Unlocking the True Power of Testing & Other Takeaways from Brooks Bell’s Interview With Ambition Data

Recently, our Founder and CEO, Brooks Bell, sat down with Allison Hartsoe, host of the Customer Equity Accelerator—a podcast produced by Ambition Data. Listen to the full podcast or read on for a few highlights from their conversation:  On what inspired her to build an experimentation consultancy… Originally, Brooks founded Brooks Bell Inc. in 2003 as […]

The post Unlocking the True Power of Testing & Other Takeaways from Brooks Bell’s Interview With Ambition Data appeared first on Brooks Bell.

Recently, our Founder and CEO, Brooks Bell, sat down with Allison Hartsoe, host of the Customer Equity Accelerator—a podcast produced by Ambition Data. Listen to the full podcast or read on for a few highlights from their conversation:

On what inspired her to build an experimentation consultancy…

Originally, Brooks founded Brooks Bell Inc. in 2003 as a website development agency. After working with a few local clients, a chance introduction led to her first major experimentation client, AOL.

Today, you might think of AOL as one of the [now-extinct] internet dinosaurs, but even back in the early 2000s, the media giant was facing its fair share of challenges. According to one story by Time Magazine, despite having 34 million members in 2002, AOL was battling slowing subscriber growth, falling ad revenue and exorbitant operational costs. 

So, the company turned to experimentation. “AOL had the right environment to build a testing culture,” said Brooks. “They had a closed technology environment, their own analytics platform, and their data was clean and connected.”

Back then, AOL relied on pop-ups to drive new subscriptions. Working with Brooks, the company issued a challenge: design a new subscription pop-up that would beat the control experience. And so, drawing from her background in design and psychology, she did—and then she did it again, and again, and again.

But that was just the start. As other large companies began to rely more on the digital space to drive their business, Brooks saw an opportunity to help them tap into the power of experimentation.

“We realized that no one was testing!” said Brooks. “No other large companies had the data, culture and processes in place to test. So we set out to help them build the data fidelity and really recreate what we saw at AOL in those early years.”

On the difference between optimization and experimentation…

It’s one of the more common questions we get: “Brooks Bell is an experimentation consultancy. What’s that? What’s the difference between experimentation and optimization?” As Brooks explains it, it all comes down to science.

By definition, experimentation is the application of the scientific method to determine something. And while optimization is one potential outcome of an experiment, true experimentation requires running tests without a prescriptive outcome or application.

To put it simply – you’re testing to learn. And as long as your results are statistically significant, there is always something to be learned from experiments—even those with flat or negative results.

On how to unlock the real power of experimentation…

Today, in the age of Amazon, a customer-centric experience is critical. But for some established companies, this requires a bigger paradigm shift in culture and processes.  

“Customer-centricity requires rethinking metrics, the type of data you collect, how teams are organized, how teams are incentivized, how you communicate and also your core values,” said Brooks.

The true power of experimentation lies in its ability to align your customer needs with your company’s strategic goals and your program’s agenda. Furthermore, you can use experimentation to learn new things about your customers in a scientific way.

“Having statistically-sound customer insights can totally change how you organize your store, how you train your team, and how you structure your website,” said Brooks. “This is where testing programs can really drive change.”

To that end, we recently celebrated the launch of Illuminate, our customer insights software for testing teams and executives. Illuminate not only provides a place to store, share and learn from your experiments, but also a means to develop impactful customer insights.

“We launched Illuminate to provide a repository of great test examples, to learn from each other, and to build a library of great test case studies,” said Brooks.  This is because outside of the testing program, any key learnings from an experiment can get lost within the data. Illuminate solves this by encouraging deeper thinking about customers, their needs, preferences, and behaviors. 

Learn more about Brooks Bell’s experimentation consulting services. 

The post Unlocking the True Power of Testing & Other Takeaways from Brooks Bell’s Interview With Ambition Data appeared first on Brooks Bell.

9 Things You Can Do Now to Ensure Your Experimentation Program Survives a Reorganization

It often begins with rumors and murmurs. Then, perhaps, a shift in executive leadership. At first, changes seem minor or isolated. But eventually, the inevitable becomes reality—a reorg is underway. Because testing programs typically work with and across several divisions in a business and don’t fulfill a traditional business function, they are particularly vulnerable during […]

The post 9 Things You Can Do Now to Ensure Your Experimentation Program Survives a Reorganization appeared first on Brooks Bell.

It often begins with rumors and murmurs. Then, perhaps, a shift in executive leadership. At first, changes seem minor or isolated. But eventually, the inevitable becomes realitya reorg is underway.

Because testing programs typically work with and across several divisions in a business and don’t fulfill a traditional business function, they are particularly vulnerable during the organizational upheaval.

However, to paraphrase Benjamin Franklin, with an ounce of prevention, it’s possible to avoid a pound of problems. Here are nine practices for surviving a reorg that you can start doing nowbefore you ever have to deal with one.

1. Calculate Impact

The first thing any new leadership will want to see is the value the testing program provides to the business. Of course, “value” is a multidimensional concept. It includes the contribution to business goals and priorities, insight into customer preferences and behavior, development of new innovations, and minimizing certain opportunity costs.

