Developing a conversion rate optimization strategy
Use this robust program to develop a conversion rate optimization strategy that will help you achieve sustainable growth.
Summarize this articleHere’s what you need to know:
- Conversion rate optimization (CRO) is a systematic process of increasing the percentage of website visitors who take a desired action, such as making a purchase or signing up for a newsletter.
- CRO can be used to improve any website, regardless of its size or industry.
- There are a number of benefits to using CRO, including increased revenue, improved customer satisfaction, and reduced marketing costs.
- Implementing a successful CRO strategy requires a four-step process: planning, experimentation, analysis, and optimization.
- Data is essential for success in CRO. By tracking and analyzing data, you can learn what is working and what is not, and make adjustments to your strategy accordingly.
- CRO should be a team effort. Involving all relevant stakeholders, such as marketing, sales, and product development, is essential for success.
Optimizing your website to ensure high conversion rates is one of the most important aspects of any marketing strategy. The thing is, it is also one of the toughest areas to deal with. Here’s a simple, yet effective, 4-step model for building a successful conversion rate optimization strategy.
While there’s a lot you can achieve with conversion rate optimization (CRO), companies that adopt website personalization and experimentation as a team effort typically achieve meaningful growth and better results. To launch tailored experiences at scale, you want to build a comprehensive CRO strategy.
A good CRO strategy has multiple contributors, with different functions from Marketing, Product, UX, and Development teams. All of them can contribute their skills and knowledge to the success of the program.
As you work within a team, you must build a foundational set of practices:
- Analyze data: Investigate your data, research, and explore your analytics and learn about your traffic and its performance, which will help you find opportunities.
- Brainstorm ideas: Now that you know what you want to improve, think about different ways to shift the user behavior. You can find more than a hundred personalization examples to kickstart the process.
- Optimize, prioritize, and re-evaluate: Transform a pool of ideas into a personalization roadmap. Adjust your experiments, and re-iterate where necessary.
Once armed with a list of prioritized ideas, the next steps come into play: creation, execution and then back to analysis. But this is an ongoing process. Make every idea a learning opportunity by thinking one step ahead. Consider what you could learn if your hypothesis was proven correct or incorrect in the case of a variation winning, losing, or lacking statistical significance. Sometimes, this analysis may lead you to new ideas.
Introducing a 4-step systematic conversion rate optimization strategy
The following chart provides a bird’s eye view of the experience optimization process and methodology. Each step in the process is represented by a unique color. The X-axis represents the elapsed time between the steps of the process; while each circle represents the estimated capacity of the specific process in terms of work involved. Of course, in reality, the amount of work may vary from one case to another. The bigger the circle is, the more time consuming the task involved is for marketers.
Step 1: Investigation
This step has a clear and straightforward objective: to gain a high trust level with your web analytics data. Generally speaking, it is a two-part process:
Data Collection – Successful companies are data-driven. While good instincts are valuable, making decisions based on data will reduce risks. You want to start your ideation process by looking at the data and finding opportunities. But first, you need to make sure that the data you’re collecting is complete and thorough. Two simple ways to start with are by defining your main segments and analyzing your conversion funnel. In this step, you should be looking for implementation errors and examining your data layer to identify missing events, untagged pages, etc.
Data Filtering – Once you have completed the initial technical step, it is time to clean up the data, making it more readable and valuable for future analysis. There are many different ways of filtering out noise from your data. You to start by consolidating canonical URLs, filtering internal IP addresses and testing environments, excluding bots and known spiders, implementing custom campaign parameters to your URLs, etc.
Step 2: Research
Now, when you trust the data and feel comfortable with its level of accuracy and organization, it is time to move forward and learn as much as possible about the business and its online activities:
Business Objectives – The main questions you should be asking are why does the website exist and what are its objectives?
Website Goals – The next part is to figure out what are the goals of the website are, which can help you achieve the desired business objectives. In other words: explore what the desired outcomes of the business objectives are. For example, the objective of a retail site is to sell products, and, as a result, grow the business and increase revenue. So the goal of the website is to increase online sales.
Website KPIs – Once goals are defined, it is time to figure out the specific metrics that are involved in the process of measuring the success or failure of the defined website goals. For example: if the website is an eCommerce brand, possible KPIs would be: (1) Reducing bounce rate. (2) Increasing average order value (AOV). (3) Reducing shopping cart abandonment rate. (4) Increasing overall conversion rates. (5) Increasing revenue.
Data Analysis and Opportunity Discovery – At this point, you are already familiar with the main business objectives, goals, and KPIs. In addition, you also trust the data and its comprehensiveness. Now, it is time to develop audience segments and look for pain points and opportunities to improve KPIs. Every business should segment its users differently. For example, for clothing retailers – segmenting users based on gender is critical. But this is not necessarily the case for a travel brand, which is why mapping different personas can be beneficial.
One way to approach this is by looking at the main categories your content is cataloged by. For example, if your business is a food delivery service – you can segment by user preference (e.g. vegans, fast-food lovers, health enthusiasts, etc.).
You should also think about your main acquisition channels: do you spend a lot of money on display ads? Do you have a lot of direct traffic? These channels behave differently with different intent levels.
