12. Faster, Better

Building Agile Processes into the Journey

During Twitter’s glory days of massive growth from 2010 to 2012, the company wasn’t doing anything drastic to produce the results it was seeing. It didn’t advertise more or increase its media coverage or improve its branding. In fact, it changed only one thing in its growth model: it simply ran more tests. This is the revelation Satya Patel, former VP of product at Twitter, shared at the 2014 Agile Marketing Meetup.¹ Beginning in the first quarter of 2011, Twitter began running ten tests each week, up from one every two weeks. This “rapid experimentation” allowed the company to find the experiences and techniques its customers were wanting and to trigger those experiences at opportune times. The result? Twitter grew faster than ever before (see figure 12-1).

In their book Hacking Growth, Morgan Brown and Sean Ellis view such rapid experimentation, or High Tempo Testing as they call it, as the epitome of modern creativity—and key to helping companies move ahead. Besides Twitter, the authors look at companies like Facebook, Uber, Dropbox, and Airbnb, peeling back the layers to explain their explosive growth. Like Twitter, those companies’ success had nothing to do with creative PR campaigns or better use of tactics. Rather, as Brown told me in a recent conversation, “They simply learn faster than all of us.”

In other words, scaling your business isn’t a matter of coming up with the most creative idea. Instead, it’s all about quickly testing assumptions to arrive at an optimum result. What’s more, you’ll need to apply rapid experimentation across the entire customer journey to achieve the best results, because such iterative testing improves any experience. Recall from earlier in the book that even a small improvement in efficiency has a massive impact on revenue—a mere 1 percent increase in efficiency across the first four stages of the journey creates a 40 percent lift in net new revenue. Therefore, we must see the entire customer journey as an experiment in which our job isn’t to “get creative” but instead to constantly test our assumptions across the journey and improve every day.

The process of rapid testing is called agile, and it is a key trait of high- performing marketing organizations. While there are other modern methods for production that sound similar, such as Fail Fast or Lean, they are not the same. Lean is about removing waste from your current production process, and Fail Fast simply refers to taking more risk. Both are improvements to your current methods, while agile is an entirely new process of constant iteration that aligns the brand and the customer. In 2018, the Standish Group found that agile projects succeed two times more often, at a quicker pace, and at a significantly lower cost than projects following the traditional production process.² Salesforce research concurs: high performers are ten times more likely to use agile methods than are their underperforming counterparts.³

Consider agile as simultaneously a tactical approach, an organizational structure, and a new lens through which to view all things. You will need such a lens, given all that the transition from traditional to contextual marketing practices requires: laying considerable groundwork and maintaining a relentless focus on creating contextual experiences (and all it implies) amid changing consumer demands—all while managing hundreds of automated programs that run simultaneously. What’s more, the bar for what constitutes a “better” experience for consumers will continually go higher, putting marketers in a very tough spot: we must find a way to consistently produce more experiences at a higher value. Agile is the answer to this problem, and a core skill of the context marketer. (In case your team needs convincing, however, see the sidebar “How to Advocate for the Power of Agile.”)

This chapter shows you how to use agile to improve and speed up the customer journey, how to build journeys using data-backed testing, and how to use agile backlogs to stay focused and get the highest return on your effort.

Crafting Better Journeys in Less Time

Agile has many applications, and customer journeys are a big one. Although the people who came up with the idea of agile were software developers trying to create computer programs, agile processes solve the exact same problems for marketers who are crafting an endless mix of experiences. That’s because marketers face challenges similar to those of developers. Think about it: developers were separated from their end users; they were trying to build products while having very little context regarding what they were solving; their projects were based on assumption without testing; they were constantly asked to make changes to the deliverable as the projects progressed; and as a result, projects would miss timelines—and not really solve the actual needs of consumers. Sound familiar?

How to Advocate for the Power of Agile

In case you’re faced with team members who really can’t understand the concept of agile or why they should change the way they create brand experiences, you can show them a very simple proof. It’s an example of the power of agile to produce anything—better and faster.

Take a sheet of paper and draw two dots on opposite corners, like this:

Ask team members to replicate your drawing on their own sheets of paper. Then ask them to place their pens on one of the dots, close their eyes, and—without picking up the pen—try to draw a straight line to the other dot. Ask them to stop when they think they’ve gotten there, and to open their eyes. Let them know they can take their time and can begin when they are ready. When they are finished, their lines will most likely look something like this:

If they happened to hit the dot, which is rare , you need to show that it was pure luck, not proof that the method is correct. To prove that success using such an approach involves sheer luck, simply ask the team members to use the same method to hit the mark again, three times in a row. The odds are they can’t repeat the feat. (The approach they just used mimics what’s commonly referred to as the Waterfall method of production, where the full plan is drawn up in advance and then executed in cascade fashion without reflection, such as on an assembly line.)
Now ask them to try again, but this time following the more iterative process of agile. Ask them to place the pen on the dot, close their eyes, and start drawing—but to stop at any point they wish, keeping the pen on the paper. At that point they should open their eyes and recalibrate (without lifting the pen). Then ask them to close their eyes again, and continue forward. Each of the team members’ second lines will likely look like some version of this—more of a zig-zag pattern than a line, but reaching the dot nonetheless:

You can repeat the agile exercise as many times as you like, and you’ll see that the agile line will always reach the mark, faster and with less energy. That’s because agile is the only way to reliably create a better experience, and it is scalable.

