Ben Gaines, Adobe - Customer Intelligence Rockstar Tips and Tricks

November 6, 2017

Customer Intelligence Rockstar Tips and Tricks

Slide 1:

I’m excited to be here with you all at the ObservePoint Analytics Summit. I’m excited to see some of the content that’s going to be share later on today, and to give you some tips and tricks into how Adobe Analytics customers are using our products to get maximum value out of it and to move toward this vision of customer intelligence that I think we’re all striving toward. I guess I don’t need to reintroduce myself, but I’m Ben Gaines. I’ve been a product manager at Adobe for about seven years.

Slide 2:

In that time, these are the countries I’ve visited. We travel the world meeting with many of you. In fact, I’ve probably spoken with many of you all over the world at various times, and it’s great. We get tremendous insights into the problems you’re facing and how we should be developing our product to solve those problems. But it’s also exhausting. As you can see, it’s been a wild seven years. It’s been about 700,000 miles flown during that time, and that’s not frequent flyer miles, which get doubled and tripled, that’s actual nautical miles flown.

Wherever I go in the world, I know there’s going to be a Marriott bed waiting for me. I’m a loyal Marriott customer, and acquirer of points, and hopefully will be cashing in those point son future vacations. As I think about a brand that’s really embraced the idea of experience, and improving their customer experience through data, and really understanding their customers, Marriott is full of that.

Slide 3:

This is the Renaissance hotel, which is a Marriott hotel in Hong Kong. I stayed there a few weeks ago when I was out there visiting some customers and doing an event out there. I had been on the road for a while and was really tired, and was reminded when I got to Hong Kong of this experience that I think Marriott has really, maybe not perfected, but they’re on their way. And I want to take you through how they think about data and the experience as a backdrop for some of the tips we’re going to share today.

If you’re not a Marriott customer, this will be new to you. If you are, hopefully you have had an experience similar to mine. After you book at a Marriott hotel, of course you get your confirmation email, but one of the things I really like about this experience is, Marriott follows up with you about a week before you get there. You’ll get an email pre-welcoming you to wherever you’re going to be. This one is from a trip that I took to Stockholm. In that email, it will have some recommendations, things to do in town, places to eat. It prepares you for being in a new place, which can really disorient you, especially when I’m off and jumping from city to city, or even country to country. Having that guide to remind me and orient me can be really important and help me feel welcome.

The next step comes as they’re mobile app. I get a push notification inviting me to check-in the day before I get there. What I love about this part of the experience is I can make requests. If I needed some bottles of water or some extra pillows or if I going to need to check-in early or check out late, I can do that right here in the check-in process while on my own time. I don’t have to remember when I get there to ask for this or that. I can ask as I go and when I’m ready.

My favorite thing they’ve done as part of the customer experience is the push notification I get when my room is ready. The reason I love this is that I hate—after a long flight, even if it’s not an international fight and I’m just flying from Salt Lake City to New York—I hate getting to the hotel a little bit before check-in time and my room is not ready. Then I’m like, what do I do? Do I wait? Do I go do something?

This push notification lets me know I don’t have to go to the hotel until the room is ready. Until I get the push notification, I can hang out, I can eat, I can maybe get a jump on my sight-seeing if I feel like it. Very often, that push notification comes in before check-in time, before the official 4pm, or whatever it is. Once I get that push notification, it may be while I’m still in the airport, I can just proceed to the hotel with confidence and know that there’s going to be a room waiting for me.

Then, of course, nothing beats the welcoming. I’m welcomed as a known customer. Of course, I’ve signed up for the loyalty program, but that allows them to welcome me in a special way. I get a little gift with some bonus points when I check-in. it makes me feel that much more welcomed. This experience has turned me into a real loyal Marriott customer. And I know that wherever I go in the world, there’s going to be a Marriott waiting for me and this experience is going to be replicated in Hong Kong, in Stockholm we’ve seen parts of it, and certainly throughout other parts of the world as well.

I really love this experience. This experience did not happen by accident. If you look at things that Marriott has publicly said about the way they think about customer experience, it’s clear that they are taking customer data and customer intelligence really seriously. They think about putting an experience like this in front of a customer like me or like you.

