Everyone tells marketers they need to add AI to their crazy tech stacks. But fools rush in, says Jess Simpson, chief strategy and analytics officer at Acxiom. To have a chance of making AI or any technology investment a success, you first need to get your data management layer right.
Cannes is over, and after countless speeches, round tables and conversations around the Rosé-cooler, the results are in: Everyone says marketers should use artificial intelligence (AI)But no one knows exactly how to use it – or how to prepare for it. And I think this gap will be one of the most dangerous traps for brands in the coming year.
Don’t get me wrong, AI is going to do amazing things for your marketing. But instead of jumping straight into the latest tech and trendy tools, you need to get serious about your data and identity strategy first.
Because here’s what no AI guru says out loud: A lot of madtech is commoditized, whether it’s AI-based or not. What’s not commoditized is your relationships with your customers and the insights you can build about them as individuals.
You can have ready-made multichannel platforms, with AI to power your segmentation and creative strategies, but that won’t go very far if you haven’t done the work to understand your customers.
So, let me take you from the glitz of Cannes Lions madtech innovation to the unglamorous but absolutely essential data practices you need to master if you want to use AI.
Consolidate your databases
When I talk about database, I’m talking about all the ways your brand collects, manages, and uses customer information – and how you connect it to the real people behind the data, to engage them intelligently and ethically.
A database helps you address customers’ desire to see deeper value. Maybe you’re trying to reach parents, for example. Don’t just offer them a coupon. By better understanding their top stressors, concerns, and interests around kids and parenting, you can help them understand how your brand will help them navigate parenting. There’s a huge difference between a customer thinking they can save a few bucks on diapers and believing your brand can help them raise their child.
It’s a complex marketing plan, as unique as a fingerprint, but you need to start building it today or you’ll have a lot of trouble down the road.
Here are four steps to help you get started.
1. Optimize your data collection
You’d be surprised how many brands don’t know exactly what customer data they’re collecting, where they’re collecting it, and how. So start by auditing and updating those practices. Work backward from what you want to understand about your customers if you know. (It’s okay if you don’t know what you need to understand about your customers yet, as long as you understand what that means for your data collection strategy.)
Are you building consent into your data strategy? Are you giving customers access to the data you have about them through a preference center so they can review everything, update or delete information, or make other requests?
By giving customers control over their data, you can learn more about them. And an open, transparent interaction like this can contribute to a positive customer experience. Remember, data collection isn’t just a cleaning operation around the experience: it’s part of the experience. And it’s only going to become more important in a cookieless world.
It’s important to emphasize the importance of a data minimization strategy that focuses only on what you actually need to create truly meaningful customer experiences, without having to always rely on personally identifiable information. This makes all of your customer interactions more relevant, whether you’re talking to existing customers when they visit your website or to an unknown lookalike audience in a paid advertising environment. You’re much more likely to convert from these deeper connections than you are with more superficial personalization at scale. (And it will help you gain trust with your customers that their data is being used ethically.)
2. Don’t neglect data organization
The brands I work with often have data in 50-100 locations – and that’s just the smallest. With large global consumer goods groups, the numbers can be dizzying. So just consolidating all of that data to get a big picture of your data set becomes a huge task – let alone processing that data into audience segments.
This is fundamental work that can be accelerated through intelligent automation. The problem is that the way this type of automation is evolving (think data science and AI-based classification) means that it’s very quickly becoming something that mere marketers will struggle to manage. So you need a team of experts.
3. Find the right talent
The exact mix of data and analytics skills you need on your team will depend on the sophistication and scope of your marketing plan, but a strong technical leader will always be a critical choice. They can help break down barriers between leadership, product, marketing, and analytics teams.
With deep experience in the highly complex madtech landscape, they can help you guide your technology investments to meet your marketing needs. Importantly, they can also match those investments with the internal talent to manage them and the partners who can help you succeed, identifying gaps to fill along the way by hiring or co-creating solutions with partners.
4. Leverage data on priority use cases
You’re finally ready to put your data to work. And if you’ve followed the steps I’ve outlined so far, you can be confident that you’ve established a solid data management foundation.
Start by identifying use cases that you know will deliver real value, and be deliberate about that value up front, as well as the timelines in which you want to see it delivered.
The key here is to set expectations based on your current capabilities. A high-level sales leader may expect great things from AI, but if you’re not aligned with the expected value relative to the resources you have in terms of people, processes, and technology, you’re going to find yourself facing problems down the road.
Do the hard work now, so everything will be easier later
I warned you it wasn’t glamorous. But without a solid data and identity strategy, none of the amazing results promised by AI-powered madtech innovations are possible. I’m not here to spoil the party, just to remind you to work hard now, so that everything will be easier later.
Maybe it’s because I had Cannes in mind, but for me, the discussion of data strategy reminds me of moving to a new country. It’s really exciting to think about the beauty, the landscape, and where you’re going to live. The hardest part is learning the language. It’s not the fun, Instagrammable part, but it’s the part that will determine whether your move is a long-term success.
Learn Acxiom Data Management Solutions can help you lay the foundation for AI and other madtech innovations, so you can acquire, grow and retain the customer relationships that matter most.
