In Stitches: Assembling a Complete Visitor Data Profile
Both Scott and I are super pumped to be here today and to be asked back to present at the 2017 ObservePoint Summit. It’s quite an honor. We’ve listened to a few of them today and we are super excited about the content and ways to learn about new initiatives. With that being said, what Scott and I will be talking about today, as the title conveys, assembling a complete visitor profile with stitching visitor data across devices. The way we came up with this concept is every year we take a look at our users and what our clients are doing, and kind of where they’re falling short. This idea kept popping up, so we wanted to talk about that and ways to remedy that. That way you can see a holistic view of your user.
With that, let’s go ahead and get started. What I’ll be covering today is the definition of visitor stitching, the value props for setting this up, the overview of the concept of enabling visitor stitching, and the difference between a visit, which is a session, and a visitor. We’ll also be talking about a conceptual explanation and how-to for Adobe and Google, and also how to validate all of this using Merkle’s ObservePoint auditing product.
As I said earlier, I know this year’s theme is setting everyone up for success when it comes to this. When Scott and I were deliberating how we could go about presenting this, instead of ideating and coming up with great ideas as to why visitor stitching is so important, we thought it’s be a better use of everyone’s time if we showed you how to do it. I think that you can take that away and take it away as an action item more than just an idea of what visitor stitching is and how you can use it. The best way to go about it is setting it up and putting it into play.
At a basic level, we’re going to go ahead and define that. It’s the process of combining visitor data across devices in an effort to create a holistic picture of that user. The biggest value prop, stated here, is that in a consumer centric world, which we all operate in, integrating data from all of our user touchpoints is most essential. It affords us, the businesses, to know how our users choose to engage with our brands and it can also provide insight into which obstacles, which I think is most important, should be removed to increase conversion with each touchpoint.
I do want to point out, your users who prefer to engage with you on a tablet or mobile, may be a little bit different from users that primarily seek out a website. Understanding where your users want to engage with you and where they want to convert, if it’s not the same thing, is very important. Without it, marketers are often left guessing and making things up, which is what we want to try to avoid.
Before we get into a visitor data without visitor stitching, you have to know at its base what’s the difference between a visit and a visitor. Any time I onboard any of our new hires, we go through this type of training. The best way I’ve seen the definition stick is, let’s go ahead and pretend you’re a person going into a mall. You are getting out of your car, you are one unique person going into a collection of stores in this particular mall location. So, as you get out of your car, you are that unique visitor and the stores inside of the mall represent visits per session.
You may take a visit to Macy’s you may take a visit to Nordstrom or Eddie Bauer, but you are only unique person going into one big conglomerate or location and paying a visit to different locations within that store. The same can be said about how users engage with your brand, they may come to your store, they may come through your web presence, they may come through the mobile app or mobile phone.
Today what we’re going to be talking about is how to track with a deterministic, which means an authenticated user. There are often times where we don’t have the…
The ability or process set up on a site to be able to have a login process. That’s really the primary thing that must happen. If the person doesn’t login on the site, there’s no way to make the distinction between that logged in person on their tablet and on their phone. There’s no way to tie those people together. That person logs in on their mobile device and we set a unique ID for that login. Now, if they login on a different device, we’re able to connect those dots and see user A, B, C now logged in on different devices. Now all their data gets collected together into one data collection. It does that for these devices going forward unless they delete their cookies or when they login with a new device, that device will get joined into the group of items.
This graphic show here shows that without the visitor stitching, if someone comes to this mobile app, laptop, tablet or desktop without this stitching enabled, you see these pieces of data all separate. There’s no collection at all. That’s something we’ve got to deal with.
Yes, because we don’t want to inflate any of our visit or visitor metrics when we’re actually seeing a unique person or unique user is coming to different touchpoints, which is what we want to talk about and clear up today.
That’s a great point. In this example here, if one visitor came through all these different devices and you’re reporting without that data being stitched together, they would look like four completely separate people and four completely separate visitors rather than that one collective piece of data. That can be quite misleading.
