Sign up for our newsletter and get the latest big data news and analysis.

Are Machines the Answer to More Personal Marketing?

Personalization is not a new concept; US marketers have been talking about it for years. But it feels as though whenever we start to get to grips with what it means and how to do it, the goalposts change as consumers demands evolve. Let’s face it, the days of being impressed by Amazon recommending a book based on the last book you read have long gone.  Consumers now expect their customer experience to be relevant specifically to them, and with pure-play digital giants rising to the challenge rapidly, businesses are under increasing pressure to evolve their personalization strategies.

Big-picture questions haunt us as we formulate our strategies, ones that point to a basic issue of what we’re even talking about when we’re talking personalization. In a 2019 Econsultancy report, 53% of marketers said that “data-driven marketing that focuses on the individual” was the most exciting opportunity in 2019, in comparison to just 7% of organizations saying personalization was top priority in the 2018 report. Strangely, back in 2017, 73% of Forbes 500 executives said they “must” deliver personalization to be successful. So, is it a priority or not?  I suspect that it is, but we’re just using different terms for the same thing. Whether we’re talking segmentation, automation, data-driven, dynamic content or AI, we’re all driving towards the same thing: delivering relevant, personalized experiences to improve CX and, ultimately, conversion.

One thing’s for certain, the appetite for personalization is there but most companies have yet to achieve it. Common challenges preventing the application of personalization strategies have, historically, included a lack of resources and expertise to deliver the amount of content and data analysis needed, lack of process, and legacy technology that wasn’t up to the challenge. These challenges still exist; however, the skills, tools, and techniques are now widely accessible. If you have the ambition and buy-in to invest, the technology is available to enable success (think Microsoft Azure and Sitecore, for example).

Demystifying the Buzzwords

There’s no point in comparing your business to Netflix or Amazon or Facebook. Because, well, they’re the white rhinos of the digital landscape. But there are achievable goals that we can all aim for. This starts by demystifying some of the buzzwords muttered in boardrooms across the nation.

Let’s start with everyone’s favorite: Artificial Intelligence. The term itself conjures up Philip K Dick-esque robots. It’s far more digestible to think of it simply as Machine Learning (ML). Machine Learning has the ability to automate analysis and detect patterns of data at a rate that would be impossible for humans to achieve. It can take data segmentation beyond simple keyword clusters, and opens up the opportunity to glean information from new data sources, such as audio, image, and video.

Team the opportunity of ML data pattern analysis with, say, audio sentiment analysis and natural language processing, and you’ve got the intel you need for powerful personalization. For example, if you’re a utility company, you could use ML for audio sentiment analysis across your call center (i.e. find all the ‘angry’ calls), and layer this with user search and location data. By doing so, you could automate a process to serve relevant content, specific to that user’s exact complaint, next time they call in to report an outage. Or they could receive an automated message detailing their issue and how you’re going to solve it. You could address the individual concerns of that one customer in record time, without them having to repeat calls or explain the issue again. How’s that for a great customer experience? Happy (or, at least, appeased) customer. Happy CEO.

Hyper-targeted experiences are made possible with dynamic content, which is the crux of personalization success. It can take you beyond trigger-based ‘they bought a screwdriver last time so show them pictures of screws on the homepage’, to a whole new level of real-time relevance. Netflix serves dynamic content based exclusively on previous viewing and preference analysis, built on extensive tagging. Very overlays usual data points such as purchase history and demographics, with real-time weather analysis, to push personalized product content on-site. Even Apple’s newly redesigned TV app has new curation features that can accurately determine what type of content you might like. And what powers this dynamic content? Machine Learning.

Omnichannel is (still) a Challenge

Understandably, personalization becomes trickier if you have an offline presence. Traditional brick & mortar retailers will be all too familiar with the omnichannel challenge. Online retailers know that consumers will interact with their brands across multiple channels, but when you throw physical stores into the mix, it makes delivering the right message, to the right user, at the right time a little trickier.

How can organizations personalize their site to promote after sales activity, if they don’t even know that a user has purchased in store? The in-store POS system would need to be able to identify the customer’s profile and map previous online research behavior, combining the records to create a fully informed personal profile. It sounds simple enough and, in theory, it is. If you can collect a unique ID at the point of sale (e.g. an email address), you can map it to their online profile. But there are multiple tech systems that need to be integrated with each other to deliver this information from the POS system to the CRM to the DXP or CMS, in order to trigger the right personalized content. It’s totally doable, but entirely dependent on investing in the right technology stack, having the correct skills to be able to connect the data and, crucially, getting your hands on that unique identifier.

I think we’d all be in agreement if we said that we want to implement this level of personalization to improve our customers’ experience with our brand. But what if said customers don’t want to explicitly share their information? Implicitly, we share our data all the time, but increasingly consumers are wary. In the EU, GDPR has been implemented to help protect consumers from unscrupulous use of their data but, let’s face it, none of us are still entirely sure what the real implications are, especially outside of the Eurozone. Does this mean that we now need to ask users’ permission to use their data online when they’ve purchased in store? Should stores guide consumers through the online experience in-store, therefore keeping all transactions in the digital realm where digital rules apply?

If we’re being truthful, we sometimes think we know how to tackle the challenge, without really understanding the wishes of our customers. In a recent Dept survey, we asked consumers whether they liked personalization or not. Over 30% of respondents said that they didn’t but, interestingly, the brands they selected as their favourite retailers were the ones doing personalization the best. So, does this mean that we need to advance our strategies, technology and execution to deliver personalization in a way that consumers don’t even realize they’re being personalized to (think Netflix or TikTok)? Or do we need to make it an active choice for consumers (think Thread)?

About the Author

Mellissa Flowerdew-Clarke is Marketing Director UK & US for Dept. Mellissa has spent the past fifteen years helping businesses to realize the potential of effective, multi-channel marketing strategies. Her background in retail analysis and expertise in written communication has helped both B2C and B2B brands understand their target customers and how to best to communicate with them, whether it be selling chocolate, cars, software, services or anything in-between. As Marketing Director UK & US at Dept, she is responsible for driving the agency’s marketing strategy in these regions, connecting and converting new clients and employees to help fuel the agency’s international growth. 

Sign up for the free insideBIGDATA newsletter.

Leave a Comment

*

Resource Links: