Easing Friction Around AI-Driven Personalization: Top Retail Marketers Share Their Advice

If you’re a retail marketer, it’s hard to go five minutes without hearing about personalization. Everyone in your organization likely understands the importance and value of personalization in theory, but that understanding doesn’t always translate into practice.

Quite the opposite, many marketing teams face serious friction when it comes to implementing those sought-after personalization strategies. So what gives?

We sat down with senior level marketers from retailers like Bass Pro Shops, City Furniture, Hammacher Schlemmer, Lucky Brand, Ulta Beauty and Quiksilver to learn more about what they experience on a regular basis and how they’re learning to work through some of this friction to get personalization off the ground at their companies.

The Biggest Source of Friction Around Personalization? Giving Up Control

Whether it’s pushback from merchandising, creative, IT or the executive level, friction around new, technology-driven personalization can — and will — come from everywhere in today’s retail organizations.

At first glance, this friction can be confusing. If everyone agrees on the need for a better, more personalized customer experience, why do people start objecting to delivering on that goal?

Largely because it means giving up control in certain situations.

The best way to deliver personalization at scale is to trust AI to make decisions. That means introducing technology that can collect, combine and act on data largely on its own based on parameters set by your team. This approach offers benefits like expanded product recommendations, better timed communications for each individual and more dynamic offers, all at scale without any manual workflows for your marketing team. 

But it also requires significant trust in the system and the data to communicate with shoppers “correctly” (the way your brand wants). And it means that some things might look different: What each shopper sees might not be exactly what a person on your team would have put together, but if you trust the data to get things right for each shopper, then that’s actually a positive.

For many teams, learning to trust the system will mean a departure from the way things have always been done — and that change will not be easy. Merchandising teams will need to get comfortable not having complete control over what products get paired together, and instead let the AI decide which combination of products each customer actually wants to see. Meanwhile, creative teams will need to adjust their expectations of how each marketing communication will look based on the brand images that get paired together. And that’s not to say the end result will be off brand, it just means it might result in a particular set of images getting featured that the team might not have considered. Overall, it’s important to note that the AI is making these decisions about what to feature based on millions of data points that go well beyond what a human could reasonably assess.

How Can You Ease the Friction? Start with Transparency

All of the marketers with whom we spoke have experienced this friction around personalization initiatives in some way or another. Fortunately, many have made significant inroads toward easing that friction. And according to them, doing so starts with transparency.

Specifically, this transparency should include taking the time to understand goals and hesitations across departments and then making key information, like the data that drives personalization, easily accessible for everyone.

Let’s break down exactly what this entails:

Have an open discussion

First and foremost, it pays to get everything out in the open. Your marketing team should sit down with your merchandising team, your creative team, your IT team and so on to talk through everything. During this discussion, make sure everyone is clear about company-wide objectives and how each team’s goals roll up to that. Then you can start finding ways to support each other, particularly through the use of any data-driven personalization solutions. 

For example, if the merchandising team wants to make sure they can sell more of any products with a surplus, how can the marketing team use technology to surface those products to the right customers? Overall, this type of conversation will help break down any perceived barriers that exist between teams to get everyone on the same page and start identifying ways you can work together. The goal should be to bring these stakeholders into the build-out process so that you can collaborate to build a program that speaks to each one’s goals.

Educate others on what the data offers and what you can control

Next, it’s important to educate those teams on everything the data can do for them. Along the way, it’s also important to make clear what you can control, because there are many areas where your team can give the AI constraints that will help achieve key goals. 

For instance, if your brand is trying to pull away from discounts, you might limit dynamic product recommendations so that customers can only see products that have been marked down once (versus two or three times) and/or limit it so that only customers with a discount affinity can see those markdowns. Or if your brand has limited inventory of certain products, you might limit the recommendations to only show products with a certain amount of inventory.

Give access where it makes sense

One of the best ways to build trust in the data and help other teams understand what they can and can not control is to give them access to the technology where it makes sense. This access will help them get familiar with the solution, which can make many conversations easier, and it might also help them find ways they can use the data in their own day-to-day.

Some of the marketers with whom we spoke shared success stories about giving merchandising teams access to product-centric data like which products are “hidden gems” (those that don’t get surfaced often but have a high conversion rate when they do) and how likely people who buy in “category X” are to also buy from “category Y.” They had similar results giving creative teams access to view email templates and proofs, which helped ease the feeling that they might lose control by not being able to look at every single email that goes out since each one would be personalized to the recipient.

While giving this type of access might not be possible in every situation, instances where you can grant that access can go a long way toward easing any friction.

Share results often

Finally, the best way to strengthen and maintain trust in the data-driven approach is to share results often. After all, it’s hard to argue with an approach that keeps delivering better and better results over time.

Not only does sharing this data prove the value of the approach, but it can also bridge gaps between teams by helping others find ways to use the data in their own day-to-day. Once again, this might mean sharing insights with merchandising that helps them make better decisions in picking products or sharing engagement data on different brand features with the creative team.

Getting Personalization Off the Ground Requires True Collaboration

At the end of the day, the key to easing friction and getting personalization off the ground lies in large part on transparency and collaboration. The more you can bring cross-functional stakeholders into the program, whether that’s setting goals, sharing data or anything else, the more they’ll understand how it works. It also means you’ll be giving them a lot of data that can help them make smarter decisions. That’s not to say the data is replacing what they do, it just means that pairing those insights with their years of experience can deliver even better results.

Interested in learning more about what it takes to deliver personalized experiences that exceed the expectations of today’s shoppers and drive growth for your brand? Click here to watch our broadcast featuring Ascena Retail Group, Beyond Transactional Commerce: Driving Profitability in a Digitally-Dominant World.