For Amy Lew, a senior user experience (UX) designer at Adobe, it’s all about marketers—what they want, what they need, and, more importantly, how they think about everything from making data-driven decisions to gaining organizational buy-in to growing relevance at scale. It’s these considerations that drive Amy’s thinking to help the marketer take more risks and trust in a hands-off approach to personalization that results in quick wins with much less effort. I recently got a chance to catch up with Amy at the Thirsty Bear Brewing Co. in San Francisco. In addition to some great beer tasting, we had a fun chat about personalization and the role of marketer trust.

The term “personalization” has been thrown around a lot recently—but the concept is far from new, and the technology itself goes back at least 15 years. I’m old enough to remember all the talk about “black box” personalization, but I’ve also seen awesome examples of it really working—higher conversion rates, great cross-sell results. What will it take to get marketers more comfortable with personalization technology? What needs to change?

Marketers have always run to their leadership saying, “This works!” when it comes to personalization—because, yes, personalization gets results. And that was good enough for a long time. But fast forward to today and everyone—including that same management—is more knowledgeable. Sure, the results are great, but now stakeholders say, “But why?” Why are we getting those results and, more importantly, how can we do more of it? They don’t want statistical probabilities and algorithms, of course, but they want to know that next level of granularity in terms of reporting. Adobe Target is meant to answer that, and that’s what I’m focusing my energy on—not just answering but elevating that conversation.

In that vein, I always hear, “Why use automated personalization if I can use A/B testing and my own predictions and knowledge to do the same thing?” It’s important to take a step back and look at the current landscape and ask yourself if that level of knowledge and actionable data is enough. Right now, I don’t think it is.

That’s where my role within Adobe Target picks up. When I think about Target, I’m thinking about how I can help you experiment more, make better decisions or unearth metrics so you can back into and leverage for more internal support and greater resources. Because with automated personalization there’s much, much more in terms of complexity than with simple A/B testing or the notion that “some personalization” is better than none. Target can take it further and, instead of creating new offers from scratch or from the user interface, we’re basically pairing audiences with experiences or offers based on algorithms and our own proprietary logic. It’s an incredible service to the marketer and opens up limitless possibilities. But of course we’re encouraging marketers to trust our personalization engine and understand that data-driven, self-learning, and algorithmic approaches to targeting content can offer quick wins with less effort. And this means making sure we give them what they need in the product to make them feel comfortable enough to relinquish some of their control.

So how, tactically, is this playing out both within Adobe Target and in the greater personalization conversation?

We’re out there talking to marketers every day—and we’re marketers ourselves. We know you need to prove true value and ROI to keep integrating a solution like Target. And it’s in our best interest, of course, for you and the decision makers in your company to say, “Target makes us incremental revenue or drives greater engagement, let’s keep using it.” So we provide everything—every metric, every data point, every next step—we can to make sure you can say that, definitively. That starts by providing strong data to support that the content, offers, products, and information we “pair” with a specific audience compels them to act, through higher conversions, greater purchasing, or whatever other KPIs we’re up against.

When we look at those numbers, we always funnel them through three tiers. First, is automated personalization performing better than if you used nothing at all? We compare the personalized set to a random set—a person who’s coming to the site and getting some combination of anything that couldbe valid, versus what our algorithm believes is more likely to get him to convert—so that we can, essentially, say that our algorithms are performing. We’re making you money. We can say, without a doubt, that automated personalization works in this scenario. If we can’t say that, we’ve failed.

The next level looks at personalization beyond just “did it work?” Here we dig in to what worked well and what didn’t. The tough part here is that different offers or content pieces are going to different audiences. So we have to look and say, “Did one set of offers do much better than others? Why?” Maybe there’s someone really fantastic and incredibly creative putting out spot-on offers that convert well, or maybe the blue just outperforms the pink, for some reason. Maybe skiing just always outperforms snowboarding. By making those observations we can start to determine smart, actionable next steps. If skiing is drawing significantly more engagement and conversion than snowboarding, maybe skiing deserves that hero spot—or maybe snowboarding shouldn’t be on the site at all, if it’s performing that badly.

The last stage is really the holy grail of Target. It’s the offer detail report. We’ve now homed in on a specific experience or offer, and can drill down on who the specific audience is for that content. In this stage we’ve changed how much predictive power an audience or variable has. Through capabilities like iterative personalization, we can get a first glimpse into what kind of offer appeals to a specific audience segment, and vice versa—what kind of offer will this particular segment respond to? There’s still work to be done, but eventually we’ll be able to have complete automated personalization, and know definitively what offer or experience goes with what audience predictively. For now, though, we can say “these are predictive—and that works.” We have an offer, we have associated variables with it. And when we can look at those pieces and say “this is worth this much,” you’re lending incredible value to an organization.

One thing we’ve been talking a lot about is trust and the notion of digital distress. Marketers don’t trust the data, the sources, and even themselves and, as a result, experience varying levels of distress that can lead to stagnancy. How are you thinking about this when it comes to designing a marketer experience?

I think about this a lot. At the end of the day, Adobe Target and its relative effectiveness come down to that trust. Marketers need to trust the reporting, data, and segmentations enough, and need to be open to reinventing and reimagining their roles. Target is designed to be completely in sync with those marketers. It’s about making their jobs easier in every sense. More to the point, marketers need to get comfortable with a more hands-off approach such as automated personalization. In my role, I’ve designed Adobe Target to alleviate some of the most common causes of anxiety for the marketer who wants to experiment more aggressively. It’s simply more foolproof—and eliminating that self-doubt is a big step on the road to breaking down some of marketers’ major trust issues.

So what’s next? What are your priorities and how are we evolving Target to address them?

Going forward, it’s all about doing even more with the data. Marketers need to think of data as part of a “virtuous cycle,” but must avoid getting overwhelmed by it. This means just-in-time data, small data not big data—and, yes, continually learning to trust more in the power of algorithms to consume data and make decisions, predictions.

We see Target taking advantage of on-the-fly audiences, and this is where real-time reporting and visualizations are going to be critical for the marketer. To be able to infer an audience and say, “This surfaced from my personalization efforts, and it has a full life cycle.” Pretty cool. You can apply other characteristics to it from your master marketing profile, do some additional A/B testing on it, and use an entire suite of digital marketing products to get the right stuff to the right people at the right time.

From there I want to see this go a step further. In an ideal world I would be able to say, “This offer matches with this very specific audience, and I’m always going to be able to feed this to these people and win.” Marketers want to be guided through the optimization process, but not give up control entirely. Let the machine do the personalization heavy lifting, but give the marketer the insights and tools to craft a strategy that gives them win after win. My job is to be there alongside them. Well, at least that’s what I’ll be imagining as I work on evolving the Target UX.

Amy would love to hear from you. Join the conversation on Twitter with @amy_lew