This is the slide that we want to avoid, which unfortunately, a lot of our clients that are coming in, we actually see this. We see them have different tracking or collection when it comes to mobile versus laptop versus a tablet.
What we want to do is unify all this so that our landscape looks like this. Anytime I come through on my mobile phone or app or tablet or desktop, I am seen as the same person or the same unique visitor. That is very important.
Now that we have gone through and defined what is the value prop, we do want to take you through a quick “how-to” to set this up. A lot of our clients out there are using one of two platforms, we’re going to show you how to do this and give you an idea of how to set this up across Adobe and, or Google, and then tie it all together with an ObservePoint audit and take a look at what you’re actually tracking to make sure what you’re doing is working and you can authenticate it.
First, we’ll start with Adobe Analytics. Like Danielle said, there are many different—you can use Adobe Analytics, Google Analytics, and there are other tools out there that we won’t get into, but we will start with both of those. With Adobe Analytics, there are two primary options for enabling visitor stitching. You’ve got the marketing cloud visitor ID service, which is there, this is definitely the much more recommended option.
This allows for simpler tool integration, using GTM, and things like that. It allows the visitor ID to use across multiple Adobe products, Adobe Analytics, Adobe Target, Audience Manager, and so on. Actions that take place within Target or Adobe Analytics or any of their tools should all still be seeing it as the same visitor across all of those Adobe products, which is a pretty amazing change. It doesn’t allow a much better, complete understanding of that visitor’s process.
Lastly, it allows your organization to continue to update to newer versions of Marketing Cloud products. If you don’t use the Marketing Cloud and you prefer to use the legacy option, you end up being stopped. They’ve stopped updating that. You can no longer use that version for your code. The legacy version is the opposite of everything I said above. It’s not recommended by Adobe. It’s more complicated to implement. Newer code versions are no longer going to be supported. And different visitor ID is set across each of the Adobe Marketing Cloud products. It no longer has that connection like we see in the Marketing Cloud ID service.
This part of the presentation, we’re not even going to walk through the legacy service because anyone who’s doing a new marketing cloud or visitor ID setup should be using that Marketing Cloud recommended approach.
In order to get started working through this, there is something that every company has to have. I’ll burn through this quickly since time is sensitive. There is an organization ID that each company has to have. This can be gotten through your administration page at marketing.adobe.com. if you’re unable to find that or you’re not an administrator and can’t access that page, that’s something you can contact Adobe about. You may have to find an admin that has those kinds of rights in your company.
If you don’t know who that is, Adobe Client Care is able to provide you with the individuals in your company that do have that type of access. They won’t do anything like add those rights for you, but they will be able to direct you to a person within your company who does have them. If you don’t have an Adobe Experience Cloud enabled account, you have to have that in order to set this up. Most companies who have created an account in the last couple of years would most likely have this, but you never know. If you aren’t able to login or you don’t have that organization ID, you will need to work with your Adobe account manager if you have one specifically assigned to your company. If you don’t, you can contact Adobe through this firstname.lastname@example.org email address and that’s a conglomeration of Adobe account managers. One of them will get back to you or they will direct you to the right person or you can work directly with them to get that.
One of the things that’s important to be considered with this process is you have to have, besides an enabled account, there are some minimum code requirements. We touched on this briefly before in talking about how the newer version of code are sort of stepping beyond some of these older options. I won’t go through everything here, but this API code file in the first column there, that’s imperative. Every implementation using this must have that. There’s additional documentation through the Adobe support section if you’re logged into that on the Marketing Cloud tools.
Within analytics, the next row down here, there’s appmeasurement.js, s_code.js, and then also something called Video Heartbeat. Those three things all relate to analytics implementation, any analytics measurement that’s been done is likely under appmeasurement.js, and as long as you have version 1.6.4 or higher of that particular file type, you’ll be set. S_code is over, sort of the legacy version of AppMeasurement Adobe Analytics implementation. It does work with the H 27, but it wouldn’t be recommended. If you haven’t already set this up in your current implementation, I would strongly recommend you move toward the appeasement implementation since you’re able to update from that place forward.