While it’s true all of these things will help communicate the importance and value of testing, nothing will be as compelling as a big annualized impact number. This number can be complex and time-consuming to calculate, but even if it isn’t an important metric for your program today, it’s worth having an analyst crunch, record and update just in case.

2. Write an Elevator Pitch

If you had five minutes or less with the CEO of your company, could you clearly communicate the mission, focus, and value of the testing program in a memorable way? If the answer to that question is “no,” or even “maybe” it’s worth spending some time crafting an elevator pitch for your testing program.

Open your pitch with a short anecdote about a problem that was solved with testing. Next, add a sentence or two about the essential mission of the program. Follow this with a sentence about the methods used. Close with a statement about the contribution the testing program has made to the business as a whole.

Once you’ve written this pitch, practice delivering it to your friends, family, team and other stakeholders. After a lot of practice—and incorporating the feedback you will inevitably receive—you’ll be ready to introduce testing to any new leader or team you may encounter.

3. Archive Results

Surviving a reorg isn’t all about making a case for testing. It’s also about maintaining a consistent pipeline after teams have been shuffled around.

The first step to protecting the testing pipeline is to create a complete, detailed, navigable archive of past test results. This is critical for developing new ideas, training new team members, orienting new teams and stakeholders, and simply making the case for what works and what hasn’t.

Recently, we launched Illuminate, our new software for enterprise-level testing teams. Illuminate offers an executive-friendly repository of your tests and any insights you’ve learned about your customers along the way. It’s direct integration with Optimizely, easy-to-use reporting tools, and custom case study generator significantly simplify the task of archiving and reporting your test results.

4. Outline the Process

Testing is complex and when teams get rearranged, processes that once flowed smoothly can become intractably clogged. To prevent this, document the process as it exists and identify areas of parallelization, possible redundancies, problematic bottlenecks and opportunities for redirection.

If you have access to a project manager, ask her to run a few what-if analyses to estimate potential problems if your processes were to be disrupted. Then, work together to develop possible solutions or workarounds to the most likely scenarios.

5. Clearly Define Roles and Responsibilities

Having a fast and nimble all hands on deck approach to managing the testing process is great until a critical person leaves the company or is assigned to a different team.

To avoid this, define the roles and responsibilities of each team member at each stage of the process. Doing this is critical for mapping the resource requirements of the testing process and quickly identifying gaps if the team is restructured.

It’s also important, however, to track ongoing responsibilities and duties in a more specific way. Having a project management platform or system that identifies what stage of the process each test is in, and which team member is responsible for that task is essential for avoiding disruptions.

6. Develop Training Programs

Looking on the bright side, a reorg could mean your testing team is greatly expanded. It might also mean fewer people are doing more, including jobs they have little experience with. In either case, having a developed and ready-to-execute training program is helpful.

Like all of these tips, the best time to develop a training plan is not the first day your new team member walks into the office. Instead, start training and cross-training your existing team right away. This gives you an opportunity to develop extensive content, deliver it, get a sense of what works, and make adjustments before a reorg renders training critical to the continuation—and not just the improvement—of the program.

7. Centralize Documentation

Having lots of documentation is useless if no one can find it. Moreover, it doesn’t help if it isn’t standardized in some way.
The archive of results, process documents, test plans, training materials, and everything else should be stored in a public or shareable archive, in a format that is easily accessible and navigable.

Using filename conventions, consistent directory structures, and standard documentation practices across the team may be mundane, but it’s just as important to the robustness of the testing program as tracking each person’s ongoing responsibilities.

8. Get essential access

One often overlooked consideration is whether the testing team has access to the essential technologies on which it relies.

Even if most development is done by an outside group, it’s important to have access to tools that enable you to upload and modify your code. Additionally, if reports are pulled by a sovereign analytics team, it’s equally important for someone from the testing program to have the access and ability to do so in a pinch.


Some solutions—like tag management systems—address this challenge. Training, cross-training, and collaboration is another helpful way to build the necessary competencies to get access to and make basic use of all your testing tools.

9. Keep it all up to date

Building the previous eight resources can take a lot of time and effort. Many teams will make any one of them a goal for the quarter, work hard to get it done, drop it in an archive, then forget it.

Months, maybe years, pass without giving the resource a second thought. Then, a sweeping reorg happens and the five-year-old process document is unearthed, dusted off, and found to be frustratingly obsolete. That’s why you must take care to not only produce these resources but maintain them as well.

A reorg can be a scary thing for a lot of reasons. However, by following these nine tips today, even the biggest organizational shakeup doesn’t have to disrupt the flow and productivity of the testing program.

The post 9 Things You Can Do Now to Ensure Your Experimentation Program Survives a Reorganization appeared first on Brooks Bell.

How to Convince Your Boss to Invest in Experimentation: 5 Hard Truths

Would you rather have a root canal, or try to build something completely new at a huge, enterprise-level company that seems to be held together with nothing but red tape and internal politics? That may seem a bit dramatic, but obtaining buy-in for experimentation is a difficult challenge faced by many of our clients. Over […]

The post How to Convince Your Boss to Invest in Experimentation: 5 Hard Truths appeared first on Brooks Bell.

