CRO best practices

Why do marketers insist on looking for “best practices” and “conversion rate optimization tips” that simply don’t work?

A few years ago, I wrote a blog post called “31 Conversion Optimization Tips – How to Increase Conversion Rates”. Early in the post, I explained that these so-called “best practices” have limited value unless they’ve been thoroughly tested. They do not take your business context into account and probably don’t work in your specific scenario. Conversion rate optimization (CRO) is context-specific, which is why not adhering to specific conversion rate optimization best practices is a best practice.

Despite that pointed preface, that post remains one of WiderFunnel’s most popular. Marketers are always looking for the next magic conversion tactic that will take their site to the next level, but unfortunately, that magic tactic doesn’t exist. If you want to do CRO for real, you need a structured process that delivers both growth and insights.

In an ideal world, you’d have a fully dedicated team running your conversion optimization program. But, in reality, your resources may be limited. If you want more resources, you need upper-level buy-in. If you want upper-level buy-in, you need to prove the ROI of conversion optimization.

A structured process is your best chance at getting valid results you can use to build your internal team and take your marketing efforts to the next level.

Beware the ‘Hot-Shot Optimizer’

There are conversion optimization “experts” out there who promise quick wins without the hassle of process, but be wary of these ‘Hot-Shot Optimizers’. They love landing page templates and swear by design ‘best practices’. They track micro conversions as goals and call winning variations without statistical significance. They value tools over strategy and love to test button colors.

The ‘Hot-Shot’ might get a few wins in the beginning, but they may not be real wins. And soon, your conversion optimization efforts will hit a brick wall. Without in-depth research, informed hypotheses, proper prioritization and design of experiments, and comprehensive results analysis, you’ll never see the full benefits of CRO on your site.

You need a never-ending testing idea engine.

In this post, I’ll show you the most important components of building a testing idea engine — a structured process for conversion optimization.

We’ve been developing and refining the optimization process at WiderFunnel for the past decade: the latest version is what we’ve called the Infinity Optimization Process. This process is designed to maximize both creative and analytical thinking so that you never run out of test ideas.

The good news is that you can adopt parts or all of it in your optimization program today! Look at the frameworks below as examples of the principles your optimization program needs to succeed.

Optimization isn’t a project, it’s a continuous cycle.

There are two sides to successful conversion optimization: one is expansive, the other reductive. This is just like in any good brainstorming session: the idea generation mindset is separated from the evaluation mindset. It’s the same in conversion optimization. I call these mindsets Explore and Validate. The central point is the nucleus where the energy is released. That is where Growth & Insights are revealed.

The infinity loop shape of the diagram is intentional — this process is continuous and both Explore and Validate feed into each other through Growth and Insights.

wf-infinity-process-diagram_infinity diagram branded poster.png

The Explore phase is the ideation phase, where you gather information and look for patterns within data. In the center of Explore is the LIFT Model — our framework for understanding customers’ conversion barriers and persuasion opportunities.

The LIFT Model, which has become the industry’s most popular conversion optimization framework, is the central point for understanding and categorizing all the data you’ve gathered within Explore. It shows the 6 conversion factors that are influencing your sales right now: Value Proposition, Clarity, Relevance, Anxiety, Distraction, and Urgency. If the Explore phase is done properly, you’ll come away with a bunch of insights that can then be validated.


There’s a lot of nuance to how this is done, which can only be learned by experience with testing on many different scenarios. You should watch this webinar recording for case study examples to learn more about how Explore reveals insights and growth. In one example on the webinar, I detail how an Explore insight led to a 26% lift in sales for one company. Could you use similar insights?

The Validate phase proves great ideas.

Validate is an iterative process that proves which of the ideas generated in Explore actually work.

Validate takes the customer insights and LIFT points from the Explore phase to be tested in this 7-step experimentation cycle, as follows:

  1. The PIE Framework

This is also a popular conversion optimization framework we developed, which allows your optimization team to always focus your resources on the most important opportunity. You simply prioritize opportunities based on three factors: Potential, Importance, and Ease. (In practice, we use many data points as sub-factors that feed into the top-level PIE numbers, but this is a good conceptual starting point for you.)


  1. Hypothesis creation, which converts the “weaknesses” identified in the LIFT Analysis and turns them into strengths to be tested.
  2. Design of Experiments (DOE). Design of Experiments is often overlooked, but this is where we spend most of our effort at WiderFunnel. To clarify, I’m not referring to graphic design, but rather how you’re structuring your experiments. Your DOE will determine the outcomes of your experiments in both insights and growth. If your experiment is designed to isolate insights, for example, losing tests will never be losing tests because you’ll have uncovered insights that you can build upon!
  3. UX/UI design, which can cause problems if you’re using a designer without experience in design for conversion. The details count!
  4. Development and QA, ensuring your code approach doesn’t cause technical problems for your A/B test results.
  5. Live A/B test, where you monitor results and make tactical decisions to prune or accelerate variations based on performance rules.
  6. Results analysis. If your hypotheses and Design of Experiments are created effectively, you’ll learn something important about your customers and lift revenue. At WiderFunnel, we call these profitable “A-ha!” Moments.

The Infinity Optimization Process enables you to take your A/B testing to the next level, allowing you to uncover and test the persuasion triggers that move your users to buy.

A mobile e-commerce subscription company discovers its customer motivation.

An e-commerce client has a subscription model: users sign up for a monthly mystery box of “Geek + Gamer gear”. We ran several experiments with the intention of testing the understanding of this client’s customers’ motivations. We wanted to know if users responded to hints of social inclusion or hints of product tangibility and savings.


Social proof variation

After a few rounds of testing, the client saw a 14% lift in sales using a combination of social inclusion and social proof. This customer tendency is one that the client can now test on other areas of their site.

Social proof clearly works…for this audience, in this context. But what works for one organization may not work for yours. You should always test that!

So, how does one come up with important ideas like this? That’s where the PODs, POPs, and POIs can come in.

This final framework I’ll share with you today is meant to help you understand how to create powerful questions about your value proposition and the persuasion triggers that motivate your users to act.


Many marketers are focused on Points of Parity: these are the features you offer that are important to your prospects that you also share with your competitors. Marketers should focus, instead on the Points of Difference: these are the features that are important to your prospects and are not available from your competitors.

Do the work to understand your audience. Look for patterns in user behavior and test the different approaches to your value proposition that are unique to you (refer to those PODs). A quick tip! Landing pages are the best place to test your value proposition because visitors to these pages are exempt from the rest of the messaging on your site.

The Infinity Optimization Process over time

Here’s a visual of our A/B testing engine in action, over time (this diagram is based off actual testing we’ve done for one WiderFunnel client). We start with one test on the homepage that reveals 5 insights.

In the next round of testing, two of these are validated, two are invalidated, one is inconclusive. We also see that an insight from a separate experiment on a content page has been invalidated.

And on and on it goes. The testing ideas never end.

As you can see, conversion optimization takes dedicated time and resources to be successful. It can’t be a side project for a single data analyst. If you want an endless stream of test ideas, you must build your strategy, build your bank of insights and store those insights so that they are accessible.