Is your optimization program discovering and testing the best ideas?
If you’re like a lot of companies, you may be thinking, “What’s an optimization program?” Many companies still think of optimization as a project or tactic, and put little serious effort into building a real program.
But that’s changing fast. The world’s best retailers and B2B marketers alike have implemented programs with refined processes for ongoing, insight-driven optimization.
The maturation of the optimization industry is happening quickly. This is a train you don’t want to miss.
Recently, I gave a keynote presentation at the Conversion Hotel conference in the Netherlands. While there, I was invited to meet with Booking.com in Amsterdam and took a tour of their optimization floor. (Yes, you read that right… the optimization floor of the building.) They have one of the world’s most advanced conversion optimization programs with hundreds of optimizers split into 30+ cross-functional teams.
Imagine the growth and insights they’re gaining with all those constantly-running tests?!
Clearly, they have the traffic and resources to support that much testing. But where does that leave those without an optimization floor?
I’m glad you asked. It doesn’t require an organization of that size to create an industry-leading optimization program.
We train companies how to run world-leading optimization programs.
Widerfunnel has always focused our business on serving high traffic companies so that we can test everything we recommend. This allows us to accomplish our four service goals:
- Prove whether our ideas actually lift business results
- Learn from the results to build a strong test archive
- Look for patterns across all those tests to build optimization frameworks
- Create the world’s best optimization process that delivers consistent results
The process has been evolving rapidly since 2007, totally focused on optimization. A few years ago, we already had a detailed and rigorous process, which I sketched out on a San Francisco pub’s table cloth.
Since then, it has evolved considerably through many more experiments. We’ve been optimizing our process for the past eight years to get to this point. And now, we’d like to share our ongoing work.
Today, I’d like to officially introduce the newest version of Widerfunnel’s optimization program.
We call it the Infinity Experimentation Process℠.
What is the Infinity Experimentation Process℠?
The most advanced organizations view optimization as an ongoing strategy. It’s not a one-time project and it doesn’t have an end point where the company’s experiences are “optimized”. Widerfunnel’s Infinity Experimentation Process℠ shows how this works for leading companies.
I’ll first describe the process and then explain why it has been designed this way.
The process cycles between two distinct phases: Explore and Validate. The central point of the process is the nucleus where the energy is released. That is where growth and insights are revealed.
It starts with the Explore phase
In the Explore phase, the team gathers insights about the business and customers through many different sources. We tap into:
Business context – We gather information from the stakeholders, pulling in their expertise. The client stakeholders understand their business, products and services, clients, and history of learning. Of course, this information must also be understood as filtered through the individual’s perceptual filters.
Persuasion principles – There’s a lot of public research, especially in recent years, about why people buy. A simple search for behavioral economics, consumer psychology, neuromarketing, and even pricing psychology hints at the rapid pace of research. At Widerfunnel, we have a library of articles, books, and insights from these topics and more that we tap into. And, our team has a Literary League (currently with 15 books in the upcoming reading list.)
Digital analytics – Investigating web analytics, click map, scroll map, and session recordings help understand how people are using your website and where they are encountering difficulties. The goal is to understand real users’ journeys, e.g. where they are going, what they are looking for, and where they encounter conversion barriers.
User research – It’s amazing how much you can learn from simply speaking with your customers. Understanding the “voice of your customer” helps you speak in their words, which increases your marketing Relevance. You can learn their voice through onsite surveys, email surveys, NPS, user testing, etc.
Test archive – Over the years, Widerfunnel has developed thousands of tests, which give us insights across all industries of a spectrum of persuasion down to transactional elements, e.g. how to design forms, how should we test a value proposition. We use that bank of knowledge to create faster and better results.
All that data-gathering is centered around Widerfunnel’s framework for understanding customers’ conversion barriers and persuasion opportunities — the LIFT Model™.
The LIFT Model is the central point for understanding and categorizing all these bits of information within the 6 conversion factors: Value Proposition, Clarity, Relevance, Anxiety, Distraction, and Urgency. The process is a messy exploration process where we gather a lot of ideas. We come away with a bunch of potential insights that need to be validated.
The Explore phase is an expansive thinking zone, where all options are considered, whether they are minor usability tweaks or dramatic concepts that challenge the business.
The Validate phase proves great ideas
Validate is an iterative process that proves which of the ideas generated in Explore work and how they work best in real life scenarios.
We take the customer insights and LIFT points from the Explore phase to be tested in this 7-step experimentation cycle:
The PIE Framework – This is one of the most popular conversion optimization frameworks, which allows the optimization team to prioritize opportunities based on three factors: Potential, Importance and Ease. Here’s more detail about how to prioritize optimization opportunities.
Hypothesis creation – Insights are only validated once they are tested in controlled experiments. The hypotheses are only testable when they are constructed using proper hypothesis structure.
Design of Experiments – The Design of Experiments is a very important step that ensures your traffic is used most effectively. The optimization planner must make a hard decision about the balance between growth and insights. If your experimental design is poor, your tests will more frequently result in wasted time, effort and traffic.
UX/UI design – In optimization-oriented organization, form follows function. While UX experts can often contribute insights toward experiments, graphic design should follow after experiments have been prioritized and planned.
Development and QA – The technical details are critical to ensure reliable experiment results. If goals are improperly set up or any technical detail is missed, your experiments can be wasted. There are many common A/B testing goal tracking mistakes you should avoid.
Live A/B test – Experiments should be run using tools designed for running controlled tests. You simply cannot use regular web analytics tools to compare “before and after” results and expect to make good business decisions. The only thing worse than not testing is testing improperly and relying on bad results.
Results analysis – The most fun and important step is here, where the growth and insights are delivered. If your hypotheses and design of experiments are created effectively, you’ll be learning something important about your customer and lifting your revenue. At Widerfunnel, we call these Profitable “A-ha!” Moments.
Why is the process important?
In practice, Widerfunnel’s process includes 69 distinct steps (with 105 including sub-steps.) But, showing every single step would be impractical and confusing. I wanted to create a new diagram of the process that represented the core components.
The shape and connections are not just for visual aesthetics. The diagram intentionally communicates the most important concepts of Infinity Experimentation:
- This is an ongoing process. It’s not just a strategy or ad hoc plan. Hence the infinity loop shape with progress arrows.
- Two distinct mindsets are required. Expansive and reductive thinking are opposed in a productive way. This yin and yang thinking is one of the distinctions of the LIFT Analysis process we use.
- Explore is not linear. There isn’t a step-by-step cycle to the explore phase. It is necessarily variable based on context and open to inspiration.
- Validate is linear. The step-by-step process is in the correct order by intention and each step is important.
- Both Explore and Validate are ongoing. An Explore cycle may inspire a series of Validate experiments, and at the same time insights from the Validate phase can lead into more Explore information gathering.
- There are two equally-important outputs Both growth and insights are the intended and central result.
Implementing an optimization program in your organization
If you haven’t been seeing excellent results from your optimization efforts, it may be due to a lack in your process or team’s expertise. Those are the two most important components of the world’s best optimization programs. Whether you choose to adopt a process similar to Infinity Experimentation or develop your own approach, building a rigorous process will undoubtedly improve your optimization program.
Subscribe to get experimentation insights straight to your inbox.