The inbox laboratory: How email can thrive on empirical experimentation
Columnist Steve Dille explains how personal data is being mined in adventurous new ways — and how email marketers can seize the advantage this practice offers.
At first glance, Facebook’s new reaction emoji might look just whimsical or downright silly. It seems like a step too far to ask users to go beyond tagging “Likes” into rendering judgments about whether they love a post, find it funny or are actually angered by it.
But there’s genius behind how Facebook is rolling out its new reactions emoji feature to subscribers. Facebook is famous — or infamous — for going to extraordinary lengths to capture finely grained user interaction data as a means for fine-tuning the algorithm that’s shaping each user’s feed.
Digging deeper for data
The new emoji are more than an enhancement that provides users with a way to post more nuanced, emotional responses: They’re one more tool that Facebook has added to its collection for optimizing the user experience by gathering data on how people engage and respond to content.
Think about it: Atop the existing layers of personal data — about buying preferences, political affiliations or other segmentation metrics — Facebook is now scrutinizing the emotional constructs that influence user behavior, both personally and en masse.
Yet the bigger picture isn’t about Facebook, specifically, or how it’s deploying a particular strategy. It’s about how every marketer needs to be mindful of the fast-tracked evolution toward empirical models and new depths of personalization being driven by Big Data.
Email marketers in particular need to commit to exploiting those models themselves by building a real-time A/B test regimen, even if they’re B2B enterprises.
The reactivity revolution
Why does A/B testing need to happen in real time? For starters, the technologies we’ve created have given us the ability to gather and process information at an incredibly quick clip.
A company’s ability to make that data commonly available to all stakeholders, and giving them the means to take action on it immediately, confer a competitive advantage when it comes to Big Data. Gathering information, no matter in what volume or quality, is worthless unless you’re making it actionable.
And that means a marketer has to gather and analyze user data fast — in real time, because that’s what consumers are attuned to in an age when buying a new hat, a 60-inch HDTV or a new car can happen with a few finger-flicks on a mobile screen.
An ability to react quickly and with personalized appeal equals consumer engagement. A recent Gallup “The State of the American Consumer” report pointed out, “Fully engaged customers are more loyal and profitable. A fully engaged customer represents a 23 percent premium in terms of share of wallet, profitability, revenue, and relationship growth.”
In an era of immediacy, modern marketing is about real-time reactivity to each individual user and fresh tools that are fashioned to capture, analyze and leverage data at the 1:1 level for each and every consumer.
A/B testing, now a real-time tool
Trying to codify personality types and messaging strategies to appeal to people on a more personal level is a concept as old as the Enneagram. And A/B testing is obviously part of email marketing’s DNA, rooted in email’s origins in direct marketing.
In time, testing grew beyond the simplistic models dictated by the early constraints of time, budget and a marketer’s ability to process data. A/B/C/D/E testing of an entire alphabet soup of variables made it possible to slice and dice audience responses in exciting new ways.
The addition of real-time analytics is what’s driven A/B testing to a new plateau of sophistication. Rather than testing a brace of static alternatives, marketers can now take a continuous approach to testing, building and refining at each step along the way.
And with today’s analytics platforms, those progressions can take place instantly.
Take Ebates and Pinterest, for example: Using multiple data points garnered from email and their websites, they’re able to test and tune messages at the per-user level, rather than by doing broadly segmented tests. It’s A/B/C/D/E raised to a nearly infinite power, and each test informs the next, always progressing toward a more finely honed and entirely personalized user experience.
Through that experience and the understanding it gives them about each of their customers, marketers can segue from being reactive to predictive.
The tools aren’t next gen; they’re now
Empirical analytics tools for email empower you, as a marketer, to optimize deliverability while also tracking each individual’s engagement. Those tools are widely available — but for some enterprises, adoption can seem like a scary leap forward, since those firms are so invested in time-honored ways of dealing with customers and prospects.
Employing these platforms shouldn’t have to override the virtues of experience and any intuitive approaches you may value. Let’s face it, there’s no substitute for the ultimate 1:1 engagement with customers, the sales call or customer service session.
In fact, jettisoning hard-earned savvy about a category or audience might knock you back to square one.
Instead, that wealth of experience should be used as a launchpad for using email marketing to test out fresh ideas for engaging your target audience, whether existing customers or acquisition targets.
How to embrace empirical experimentation
Let’s look at two different approaches to deploying an empirical testing approach to email (and other integrated) marketing:
- Company A is an established marketer that decides on a baseline approach where their time-tested product mix, message strategy or other factors are locked in, and any A/B testing is strictly iterative and done sparingly.
- Company B is a startup or challenger brand that’s committed to rapidly testing multiple email marketing innovations, messages or hypotheses in parallel or rapid succession — the “fail fast, try again” approach.
Each has its merits — and demerits:
- The baseline-hugging approach of Company A keeps them anchored in the category and in their customers’ perceptions, but it can keep them from exploring options and innovations their competitors might use to leapfrog them and from staying in step with consumers’ ever-evolving tastes and interests.
- Company B might be that very leaping frog, but there are always risks of exhausting resources before gaining traction, fragmenting whatever brand identity they own or only realizing marginal response from narrow segments.
The best route for most marketers probably lies in hybridizing the two philosophies.
Let’s go back to Facebook and its “reaction” emojis: Facebook overlaid a proven baseline data-sourcing engine with a tactic of adding refined levels of insight about each user. But before rolling it out across their entire platform, they tested and retested it relentlessly against various subsets.
So our Company A might instigate a Company B-style campaign of judiciously yet continually testing different email messaging, design and interaction options against varied lists or segments.
The inbox laboratory
Real-time analytics turn the inbox into an unmatched testing ground for extracting deeper, smarter insights about every consumer.
The instantaneous 1:1 relationship more and more consumers demand from marketers means we, as marketers, have got to engage in a meaningful and constant conversation with each member of our audience.
And just as in any conversation, we only get to know them by asking questions, trying different inflections and gestures and adjusting what we do and say so we’re both comfortable — and that ends up making the relationship behind the dialogue deeper and more enduring.
Opinions expressed in this article are those of the guest author and not necessarily MarTech. Staff authors are listed here.