The benefits of personalization at scale require understanding what “triggers” are and how they really work.
Personalization can dramatically improve the effectiveness of marketing campaigns, but scaling it across an organization requires sustained focus and teamwork. Julien Boudet, a McKinsey partner in Seattle who advises consumer-focused companies on marketing, e-commerce, and operations, joins McKinsey’s Barr Seitz to discuss the hurdles companies face and the benefits they can reap when they personalize at scale. The interview has been edited for clarity.
The main challenges to scaling personalization
Doing personalization is not new. A lot of companies have tried it and believe in the value of personalization. What is often hard is doing it at scale and sustaining it within the organization.
Companies attempting to scale personalization face several challenges, one of the main ones being how to build an operating model to support it. Doing personalization is not a one-person job. It requires a cross-functional effort with people from marketing, engineering, analytics, finance, legal, and creative. You need to get all those people operating together.
Technology presents another challenge, since it’s vital for the ability to scale the impact on personalization. It’s a very fragmented, fast-moving space, however, and finding the right vendors, the right partners, and getting the most from their solutions is never easy.
Data analytics poses another hurdle. Many companies have a lot of rich data, but it’s often siloed across the organization. You need to figure out how to put all the data, including third-party data, in one place and create what we call a Consumer Data Platform (CDP). The goal is to create a unified view of the consumer across all the data you can access in the company.
What it takes to make personalization work
There are a few things we’ve seen successful companies do. One lesson is not to wait for perfection. At the end of the day there’s an enormous amount of data that can be pulled together, but you don’t need all of it to begin.
It’s also vital to remain focused on the problem you’re trying to solve. A lot of companies often start building infrastructure, trying to tackle too many issues and solve too many problems. You need to be very crisp about the consumer opportunities you want to pursue, the data you need, the people required, and the organization of your effort.
One strategy we’ve seen companies do really well is creating what we call a war room, or a personalization lab, to help disparate parts of the company collaborate more efficiently. For example, marketers working on new initiatives will require support from engineers to make it happen. Changing landing pages or some other functionality requires analytics to know if the campaign is actually working.
A war room allows creative people to develop new campaigns much faster, and we’ve found that applying these agile marketing techniques to personalization drives significant impact—not only on the output, where we have seen a significant lift in campaigns, but in the input too, which is the number of tests.
In some cases we’ve seen companies increase the number of tests from 2 to 5 to 20 to 30. And that’s a direct result of all those people working together on a daily basis toward the same goal.
How a company can unlock value from customer data
Get a clear view of your consumers’ pathways, and determine the steps they took along the journey.
Once you have a clear picture of your consumer, you need to make it accessible. One challenge companies often struggle with is that data ends up being virtually locked up within one team, which then accesses that information and provides it to the rest of the organization. This makes it really hard to scale, because you’re relying on a few people and a lot of manual queries to create reports for the whole organization.
The more you think about your data, the more you should think, “How do I make that data accessible to the organization through tools?” This idea of making data almost a service is critical to scaling personalization. Then you need to figure out which consumer actions you want to stimulate. Having a clear goal allows you to look at the data and understand where the opportunities to accomplish it are.
A good example how to use analytics to improve a company’s personalization
A large retailer has done something pretty interesting when it comes to campaign reporting. A lot of companies struggle to actually get real-time information about the campaigns they launch. Typically, a marketer creates a campaign and passes it onto marketing operations for the launch, but then it takes a few days to actually get the report. But this retailer provided marketers with access to a tool that gave them information in real time about their campaigns, with a segmented view of not only how the campaign worked in terms of open rate, click-through, or conversion, but how it performed across different segments they targeted.
This is important because they now had access in real time to the campaigns they’d just launched, which can help to make changes as well as to inform the design of subsequent campaigns. It’s injecting the concept of test-and-learn throughout the organization.
The benefits of this change were twofold. First, the time spent on campaign design became shorter, because marketers were getting a clearer view of what was working and what wasn’t much faster than before. This helped them focus on what they wanted to test next.
Second, from an impact standpoint, the actual conversion of email went up by a couple of percentage points for each campaign across the whole organization. Across the whole organization those improvements were significant.
Discovering the “triggers” that help with personalization
Let’s say a consumer visits your site and browses for something without buying it. That means they’re in the market for a particular product. That consumer signal is what we call a trigger, and it leads to the delivery of a certain message.
It’s all about building this library of triggers that are readily available in your system, ready to be activated in real time. And then creating the engine, what we call the brain, that will decide the right trigger for the right consumer at the right time. This really helps you scale personalization.
We’ve found that timing matters when it comes to triggers. Getting the message out within 24 hours is the most important thing to get right. After 24 hours, people may already have purchased it elsewhere.
Companies that do this really well manage to make 20 to 45 percent of their campaigns trigger campaigns. That matters because a small percentage of those trigger campaigns drive the most value from your program. Five to 10 percent of your trigger campaign might represent 20 to 30 percent of the value of your email programming.
When it comes to creating triggers that are specific to your consumer base, we often find it helpful to ask, “What are the key behaviors along the journey that help a consumer move from a first transaction with you to becoming one of your most valuable consumers?”