Slide 4:

Andy Kauffman, who runs digital marketing and, said—this is what you see here, is the end of a quote, I want to read you the first part of the quote as well—he said, “think back years ago, before the digital age, people had to go into the hotel for their experience to start. Now those experiences start well before they step into our hotel. It may start when they interact with us on our website, it may start after they’ve made a booking and they check-in with their mobile app, or it may start when they are using some of our mobile guest services. All in all, our job now is to enable the journey for that guest to be seamless all the way through that experience.” And then he says, “Marrying the data we collect on social media…with the rest of our guest data to create a holistic engagement at any moment in time generates loyalty and revenue.”

In my experience, that’s certainly true. This is something that we’re all probably striving for. If you’re on this call, if you’re part of this summit, I’m guessing that you believe this. You believe that guest or customer data, used correctly, applied correctly to the customer experience, creates a seamless, holistic—as Andy Kauffman says—a holistic engagement at any moment in time does lead to loyalty and revenue. But I think the important thing I take from this is that Marriott is not a brand that was born in the digital age. This is a 90 plus year old company, so they’re legacy is not in digital. They’ve had to learn. And I know as I’ve talked to many of you and your peers all over the world, there’s this struggle to kind of bring your brands into the modern age by using data.

I share the Marriott example, one, because I think it’s a great example, but also to emphasize that this is something we can all do. It’s hard. It’s hard work, but there are things we can do to get started. The tips that I’m going to share in just a minute are all about getting started on this journey toward customer intelligence and specifically, toward the kind of customer intelligence and the kind of change in your organization that can help you get closer to this experience that we walked through for Marriott.

Slide 5:

I think we would probably all agree that an experience like the one that Marriott has put in front of customers is the kind of thing that differentiates. It’s not necessarily my stay at the Hong Kong Renaissance was better than it would have been at another, maybe higher end, hotel in Hong Kong, but that experience keeps me coming back over and over again. Experiences really do matter more than ever to customers, especially as a lot of product and services become a lot more commoditized.

I’m going to talk a lot about—hopefully, if I remember to—I’m going to try to connect a lot of what we share back to the customer experience because that really is what we want to improve with data. It’s not just data for data’s sake, but experiences are things that are going to allow your brand to differentiate.

Slide 6:

To that point, it’s really data that should be driving experience. And not just data, I guess, but customer intelligence. Taking that data and turning it into insight about your customers, about different segments of your customer base and how to reach them more effectively. I’ve seen throughout my experience with Marriott and other brand I’m loyal to, that they are starting to do a much better job recognizing me wherever I’m interacting with them and that makes a difference. They are becoming experience businesses.

Our CEO of Adobe, Shantanu Narayen, said in 2016, “Getting content to the right person, in the right place, at the right time takes data. We need insights, we need intelligence, which are the engines of these highly personalized, and relevant experiences. Robots will never do great marketing. Human intuition can never be replaced, but if we can harness that power of computing, man plus machine, to work faster and smarter, we think we can make big things happen.” And that’s what we are beginning to see.

As I talk to brands like Marriott and others who are starting on this journey, there are some clear themes that emerge and things they are trying to do. And I think I should say before we get into anything else, I don’t think anyone is quite there yet.

Slide 7:

These are the four areas we’re going to focus on that we believe are part of customer intelligence in 2018. If you want to start your journey toward customer intelligence, these are the places to start. Getting as much customer data as you can to create a 360-degree view. This idea of a citizen data scientist or learning tools that can help make sense of all that data for you and your colleagues. Being able to democratize that data so the people who are on the front line of experience have these insights and can affect the experience where the customer really is across channels. And then lastly, that ability to integrate those insights into other systems to operationalize customer insights.

Slide 8:

With that, let’s dive into the tips. I’m going to share eight tips. I’ve organized them into these four categories, so that wherever you feel like you should be starting, you can go from there. The only other thing I will say is that the tips are going to be focused on recent innovation in Adobe Analytics, but we’re going to talk about new ways to apply some of those innovations. So even if you think you’re familiar with a feature, hopefully there’s something new in there that you can take advantage of in a new way.

Slide 9:

What can you do now to start building out a 360-degree view of the customer? As of last week, we have done a server integration between Adobe Audience Manager, our DMP, and Adobe Analytics. What’s great about this is it means that all of that first, second, a third-party data that you’ve got flowing into Adobe Audience Manager is now available for you in Adobe Analytics as a dimension so that you can do things like customer journey analytics to show how people are moving from one audience to another.