But then the software developers realized something: the problems they confronted were simply by-products of the processes they were following. Their solution? Craft a new process—agile. Founded on a core set of principles, agile sought to provide the why that outlined the process and the structure. Agile as a process is based on multiple iterations rather than a single attempt, and it values objective feedback over subjective opinions. As a structure, agile focuses on collaborative teams rather than siloed channel teams.

We can think of agile processes, therefore, as springing from the structure of the problem itself. How to create the most effective and efficient journey? By involving all parties in a process of collaboration and testing, agile journeys simply produce better outcomes faster. Collaboration happens both internally and externally. Customers are involved at each stage of the building process, where their feedback combined with quantitative results provides the direction for the next iteration. Internally, collaboration is created by a new team structure called a pod: a diverse group of skill sets working together to test ideas, gain instant feedback, and iterate to create a better experience. (See the sidebar “Capital One: Case Study of an Agile Team.”)

Such teams—pods—follow agile processes, as conceptualized by the image in figure 12-2.

Specifically, the agile process begins with user stories. These stories tell you how the person is naturally engaging across the journey, which is why the first chapter of part three outlined in detail how to obtain those stories (a.k.a., market research). Then the process moves to creating a minimum viable experience (MVE), which the pod crafts as a hypothesis. The MVE is not a finished product or experience; rather, it’s only the first test, usually the easiest way to test a basic assumption.

For example, Santander Bank gave up long production schedules and lengthy approval processes to execute a single marketing effort, switching instead to employing many smaller low-risk (MVE) efforts that were quickly launched and measured. Successful efforts were reinvested in and grown, while unsuccessful efforts were abandoned. A case in point is the bank’s sponsorship of Boris Bikes—a public bicycle-share company in London. Santander began with a small number of bikes and pickup points. Before deploying new bike stations, the bank sought customer feedback and data—such as, what are the most commonly used bike routes—to decide where and how to best grow the effort. Santander’s agile approach across its marketing efforts has led to a 12 percent increase in loyalty and an account satisfaction increase of 10 percent, which is the bank’s highest level of customer satisfaction in seventeen years.⁵

Capital One: Case Study of an Agile Team

What do Alec Baldwin, Samuel L. Jackson, Seinfeld’s dad, and a Viking all have in common? They’ve all asked you, “What’s in your wallet?” That’s the well-known catchphrase of Capital One Financial, the progressive credit card company that pioneered new credit models, online banking, mobile banking, and now agile marketing. You may not think of a financial company as being progressive, but this one was founded, in 1988, on that very idea: that the power of information, technology, and testing could be harnessed to bring highly customized financial products directly to consumers.

Indeed, the bank was one of the first companies to put agile to the test on a large scale in a marketing department. Through conversations with past and present unnamed sources familiar with the effort, I was told the organization viewed this as a pivotal business decision. The message to the marketing team was, “We moved to agile, and burned all the boats. There is no turning back.” Capital One made the move for a few key reasons. First, it had used agile in other parts of the organization for many years, so it was already a core of the business.⁴ Second, the executives saw how divisions outside of technology could benefit from agile as well. They tried out the idea in a very small niche test, watched by management. They liked what they saw and went all in.

As a result, I was told by my source, not only has the quality and speed of marketers’ work improved, but also the fundamental way people viewed their work. Agile allowed them to align everyone’s goals, and the process supported and reinforced this alignment. Agile helped shift everyone’s concept of marketing from subjective to measurable and optimized efforts for the most value per unit of time possible. By reorienting marketing around agile, suddenly the division of 130 people became dynamically responsive to customers’ needs.

My sources said the key to the bank marketers’ success was how they structured and built the agile teams. To begin, they identified work that could easily be done in an agile format, where multiple iterations were possible. Work such as content creation, digital experiences, and customer journey management are easy places to begin since they allow for rapid deployment and instant feedback. Capital One chose twelve specific teams, almost half of the marketing organization, to produce work in an agile way. (Some teams, such as the Events team, which needed to follow sequential processes, were not moved to agile initially.)

Each of the new teams was renamed as a “pod,” and the pods were broken down into four layers, with an accountable executive (AE) at the top, then a project manager (PM), a project owner (PO), and finally the trades. The trades consisted of copy writing, graphic design, art direction, data analysis, and coordination. They were supported by external partnerships, such as channel sales, to support larger business goals.