Every business has a primary conversion funnel where you define what success looks like. The funnel maps the different steps a user needs to take until a conversion (i.e. your primary metric) is completed. If we take an eCommerce site as an example, the funnel steps could be:
After building this funnel into your analytics platform (e.g. Google Analytics), you can start analyzing it:
- Compare behavior across the conversion funnel for your main segments (e.g. mobile users vs. desktop users, direct traffic vs. paid traffic). Is there a segment that performs poorly in one of the steps compared to other segments?
- Have you witnessed user dropoff in one of the “steps” compared to the industry average?
- Compare it to different timeframes: do you see a decline in the conversion rate during one of the steps? This may indicate that something in your program is broken and needs fixing.
After that, think about your secondary KPIs (e.g. newsletter subscription, register to loyalty program). Define those KPIs and analyze these conversion funnels as well.
Step 3: Optimization
Now, it would be a good time to take the opportunities you have discovered and put them into practice:
Hypothesis Consolidation – Once you have analyzed the data and have identified your opportunities, it’s time to get creative. How do you plan to solve your pain points? No one knows your business more than your team. We recommend involving everyone on your team: customer support, marketing and even the development.
A good idea is based on a hypothesis. A hypothesis clearly indicates what change you are considering, what you believe the outcome will be, and why you think the change will drive this outcome.
Taking the time to write your hypotheses correctly will help structure your personalization ideas, get better business results and marketing insights, as well as avoid time-wasting tests.
If your hypotheses are oriented toward improving your metrics, your team can focus on digging into the data and designing strong tests.
Your hypothesis should include three elements: “If [The Change], then [The Expected Impact] because [The Rational].”
Here are a few examples:
Ideation: How to develop ideas for your CRO programs
Developing Optimization Plan – Based on your hypothesis, start building the optimization plan. In particular, explore what the specific changes are that you would like to implement and test by preparing the new copy, designing materials, and such.
It’s easy to get carried away with ideas, so now you probably have more than you can actually handle; more ideas than you can actually spec, your design studio can design, your dev team can develop, or you can properly analyze. How do you decide what to do next? A good personalization program includes a prioritization system. A simple way to do it is to consider effort (the amount of work needed) and impact (how much it will “move the needle”).
For every idea, estimate the amount of work needed. Think about all the people that will work on it. There’s the marketing and product teams (e.g. writing the spec), the design team (e.g. if graphic design is needed), the dev team (e.g. developing it), and the QA team (e.g. testing it). For each team, how simple or complicated it will be to bring this idea into life? Don’t be optimistic, be realistic. This is the “price” you will need to “pay”.
For example, showing a recommendation widget upon exit intent is easy to launch. There’s an out-of-the box Dynamic Yield template and trigger for that. However, changing the layout of the homepage requires a lot more effort – getting the approval from all stakeholders for this drastic change, redesigning the new layout and all homepage promotions to fit the new layout, the custom code that your dev team will need to write, and the heavy QA that will be needed.
For every idea, try to estimate the impact it will have on your business. Again, don’t be optimistic. Ask yourself:
- Will it target all users or just a small segment?
- Will it run on pages with high traffic or pages that are rarely visited?
- Will it affect an area of the page that is above the fold or below the fold?
- Will it affect the conversion funnel directly (e.g. get more people to add items to the cart) or affect a secondary metric (e.g. acquire more newsletter subscriptions)?
An impact-effort matrix
An impact-effort matrix is a decision-making tool that will help you turn your pool of personalization ideas into a prioritized list – a roadmap. By scoring the effort and impact of each idea, you can “calculate” a priority score.
After scoring your list of ideas, see where they are mapped in the Impact-Effort Matrix:
Impact-Effort Prioritization Matrix
If there’s an idea that you believe will have high impact and low effort – Do it now. On the other hand, if there’s an idea that will require a lot of resources to develop and execute, and will likely not result in significant impact – consider not de-prioritizing it, or in some cases, not doing it at all.
By scoring all of your CRO ideas independently and placing them over the above matrix, you will get an overall priority score between 1 to 5:
- High impact and low effort.
- Medium impact and low effort, high impact and medium effort.
- Low impact and low effort, medium impact and medium effort, high impact and high effort.
- Low impact and medium effort, and medium impact and high effort.
- Low impact and low effort.
Implementation – Now, put all of the ideas into action and go live on the site. The time and effort that will be required from you in this step will vary depending on the accessibility and usability of your marketing optimization software. When implementing, remember to configure different content variations, as well as targeting conditions, and, if using personalization platform, content personalization triggers as well.
Step 4: Evaluation
Once enough data is gathered, it is time to analyze results and arrive at conclusions:
Report Analysis – Explore the optimization reports, look for noticeable changes. Identify new audience segments, methodological pitfalls and find out where you succeeded. Do not be afraid of finding failed initiatives, they would only help you gain valuable insights about what is working better for your target audience.
Read further on how to effectively analyze and interpret A/B testing results and learn why some failed results can actually be winning.
Optimization Adjustments – If necessary, modify your optimization initiatives with different targeting conditions and segments. Always double-check yourself and preview the results to avoid any errors. Adjust and be patient. Depending on traffic, it may take a while before getting high-confidence, sustainable results.
Conversion rate optimization is not a linear process, but a circular one. It includes researching, A/B testing and analyzing. Once you are done with the final evaluation step, it is a good idea to go back to Step 3 and formulate new hypotheses and optimization ideas.
This post was an overview of our internal conversion rate optimization strategy. I hope that you will find the suggested plan useful and valuable to your own optimization efforts.