Personalization: What You Need to Know Right Now (Part 2)

When it comes to personalization, how important is data collection? What’s the best advice for companies that want to step up their game? And how might personalization technology be improved in the future? In this two-part series (read part one), we ask a senior member of our data science team to explore the trends, challenges, and possibilities of personalization.

How important is data collection when it comes to personalization?  Where do brands tend to struggle?

Data is the foundation of any personalization campaign.

In the digital world, most tools have been designed for the savvy marketer rather than for a data scientist and his or her modeling efforts. So the first thing you need to implement is some type of data warehouse that stores user-level data and joins profiles by a common visitor identifier. This drastically differs from the aggregate information that most digital analytics tools will surface.

Concurrent with this data warehouse effort is mapping all relevant dimensions of the data that need to be collected. For example, you will want to:

• First, obtain digital behavior data, including all the actions a particular session included
• Then combine this behavior data with any available offline data, including, for example, records of which items have been purchased over the phone or in the store
• Finally, augment this information with any relevant and appropriately obtained third-party data.

If the right data is not available, then the personalization campaigns won’t be relevant or timely.

Though rarely discussed, the transparency in which the organization collects data and the disclosure of what they intend to do with the data are equally important. Clear policies and opt-outs should always be part of data collection and utilization.

What data has been most valuable to personalize against (e.g., past transactional history, current behavior, third-party data, first-party data, social data, other)?

All data sources can be valuable—it just depends on the specific needs of the campaign. Statistically modeling a desired action, such as a purchase, will uncover what particular variables or data sources are relevant and important.

That said, the recency and frequency of behavioral activity are commonly the most predictive indicators of a future action.

What advice would you have for a brand looking to embark on the personalization journey?

Achieving alignment and senior-level support for the personalization efforts is most important.

Educating the leadership about what is technically possible and how that supports the strategic direction of the business will help to secure the initial funding for the personalization campaign. It will also sustain the program in the face of potential short-term setbacks.

Once the program is off the ground and running, focus on small but data-driven ideas that have the potential for a large impact.

What is typically the biggest challenge to executing a personalization program?

Personalization efforts require a big investment in time and resources, often across many parts of an organization.

Most people have managerial or self-appointed mandates to prove a certain return over the next 60, 90, or 120 days. But the ROI of personalization efforts may not be able to be measured in a short, concrete period of time.

Finding the right senior leader to sponsor the program, provide resources, and drive adoption is usually difficult, because the real value may not be realized right away.

Where do you think the industry is going when it comes to personalization and customer journeys?

I think the industry will continue to evolve. I expect to see an increase in the sophistication of the analysis that serves as the backbone of personalization efforts.

I also expect to see continued improvements in the technology, making it more seamless to both perform the analysis and implement the campaigns.

Right now, tools have just begun to emerge that claim to offer auto-personalization. A selection of algorithms are available to model customer behavior and assign experiences—all within the testing tool. The potential downside is that customer are put into groups that may not be consistent with their preferences and goals. That’s why it’s truly beneficial to have a talented analyst compare the experience displayed with the customers’ purchasing behavior. The bottom line is that an analyst needs to make sure the correct experiences are being displayed to the correct customers.

Ultimately, the industry will need to strike a balance between auto-personalization and statistical rigor. In doing so, those without a data science background can more easily execute campaigns.

However, those who do have a high degree of statistical aptitude will always need to approve of the inner working of the algorithm and how it relates to the specific data being collected.

What has been the most interesting thing you have learned from helping customers with their personalization programs?

Again, finding the right senior leader to sponsor the program, provide resources, and drive adoption is usually difficult because the real value may not be realized during the expected tenure of any of the participating individuals.

Reid Bryant is a data scientist at Brooks Bell. He uses advanced analytics and applied statistics to create data models, refine methodology, and generate deep insights from test results. Reid holds a Master of Science in analytics from the Institute for Advanced Analytics at North Carolina State University.

—

Brooks Bell helps top brands profit from A/B testing, through end-to-end testing, personalization, and optimization services. We work with clients to effectively leverage data, creating a better understanding of customer segments and leading to more relevant digital customer experiences while maximizing ROI for optimization programs. Find out more about our services.