While the AE determined the team’s direction, the PM was in charge of designing the time frame for the project and the objectives the team intended to achieve at each step. The PM answered questions such as the following: What does MVE look like? What will it take to make this happen? Which channels should we use? In what way? The PM and the PO then worked together on a detailed plan, with the PO setting the intent and priority of the outcomes. Next, the project would be decompressed into tasks that were placed into a queue. The PO and the entire pod would then do a forced ranking on those initiatives and break up the work. The team then took the top three things and put those into production. The rest went on the backlog.

The pod used a standard agile process. It began by collecting user stories, moving to a quick production cycle, followed by review and iteration. As a part of the review, the business analyst showed the pod the results, and the pod would fold the findings back into the planning. Sometimes the data might change the order and priority of tasks and move things into or out of the backlog. In addition to reviewing the project, the pod would also look at whether the process used was efficient. So pods were iterating not only their work but also the process of how the work was done.

Each pod was supported in two key ways that allowed it to succeed. First, each had external partnerships, such as key technology administrators and subject matter experts. But more important, each was assigned an agile guide. Since Capital One had already been using agile on the product side, the bank provided its marketing pods with experts in the agile process. These agile experts became internal consultants, aiding the agile teams and the executives to ensure the effort was a success.

Technology and innovation have been the core of Capital One, allowing the company to create variable-interest-rate credit cards before anyone else, and leading it to the fruits of agile marketing. Moreover, the bank’s focus on innovation and technology has placed it on InformationWeek’s list of the five hundred most innovative users of information technology as well as Fortune Magazine’s list of the World’s Most Admired Companies, and earned it excellent placements on yearly Fortune 500 lists. It is also the ninth-largest bank in the United States. These are massive achievements for a relatively young company, and one in an industry traditionally lacking innovation.

Changing the structure of your entire marketing organization is likely one of the most extreme things you can do. Yet Capital One has demonstrated the necessity today of such radical change—both in how we think about marketing and in how we execute it.

Returning to figure 12-2, the MVE is followed by a review of the test, utilizing more customer feedback and quantitative data to determine what the best next steps are, and then the pod iterates again. The power of the agile process comes from that cycle of collaboration and iteration. Such constant feedback, collaboration, and testing allows an idea to reach its highest value every time. By iterating through this process multiple times, you decrease the risk for failure and increase the value of the experience because you are building from the previous learnings.

Data-backed Testing—not Biased Assumptions

Traditionally, marketers follow a customer journey strategy that takes months to carry out. First they diagram on a massive whiteboard the series of steps they envision. They take pictures of it and send those to the brand campaign team to build out. The images reveal the loads of new content that needs to be written, shot, and produced. Then there are accompanying programs to be engineered, landing pages to design, and more.

But even if you’ve based that campaign on consumer stories and interviews, you are still building it with bias. Just because you think your content and brand experiences will work, that doesn’t ensure they will. Without testing your assumptions, you’re betting a whole lot of time, effort, and money on your own unproven bias.

By contrast, an agile process builds customer journeys in blocks or small steps, where each set of experiences is created and tested iteratively until the result is clear—using real data rather than biased guesswork. This process of partial creation and iteration is quicker and much easier to balance with your regular workload. Rather than investing a lot of time to build one big program, you’re building one step of a program at a time. That allows you to switch between multiple programs, optimizing each one as you go. The process might look something like figure 12-3.

The first and second actions in the journey are the test—will the customer open the email you sent?—while actions marked as 1, 2, 3 have yet to be built. They will be built in time, using the data gathered from the prior test.

FIGURE 12-3

Creating an automated journey using agile

Just how long you have to build the next steps is totally up to you, but use the time between actions as your guide. In this example, the automated program starts out with a two-day pause, meaning a person is added to the program and two days later the first email goes out. The pause in this example is marked with the small clock and a number of days above the action. If the person engages, the program waits two more days before its next action; and if the person does not engage, it waits three days before it takes an alternative action. By adding up the pauses, you can see the brand has six days from when the person first enters the program until the next action (represented as box #1) in the program has to be built, leaving ample time to review and iterate on the first steps.

Following the same logic, the brand has eight days to create the action marked with #2, and twelve days to create action #3. This process continues until the journey is complete, at which point it will have been built over time, and optimized with the data gained from each test, rather than biased guesswork. The result is that you can actually build your brand experiences faster—and achieve better outcomes.

By relying on audience engagement and changes to that engagement over time, any bias on the part of your marketing teams is removed. What bias, you ask? Let me be clear about how much bias drives marketing decisions. Ask yourself, when was the last time you picked up the phone to talk to someone who had engaged with your marketing, with the express goal of learning how to make that experience better for the next person?