If someone qualifies, they move from a new customer to a loyal customer, for example, or they move from a gold-level member to a platinum-level member. You can actually show that as you see here in the screenshot. That ability to bring second and third-party data segments into Adobe Analytics really opens the doors for you to understand your customer a lot better. That 350-degree view just became quite a bit more attainable for many of you.

If you don’t own Adobe Audience Manager, I’m certainly not here to sell, but that’s maybe something to look into if you’re interested in integrating your second-party data into Adobe Analytics and applying that to your first-party behavioral data that you’ve already got. If you are using Adobe Audience Manager today, you have the ability to plug this in. You can see in the screenshot the flow visualization, and you can see how people are moving from one audience on the left, and they’re qualifying for additional audiences as we learn more about them. You start to get a sense of that journey and see how people are moving between these things.

Then the ability to use these segments that you’re bringing in from Adobe Audience Manager in the Segment IQ or Segment Comparison feature. You can compare, for the first time, demographic traits and psychographic traits, interests, people who are in-market for certain products and how they behave differently, using the Segment comparison feature which hopefully some of you, or all of you, are familiar with from past presentations and your own experience. Being able to plug in—to take a simple example—how do my 31 to 40-year-olds behave differently than my 41 to 50-years-olds? Are there different things or different propensities that they have?

That kind of insight has not previously been available, and can now be brought in without a single change to your implementation and you can run that right through Segment Comparison. So, something to be aware of, hopefully for everyone on the call.

Slide 10:

Another one, this is actually not something that we’ve done recently, but it’s definitely something worth knowing about, and I would say it’s one of the most underappreciated capabilities of Adobe Analytics, is the ability to bring in offline transactions using transaction ID data sources. The was this works is at the last point of an online transaction, where it moves form online to offline, for example lead generation, when a lead is generated, you set a transaction ID and you pass that transaction ID both to Adobe Analytics and into your CRM system or ecommerce system depending on what you’re collecting.

Then whatever happens that’s associated with that lead or that transaction ID, you upload back into Adobe Analytics. What that gives you is fully segmentable, fully breakdownable offline data. For example, you can see here down on the right, we’ve got this Order Returns, and that’s something that happens offline, but can be tied to an online transaction, occurred at purchase, and those orders returned can be uploaded back into Adobe Analytics so you end up with Orders, Orders Filled, and Orders Returned, so you can do a truer analysis of people who actually kept the product. That’s the kind of thing that’s possible here.

You can also upload cost data. If you want to identify profitable orders versus unprofitable orders, cost is typically not something that people would pass directly into Adobe Analytics from the site, but if you set the transaction ID for every purchase, you can then upload the cost associated with that purchase back into Adobe Analytics and segment against it. Segment for high-value customers versus low-value customers, that kind of thing. Strongly recommend looking into transaction ID data sources, if that’s not something you have done previously and if you have a model where transactions or returns or other noteworthy events are occurring offline that could be tied back to online data.

Slide 11:

Machine learning. What are some tips that you can apply here?

Slide 12:

Hopefully most of you are aware that last year we introduced intelligent alerting where alerts in Adobe Analytics are now based on anomaly detection, not based on a static percentage increase or decrease. This on its own is not new, hopefully you’ve seen this, but some things to be aware of in terms of tips: creating alerts is a lot easier than it used to be. This is one that can certainly save you time in keeping your organization informed. As you’re working in Analysis Workspace, you can right-click on just about anything, any cell or dimension value, and create an alert directly from that selection. That’s simply a time-saving tip there.

The other tip that I love related to intelligent alerting is setting these up and sharing them, or having them sent to a group of users rather than a series of email addresses or series of individual users. If you have user groups that break down by business unit or job function, input those groups here in the recipients, as you can kind of see that outline there in red. The advantage of that is that as colleagues move in and out of those groups, they automatically get subscribed to these alerts.

This reduces the load on you as an analyst to manage all of these alerts, and if you’re like most analysts, you get asked to set these things up. It’s less common that your colleagues are setting them up right now. This just makes your life easier as an analyst so you can send these alerts out. These alerts are based on anomaly detection so they’re much more relevant, fewer false positives, and generally a lot more useful, a lot more actionable for the recipients of these alerts than previous alerts or other ways of keeping people up to date on what’s going on with your data.