Adobe’s made some vast improvements in how these visitor IDs work and how much more seamless they are in determining who a visitor is and isn’t as they go throughout your site. Anyway, those are the recommended versions. I won’t get into the specifics there, but it’s things to keep in mind. That would be a good point of reference.
Mobile for SDKs if you’re doing mobile apps implementation or tracking. You need to have the minimum analytics SDK version or 4.11.0 for Android and the same for iOS. If you do not have those versions, they might not work, you might not have much to track.
As far as implementing the process, you have the Marketing Cloud ID service tool that you can login into Adobe’s site management tool called DTM or you can go through an older, more traditional method and not use DTM, or if you don’t use Adobe’s DTM, which is fine, DTM just makes it simpler.
These are the quick steps of doing that. Basically, with DTM, you select the appropriate property, add a Tool, you select from the dropdown Marketing Cloud ID Service. Once you select that, the org ID should be there for you already because you’re using an Adobe tool. You enter your analytics tracking server. It’s the same value that you’re using with your Adobe implementation now, if you already have one, and you create the tool.
Once that’s set up, it will create a code-version of that library, which brings us back to the chart, you need to make sure you’re using the newest version of your appeasement code so that new version will work. If not, you can go to the chart to reference and select another version and it will allow you to make that change.
Next you need to have a visitor ID value. This is the unique identifier that determines who this person is. Once I’ve logged into a site, I now have a login ID or username and that’s the value that most companies will use and set that visitor ID cookie in order to allow you to retrieve that data and then pass that on all user requests following. Most companies put this in when they’re using DTM and they’ll pass that data into a Data Element that you can then use as part of setting up your Adobe Analytics tools setting sections. It’s a rather simple process. Again, it is document in the help section, but I thought we needed to point that out here.
If you’re not setting using a Data DTM Element, then you can set your value in your Custom Code within the Library Management area, or the Customize Page Code area of the analytics tool as well.
Regardless of how your implementation is down, they will all need this VisitorAPI.js file. If you’re doing it through a manual implementation without DTM, this is how it would be uploaded to your site. And you would need to have a call to that file on every page on your site within the head tag of that page. You can see already, that the process without DTM is more complicated, more confusing. As a quick point, you can see there’s a screenshot of what the code looks like, what values need to be placed. Organization ID on that first one and the tracking are all the same.
The Adobe process using DTM is so much cleaner. You don’t have to populate these things. If you are choosing DTM and you want to go down this path, there are help documents do give some additional details on what these values actually mean. Here at Axis41, we can help with that kind of thing as well. That’s something that you would want to look into to show to keep it simpler.
Like I mentioned before, all products require the VisitorAPI. Each Marketing Cloud tool has a slightly different process for implementing the visitor ID. We focused on Analytics in this case.
Last, you want to confirm they’re functioning. In the network tab of your browser or using a tool like Charles or Adobe Debugger, what you look for is a value called MID. That giant, long number there is your visitor ID. That’s what you would look for, so if you’re seeing that image request, you have gotten your visitor ID, the Marketing Cloud is being set up and that’s how that value is being populated. You can also check your website cookies and confirm that this cookie that begins with the name AMCVS contains your appropriate organization ID. Those next 14, or whatever characters there, are what indicate your Adobe organization ID. So, a couple things you can check to be sure.
For Google Analytics, their process is a little more menu based within their Google Analytics tool, but there is also some code that needs to be done as well.
First, you set up the user in your Google Analytics account. In order to do that, you also have to accept their user ID policy, you also have to add your user ID in their tracking code, but same thing—the user ID is your unique identifier, your login ID or login name. you’ll confirm the needed option for session unification—I’ll get into that in a little more detail. Then you create a user ID reporting view, which is the way that data is displayed in Google Analytics, which is quite different.
