of five tests that can be run concurrently on any one account.
  • OptimizelyOptimizely positions itself as the ‘world leading experimentation platform’ and it has traditionally been the market leader for on-site A/B and multivariate testing. It is incredibly powerful and provides the capability to complete complex multivariate, multi-page tests. Recently though the business has moved towards an enterprise business model which has eliminated the free plan and has priced out many businesses.
  • VWOIf you require more complex tests beyond the capabilities of Google Optimize but don't have the budget for Optimizely then Visual Website Optimizer (VWO) is your best bet. VWO has much of the advanced testing capability of Optimizely including a terrific interface to easily modify web pages and clear insights and reporting dashboard. A personal favourite of mine.

If you are testing significant functional changes then developers may need to get involved in order to create and implement the test variations.Once you have implemented the changes to be tested you will need to identify the objectives that will serve as goals to measure success. As mentioned previously it is critical to measure not only the micro goals (eg button clicks) but also the macro (eg revenue) for each and every test.

Measurement and reporting

Assuming you have implemented all the correct tracking and have both micro and macro goals as a measure of success this stage of the testing process is a waiting game. Depending on the volume of traffic going through your website or application your tests could last anywhere from days to months. Using a duration calculator as previously mentioned should avoid ridiculous wait times.Once enough traffic has passed through for your test to be considered statistically significant (the testing apps like Google Optimize will let you know when this is), then you can end the test confident that the results are valid and the hypothesis you are testing is either proven or disproven.It is important to communicate the results of tests not only with the key stakeholders involved but the business more broadly. Whether the hypothesis was proven or not, sharing the results and the learnings benefits all involved and perpetuates a testing culture within the business.A good way to communicate the results is with a reporting scorecard, preferably something that contains:

  • What was the hypothesis
  • Who was the audience
  • Visualises the data showing the results, for both the micro and macro objectives
  • Articulates key learnings
  • Lists action points and potential business impact

A nice visual scorecard showing the changes made and the macro impact that it had is sure to impress management.


Once tests have been completed, the reporting communicated and the changes implemented permanently, that isn't the end of the conversion rate optimisation process.CRO is a perpetual cycle that is focused on making small, incremental improvements that, when stacked on top of each other, make a significant impact on business' macro objectives.Making iterative changes is key to the continual improvement of your conversion rate - the fact is unless you are converting every single customer who lands on your website into a sale, there is room for optimising their customer journey.Using the learnings from previous tests you can continue to generate ideas and hypotheses, test different parts of your website and slowly but surely increase your conversion rate. As you go through the testing process you are bound to uncover insights and pain points that you didn't know existed. Or alternatively, you might discover a lack of data - maybe you need a survey to ask why customers are abandoning their shopping cart.Conversion rate optimisation is far from a stand-alone piece of work, or a set-and-forget process. It is an ongoing, iterating process where we are perpetually learning and improving the customer experience which in turn achieves our business objectives.


If you have gone through this article and feel a bit overwhelmed at the process of conversion rate optimisation you need not be. What I have outlined here are the elements of a full-blown conversion rate optimisation strategy, however, the best advice when starting out in CRO is to start small.This means to start off with a single idea to test, making sure that it is aligned with your objectives and go from there. Positive results serve as great momentum to invest more time and effort into a comprehensive CRO program. That said, even simple tests take time to research, implement and report on, and often CRO work will slip down the list of priorities in a busy organisation.The benefits though are significant and cannot be overstated for any brand that has a digital presence and any sort of conversion activity online, so if you would like a chat about how we at Bang Digital can help you with conversion rate optimisation, contact us today.