Would you rather have a root canal, or try to build something completely new at a huge, enterprise-level company that seems to be held together with nothing but red tape and internal politics?

That may seem a bit dramatic, but obtaining buy-in for experimentation is a difficult challenge faced by many of our clients. Over the last 15 years, we’ve identified some hard truths about getting others to invest in testing. And in the spirit of “being real,” (one of our core values) we’ve decided to share them with you.

1. You have to bait the hook to suit the fish.

“The world is full of people who are grabbing and self-seeking. So the rare individual who unselfishly tries to serve others has an enormous advantage. He has little competition.” – Dale Carnegie, How to Win Friends & Influence People

We’ll take good ole Dale Carnegie’s advice on this one: if you want to convince someone to do something, you have to frame it in terms of what motivates them. In order to do that effectively, you have to be able to see things from their point of view.

“My go-to strategy for gaining executive buy-in for testing programs is to focus my discussions with them on the business impact of a well-run and successful testing program,” says Kenny Smithnanic, Director of E-Commerce at Ultra Mobile. “Most executives I’ve worked with want to see better company performance, even if they aren’t directly evaluated on this. So, I tend to discuss the exact revenue, orders, leads, and/or average order value increase that our business could realize from the accumulated gains of a successful testing program.”

When you’re pitching your experimentation program, position it as a program that works in service to other departments. Consider also your company’s big picture objectives and pinpoint where experimentation can work in direct support of those.

2. You’re gonna lose if you make it all about winning. 

Managing expectations is critical. Optimization is as much about learning about your customers as it is about landing some quick wins. And learning sometimes requires failure. Or, in this case, a few flat or losing tests. In fact, the industry win rate is actually 25%. If the expectation is that every test has double-digit lifts, it’s going to be a rough road ahead. 

All that said, here are a few strategies to manage expectations for your program.

Include executives in the ideation process. Let them see how your team is using data, generating lots of great testing ideas and working collaboratively. Including them in this process also provides them with a sense of ownership and investment to see how a strategy performs.

Develop a consistent (in frequency, timing and format) reporting structure. As you’re reporting the results of each experiment, begin by framing it in the context of your company’s larger goals. Don’t make the mistake of only focusing on the winning experience. Rather, walk your stakeholders through each variation individually and the resulting insights based on their performance.

State what’s next on your priority list, based on these results. This builds anticipation for your next round of tests, setting you up for future success, regardless of individual results.

3. Proving people wrong doesn’t convince them you’re right.

Avoiding pushback to your ideas requires three things:  Knowing what you’re up against, anticipating challenges and being open to others’ solutions.

Because executing tests often requires cross-functional support, you may have to take a test and learn approach to your experimentation program. You might also have to make a few compromises or adopt short-term solutions along the way.

Ultimately, if you’re seeking to make experimentation a core competency at your organization, you have to bring others along in the process. This means that as problems arise, it’s okay to recommend a solution, but you also have to be open to trying others’ solutions as well.

Only by solving problems together will you be able to build a high-functioning experimentation program and team.

4. You’re not going to blow anyone’s mind by using big words or snazzy acronyms.

In fact, when you’re trying to get people on board with testing, using super technical jargon can actually work against you.

Consider, for instance, the difference between “experimentation,” “testing,” and “optimization.” To you, these terms might mean the same thing, but to your boss, they may evoke different meanings.

So when you look to pitch your experimentation program, use terms that are familiar, accessible and aren’t at risk of being misinterpreted.

The same goes when introducing testing terminology to the broader organization. There are tons of acronyms associated with experimentation—RPV, AOV, UPT—the list goes on. Be sure you are decoding them until your team is more familiar. Revenue per Visit, Average Order Value, and Units per Transaction communicate far more meaning than an acronym.

Finally, consistency and intentionality are key here. Develop a common language for experimentation through trainings, lunch-and-learns and/or a company-wide roadshow. Finally, be sure that these terms are reflected in your internal processes: within meetings, in status reports, playbooks and in other means of communication.

5. Getting buy-in is not a one-time event.

You have to lay the groundwork. You have to build trust and credibility. You have to follow through. And through it all, you have to be a generally good person to work with.

Experimentation is amazing because it is supported through quantitative data and statistically significant results – making it a very quantifiable and persuasive case,” says Suzi Tripp, Senior Director of Experimentation Strategy at Brooks Bell.  “When you can share an experiment and tie it directly to incremental gains, it’s a powerful statement.”

But even if you get executive support for your program, you simply can’t expect everyone else to get on board immediately. In fact, having that expectation can lead to conflict and tension between teams, and cause a lot more internal problems.

To address this, we suggest working gradually: start out by working with one group, generate a success story, and share it through your internal communication channels, and keep going from there. Only by sharing your stories will other departments begin to be interested in harnessing that success as well.

Have ideas for other strategies to obtain buy-in for testing at your organization? We’d love to hear what’s worked for you. Share your story in the comments.

The post How to Convince Your Boss to Invest in Experimentation: 5 Hard Truths appeared first on Brooks Bell.