Slide 13:

The other capability to be aware of here, and I’ll share some tips on this as well of course, is Segment Comparison. As you hopefully know, Segment Comparison lets you compare two segments and automatically it uses machine learning to compare those segments to find statistically significant differences between them across all of your dimensions, metrics, and segments. It will tell you how two segments behave differently based on the different dimensions and dimension values that we see for each segment, meaning the things they actually do, the values of eVars and props that get set by those segments more often than not than other segments. As well as the metrics, of course, that they generate and the other segments that those customers belong to. That’s Segment Comparison.

Some things to do here that we would recommend is you can take your third-party audiences that we talked about earlier and use those to determine high-value characteristics in Segment Comparison. You can compare two segments based on third-party data like demographics or psychographics, see how the segments are different, and what you’ll get back are the characteristics of those third-party audiences. Those will be first-party characteristics, so what you can do then is turn those characteristics into their own segments. What you end up with then is first-party segments that look like the third-party segments, but allow you to extend those out a little bit, extend out the analysis so you’re not limited to just the third-party attributes, but you’re also bringing in first-party attributes as well.

Then of course, you can send those to Adobe Target to target people. You can send those to Adobe Campaign to run email campaigns or get those out through our API and integrate them into something completely different that maybe we haven’t even thought of yet.

Loyalty analysis. Determining how customers behave in loyalty tiers, in one loyalty tier versus another. This is a great place to start here, if you’ve got that data coming in through customer attributes or classifications. All of your classifications will work here in Segment Comparison.

Then lastly, one of my favorites, and something that I think we’re probably all in the boat of trying to understand how people interact differently with our web properties versus out app properties, so you can plug, you can segment by web versus app, compare those two segments, and see what behaviors are different between the two in an environment where you have some functionality bleed over between web and app. Just some ideas on where to get started with Segment Comparison. And some interesting applications of this tool that we have seen actual customers of Adobe Analytics get tremendous value out of as they’re trying to understand those customers to improve the experience through the operational channels that I mentioned, Target, Campaign, APIs, etcetera.

Slide 14:

Democratization of insights. This is my personal favorite topic, I think. We’ve hopefully come a long way helping you democratize data with Analysis Workspace and some of the other tools that we’ve introduced in Adobe Analytics.

Slide 15:

I want to pick on one in particular. This is the similar screenshot to the one that I showed earlier, but it’s about flow and showing how people move through a journey. What I love about flow for data democratization is that it is an exploratory tool. Flow gives you the ability to just drill into the things that are interesting to you just by clicking on a particular step. So, if I wanted to see how people move from page to page in this screenshot, I would just click on one of the pages here that you see on the very right. It would give me the next step in that journey. I can start to figure out what customers’ natural journeys are through my site.

I mentioned that one of the great things you can do here is show how people move from audience to audience, which is what you see here on the left of this visualization. Another important thing to note here is that you can create interdimensional flow. You see here that we’ve got audience name on the left and then we’ve switched to page. There’s no limitation to flow in Analysis Workspace on the combinations that you can make. You can show how people move from marketing channel to entry page to product to search results. These are all different dimensions that you have likely in your implementation and this tool will allow you to show that journey even if it’s not just from page to page or site section to site section. I love that about flow.

Incidentally, one of the things some of you may be thinking is, “Well that’s great, but I’ve tried this and my flow visualization is kind of polluted by reloads. People refreshing the page and causing their next step in their journey to look just like the last step in their journey.” What you can do there—shout-out to our consulting organization for recommending this—you can build a segment that excludes reloads and apply it to your flow visualization and that will help you solve this problem. There’s another additional tip for you as well.

The other thing that i’ll mentioned is that ability to show how someone moves from one segment to another, as you see here on the left side of the visualization, that really is powerful. If you want to show how people move between loyalty levels, for example, that’s something you can easily do here. Super powerful. Not something that we’ve been able to do before. Customer attributes in Adobe Analytics are great, but they’re not time based, so if you’re a gold member, you’re always a gold member. Whereas with flow and with audiences coming in from Adobe Audience Manager, you can get that time-based view. And as I already mentioned, excluding reloads.

Slide 16:

Then I just wanted to call out some of the data storytelling features that we’ve added in Analysis Workspace that we’ve added as a tip to help you democratize data. We have added a number of these things. We’ll kind of go through them. First, text boxes and descriptions, we’ve enhanced those greatly. You can now put links to external documentation, so something that a lot of you have wanted to do. You can add a link to your internal Wiki or other documentation that you have to help people understand the data you’re sharing with them.