Next, we’ve got where to access the setup of the user ID in Google Analytics. These are some quick steps. Sign in, go to the Admin section. If you aren’t an admin, you won’t be able to do this. Click the tracking info, user ID section. Next it will force you to agree to their user ID policy, basically saying it’s okay they keep this data and pass it from visit to visit. They’re going to do that anyway, but since it is your customer’s unique ID, you’ve got to set that value. Once that’s enabled, you switch that to on. It’s set up and you click next step.
This is the piece of code you add to the tracking code. Just the ga(‘set’) equals some value. Again, that same value we’re talking about for Adobe Analytics, it’s your login name, user ID, whatever. Just like in Adobe Analytics, unless a user is logged into a device, this data will not select. If they’re logged in on one device, but not logged in on the other, there’s no way for Google to know either that those two are the same person.
Google does have some other backend things they can do since they are Google and they own some other data points on that person they’re touching, that they might be able to home in on some of that, but for the most part, that’s not happening for you. That’s why we have to do this and that’s why people miss so much of their data. The user ID can be setup for both website and mobile app. Both types will be accepted into reporting. Whether a person’s using their website or mobile app or whatever, this can be in the same reporting area. It’s just a matter of setting up your views to deal with that.
It’s certainly recommended that you enable this for your app, about anything that’s touching this particular site and user so all that data is collected into one. Just like we talked about with Adobe. With the unification settings, it sounds complicated, but basically if a visitor comes to the site, they visit some pages and then login, if the value set is one, then those seven pages that were seen before they logged in will still be bundled with that unique visitor. If that value is set to Off, then those first seven page views or visits will be deemed as a separate visitor until that person is logged in. I’m not terribly certain why you would want the second option.
Create a user ID reporting view. This is just how it’s seen in reports. You create a new reporting view. Once you’ve created that new reporting view, we recommend that you include the user ID name that you create just so it’s easier to determine what this view was created for, whether it was with or without that user ID. Viewing options setup, again, just the way the data is seen and reported. If you look at Adobe reports, you’ll see users who were authenticated and users who were not authenticated in two separating reporting windows. That’s one of the massive distinctions or differences between Adobe and Google, and it can be and bad, but it is something that Google does handle differently, so I thought it’d be important to call that out.
The great part is, no matter which platform you’re on, ObservePoint’s auditing technology is agnostic. For every implementation that we do here at Merkle, we always start it, when we talk to our clients, with an audit of their current landscape, as far as connectivity or if we’re looking at information architecture. And also looking at missing tags and opportunities for actively closing the gap. And this we actually see a lot of.
Any time we onboard a client, we go ahead and run and ObservePoint audit as the initial step and also run it as our closing step. The great thing about the ObservePoint audit is it actually gives you’re a grade saying out of 100 percent, this is where you stand. When we’re going through and running our systems of checks and balances and making sure, what the strategy has gone into for a client is actually delivered upon.
This I know for me is one of the best moments of interacting with the client. They’ve paid all this money to have some work done by our company, we walk through this process, we set everything up, and we’ve got a really easy way of saying, “This is what you had before. This is what you have now.” And it absolutely proves that what we told them we were going to do, is what we did. That’s something that is a great bit of value for anyone in their organization, whether it be initial folks you work with as far as setting up the scope and the plan, even though interacting with the data on a daily basis allows them to be much more confidence with what their data is and why is was set up the way it is.
It also is a good way to ensure accuracy. Once these attributes like visitor stitching and other things that we agreed upon when we are setting up strategy with a client and it’s our responsibility to deliver on, this is a great way to ensure that A) they’ve been done, and that they are accurate.
With that, we encourage you to go through this on your own time. Write down some action items and set this up with your clients or with your organization that you’re currently in, and hopefully start stitching your data together to create a holistic view. Thank you.
Thank you. It’s been great talking to your guys today.