How To Increase Conversion Rates
How To Increase Conversion Rates
This is the most basic yet dreadful and complex part of the product manager job.
Let’s see how avoid the most common mistakes that many product managers make. The focus is on moving from random attempts to increase conversion rates to a methodical approach that improves both KPIs and user satisfaction.
The ‘spaghetti on the wall’ approach
The ‘spaghetti on the wall and see what sticks’ approach is a known product metaphor for making random, untargeted changes to see what might work, without a clear strategy. This method is tempting when product managers feel pressure to improve performance metrics quickly. However, it’s similar to flipping a coin, it may work and it may not work and most importantly you rarely know what part of your improvements actually worked.
Instead of addressing the root causes of issues, this approach scatters efforts across multiple aspects of the product, hoping that some change will stick and make a positive impact. Since changes are not based on a deep understanding of user behavior or data-driven insights, there is a high risk of implementing solutions that will not solve the customers’ problems that are causing the drop-offs.
This way of working will also create frustrations within your company and even with your customer base. Your company will see you consuming resources and time while the metrics will barely move, while your users will be confused by frequent and random changes. Their trust in the product will erode, as the overall user experience will degrade.
While the ‘spaghetti on the wall’ method might occasionally lead to a successful result, it is not a sustainable or efficient way to optimize conversion rates. It bypasses data analysis and user research which help you build conviction and help you achieve the improvements you are looking for in a much faster and methodical way.
The right approach to increase conversion rates
Understand your data
Start by mapping out the user journey and analyze the data to determine where users are dropping off or taking actions that do not lead to conversion.
Look for any pattern among users who fail to convert, as well as those who complete your flow. It’s also important to segment your data to understand different user behaviors based on demographics, user types, or source of traffic. This segmentation can help you find insights into how different groups interact with your product and where they may find friction points.
You should also track and analyze user behavior over time. There may be changes in user behavior due shifts in market trends, recent product changes, or bugs. For example, a sudden drop in conversion rates after an update might be due to some usability issues with a new feature.
Develop hypotheses
Based on your data analysis, formulate hypotheses about what might be causing the drop-offs. For example, if users are abandoning their shopping carts, it may be due to a complicated checkout process or unclear pricing information.
Your hypotheses should be specific and tied directly to the insights collected from the data analysis. For example, if younger users are abandoning their carts more frequently than older users, you might hypothesize that the payment options are not aligned with younger users’ preferences.
It’s important that each hypothesis is actionable and testable. This means framing your hypotheses in a way that allows you to clearly test them through changes to the product or additional user research. For example, hypothesizing that simplifying the checkout process will reduce drop-off rates gives you a clear path for testing.
Talk to your customers
After developing hypotheses based on your data insights, the next step is to collect user feedback to validate these hypotheses. You should engage with your customers to understand their experiences and get the insights that can either confirm your assumptions or point you in new directions.
There are many ways to engage with your customers.
Surveys are a common approach, because they allow you to collect quantitative data as well. However make sure your survey is designed to ask pointed, clear questions that are directly tied to your hypotheses. For example, if you hypothesize that users find the checkout process too complex, your survey should include specific questions about each step of that process.
Another way to engage with your users is to schedule more In-depth interviews as they can give you qualitative insights that surveys might miss. During interviews, you should encourage users to share their thoughts freely, which can uncover deeper reasons behind their behaviors and opinions. For example, users might tell you that they find some terms confusing or that navigation buttons are not where they expect them to be.
You can also combine these methods together. For example you could initially survey customers to get an understanding of the macro problems in your product and then use interviews to dig deeper into those problems if they happen to be too vague or need further research.
Usability testing sessions are also a good tool in a PM arsenal. During usability sessions you observe users interacting with your product in real-time to see where they find problems or have hesitation.
It’s important to structure all these interactions in a way that they do not lead the user but rather allow them to provide genuine responses. This might mean setting tasks for them to complete during usability tests or framing interview questions in an open-ended way.
Ideate solutions
Once you’ve validated your hypotheses through user feedback, the next step is to ideate solutions that address the identified problems effectively.
A good, collaborative way to come up with ideas is to organize brainstorming sessions with your team to generate solutions that are tailored to the specific issues your users are experiencing. You can use techniques like design thinking, which emphasizes empathy with users and encourages thinking outside the box. In these sessions, no idea is too small or too far-fetched to consider. The goal is to generate a wide range of ideas that can then be refined and evaluated for feasibility.
Once a broad set of ideas has been generated, narrow them down by considering the impact on user experience, technical feasibility, alignment with business goals, and potential ROI. Prioritize ideas that are most likely to effectively solve the issues identified earlier in the process.
Prototype and test
After ideating potential solutions, prototype and test the best ideas to see how they perform in real-world scenarios. In this way you can build confidence that those solutions solve the identified problems without introducing new issues.
The prototypes should be functional enough to simulate the actual user interaction but don’t need to be fully developed. Depending on the nature of the solution, prototypes can go from simple mockups or wireframes to more interactive versions of the product.
Once the prototypes are ready, conduct testing sessions with actual users. Design the sessions to observe how users interact with the new features or changes and look for any immediate reaction, ease of use, and whether the changes actually solve the issues they’re meant to solve. It’s also useful to collect quantitative data during these tests. Metrics like task completion rate, time on task, and user error rates can give you good evidence of a prototype’s performance.
Based on the results of these tests, refine your prototypes and don’t stop after the first one. Make adjustments based on user feedback and behaviors. You may require doing several rounds of prototyping and testing to get it right. Each iteration should bring you closer to a solution that solves the problem and also fits into the overall user journey and product ecosystem.
Implement and monitor
Once a prototype has been tested and iterated on, it’s time to build the solution and monitor its impact. Depending on the scope and scale of the change, you may choose to implement it in stages. For smaller changes, a full launch might be appropriate, while for more significant changes, a phased approach can help mitigate risks. This allows you to monitor the impact incrementally and make adjustments if there are unexpected issues.
Once the solution is live, use the same metrics that you used during the testing phase to measure the solution in production.
Monitor your main KPIs closely to check whether the changes achieve your desired outcomes. This should include user engagement metrics, conversion rates, and any specific metrics related to the problem.
Thanks for reading! If you liked this article and you’d like the get my book ‘Navigating the Product Galaxy‘ click here.