I’ve put this question to tens of thousands of marketers from around the globe, and less than 1 percent of marketers have ever made such a phone call. Now think, what product manufacturer would produce a product and never ask its audience what they think of it? Agile allows marketers to transition into a new world of building from data and executing via automation, and the review process is a critical component.

Some marketers make excuses to justify their lack of review. They say, “Customers don’t want to talk to us,” or they believe their customers will have their phone calls blocked. Sure, not everyone wants to talk to you, but many do and are happy to. To increase your odds of success, make sure you have a list of about twenty people you can call with a set of quick, easy, and specific questions. Personally, I find the best system for reviewing an effort is to ask three questions:

What brought you to that moment?
What were your expectations, and did we meet them?
Have other brands produced better experiences? Which ones?

I’ve found that calling twenty people and asking these questions will net you around six conversations, providing you with ample information to guide your iteration. The answers to these questions will give you the insights you need to refine your strategy, reach more people in context, and meet their desire in the moment. Once your program has been completed, much of it can run on its own. In time, you’ll likely have hundreds of programs running across the entire customer life cycle, leaving you time to focus only on managing programs and introducing the refinements demanded by changing conditions. You’ll begin now to manage your programs in an agile way, using data to focus you on the experiences needing your attention in the moment, driving the most value for your brand with your effort.

Using Agile Backlogs to Keep You Focused

Now that you’re starting to see a new world where you build your customer-journey programs in iterative steps, and where your daily tasks shift to managing and optimizing hundreds of automated programs, there’s one final aspect of agile that’s critically important: producing the highest value per unit of time. Key to this is the backlog.

What makes agile so valuable is its ability to focus a team on the most important short-term goals for testing ideas and immediately incorporating insights. In other words, agile helps brands focus on and produce whatever experience has the highest value now. But since time is limited, that means a lot of good ideas won’t take priority—and could easily get lost (and often do). That’s why backlogs are a crucial element of the agile method.

Backlogs are living lists organized by the marketing team’s collective view of an item’s business impact. At the top of the list, the team places the make-or-break items for reaching business goals; at the bottom are nice-to-haves. To determine this “collective view” of where an idea or task belongs on the backlog, teams use open discussion coupled with a simple 100 Point exercise.

The 100 Point exercise asks team members to assign points to ideas and tasks that, in their view, have the biggest impact on business goals (not marketing goals, but goals for the business overall). Each team member gets 100 points to assign to the tasks on the backlog. Points are awarded based on how critical the item or task is believed to be by the team member. Whereas a believed business-stopping item may receive thirty points, a new blog post may receive only one point. Once everyone has deployed their points, the totals for each item are then added up and the priorities reordered, with the items winning the most points at the top. This ensures that the projects with the biggest impact on the business are completed first, with the others taken on in turn.

Backlogs also help your team manage its new workload, with its many moving parts. New requests are added to the backlog and prioritized by consensus; so instead of tackling items based on when they come in, they are completed only when the team rates them as the best use of resources. That way, the backlog gives each team member the leverage to say no when there’s a better use of time, which leads to one of the biggest benefits of the agile process: the team’s sanity. You not only have a powerful way to rank tasks based on business impact—not on who has the loudest voice—but you also have a way for team members to communicate their goals and achievements that the rest of the organization can understand.

The concept of the backlog also implies a new way for the marketing team to operate day to day. Rather than working on projects from whoever has the loudest voice, team members stay focused on producing the highest returns. Customer journeys are monitored and constantly optimized, and daily tasks are sorted based on potential impact.

I’ll close this chapter with an answer I got to the question, What’s the best part about agile? “The big win is in people looking at their work in a fundamentally new way,” said one of my unnamed sources from the bank. “They are now very focused on how we add the most value with the least amount of effort.” Capital One should know: the bank made a big bet and moved the operations of its entire marketing team to the agile method and hasn’t looked back.

That focus on value is an important part of making the case for context marketing. As the role of marketing takes on a bigger piece of the business, it is imperative that we seek not only new marketing ideas and tools but also new ways of working in order to handle the expanded task load. Agile can sound like a complex answer to a complex problem, when in fact—like most brilliant ideas—it’s actually a simplification, providing brands with more value for less effort.

What’s more, embracing agile is how we move toward the final step of the context marketing revolution: becoming a contextual organization. The transition to a contextual organization requires more than just adopting the context framework. It requires a reconfiguring of the business model and marketing’s role in it, new executive leadership, and a new way of reporting on marketing’s value. These changes are not small, which is why understanding how to test and prove results via the agile method is critical. Now that you have a solid understanding of why we need context and how to execute it, let’s take the final step and look at how high-performing marketing organizations realign their businesses around a new idea of marketing—the context marketing model—and complete their transition into becoming modern brands.