You can also link to specific visualizations. If you put a table of contents at the top, people can click on those and it will actually link them down in the project to the visualization you want to link them to. You have control over font settings so you can call out important data points. You can see you’ve got some green and some red. It really become a much richer experience to give someone an overview or to describe the data. I mentioned you can add links to other projects as well. You can link out to a different dashboard or a different project as well. You can right-click on any visualization to get a link to that visualization that you can drop into the table of contents so users can jump to specific panels and visualizations.

Slide 17:

Last in this category is our integration with PowerBI. Hopefully you’ve heard about this. You can access sort of out-of-the-box Adobe data in templates, using the content packs that exist in PowerBI today, if you’re a PowerBI user. You can publish Report Builder workbooks, which is great so you can go build your dashboard in Report Builder, get it looking exactly the way you want it, publish it to PowerBI, and then your colleagues can get that anywhere. What’s important about that is that PowerBI has a great mobile app. So, if you’re looking to get a really rich Adobe Analytics experience for your colleagues, for you executives, on the go that they can access of their phone, pushing these dashboards to PowerBI and the having your colleagues download and install the PowerBI mobile app is a way for them to get some really good insight on the go.

Slide 18:

Last, but certainly not least, integrations.

Slide 19:

These are mostly informational. I don’t have any mind-blowing tips here, but some things to be aware of that you may not know that we’ve done. First are livestream triggers. This is a real-time, event-level integration between Adobe Analytics and Adobe Campaign where you can define segments like cart abandoners or highly engaged users and then as data comes into Adobe Analytics, it’s sent directly to Adobe Campaign to kick off remarketing campaigns, to boost engagement, and to increase touches between customers and brands. This is something that you don’t have to own livestream in Adobe Analytics to use. If you own Adobe Analytics and Campaign, talk to your account team at Adobe about getting up and running with livestream triggers. It’s a really cool, powerful integration.

Slide 20:

Then last—I believe last—in here is a little bit of a different kind of integration. It’s not as much about putting the data in an auctioning tool or an operational tool, but being able to get data from Adobe Analytics Directly into Slack where people are actually talking and engaging and discussing the business. There are a couple of these Slack integrations. I’m going to call out one that was written by our friends at 33 Sticks. You can go to to install this on you Slack team. It’s awesome. What it lets you do, is it lets you request data for simple queries, but you can get top products over the last few months—you can control the date range, control the metrics and dimensions in there. You can ask it these questions and it uses our APIs to retrieve the data and spit it right back out into Slack where your team can see it.

My personal favorite tip here though, actually isn’t shown on the slide and it’s the fact that you can use this Slack bot to create new users in your company and assign them to groups. If you’re like me, you hate having to login to a product just to create a new user. You no longer have to, you can do that directly in Slack using this Slack-bot and save yourself the hassle of logging in. so a couple of really great applications of Adobe Analytics data being used in other systems to improve experience or to democratize data and analytics.

Slide 21:

I am out of time, but if you have a moment while I’m wrapping up, hop over to and we would love to know which of these tips you think will be most powerful for you in your organization, so that we can help plan future events and future content, things at summit 2018 and so on. We would love to have your feedback on which of these tips you think will be most impactful for you. Again, you can see the URL there:

Slide 22:

Very last slide. Some other ways to get some great tips. subscribe to the Adobe Analytics YouTube channel for more tips. we’ve got a ton of great stuff there. We’re continuing to churn out videos and we’ve heard from many of you that that’s been a great way for you to get to know the product better and to come up with ideas to move toward that customer intelligence vision. Join the product forum at to give feedback and interact with Adobe and with your colleagues. And then shout-out also to the digital analytics Slack team. You can go to to join that team. I’m out there, many of my colleagues are out there listening and talking with many of you, so we’d be happy to brainstorm tips and tricks for customer intelligence out there as well.

Slide 23:

With that, I want to thank ObservePoint again for having me. Always a blast and always great to be with practitioners and others in our industry.

Previous Video
Brent Dykes, Domo - Effective Strategies for Building a Data-Driven Culture
Brent Dykes, Domo - Effective Strategies for Building a Data-Driven Culture

Learn about the common barriers as well as best practices and strategies for building a data-driven culture.

Next Video
Daniel Herdean, Cognetik - Combining Multiple Data Sources to Tell the Whole Story
Daniel Herdean, Cognetik - Combining Multiple Data Sources to Tell the Whole Story

Learn how to become an analytics hero by connecting clickstream, paid, social (Twitter, FB, YT) and CRM dat...