Eric Matisoff, Adobe - Become an Adobe Analytics Master

November 22, 2016

Slide 1:

Hey everyone, this is Eric Matisoff. Today what I’m going to be talking through is some really fun and powerful tips within Adobe Analytics Analysis Workspace to get you to become even closer to that analytics ninja that we all want to be so you can make the most of the tool. The good thing about my session is I don’t want to bore you with too many slides, so I only have two. The first is my intro slide.

Slide 2:

The second one is how you can get a hold of me. Feel free to reach out to me on Twitter or the measure Slack. I’m usually very available on both and I look forward to connecting with you after this session. And if there’s any additional questions around how I was able to do any of these different features, you can ask me there. Of course, asking in the Q and A pod is a perfect way to do that as well.

Demo Screen 1:

What I’ve done is I’ve opened up Analysis Workspace here. For those of you who have never opened up Analysis Workspace within Adobe Analytics before, shame on you! Do it now. It’s the greatest tool to come to Adobe Analytics in a very long time. We love it. It’s been adopted incredibly quickly. Some of the features that I’ll be walking through are brand new, some just released last week, some released just a few weeks ago, all extremely powerful, super fast. The goal of which is to enable analysts just like yourselves to be able to dig into the data and find it in a more accessible way.

I’m going to be bouncing around across a few different tabs here. The first one that I’ve got pulled up is a relatively, maybe somewhat boring looking project to some of you, but you’ll notice that what I’ve done is I’ve actually recreated the marketing channels report that you see within reports in analytics. That marketing channel recreation project that I created, again super, super simple. I just pulled in Last Touch Channel as my dimension, and put it side-by-side with First Touch Channel. Then I’ve have for this report suite, it happens to have revenue and orders within it, so I’ve pulled in both of those metrics into both of them, and I can see side-by-side last touch versus first touch for each of my different marketing channels. First of all, that’s great. I happen to be a big fan of being able to do that within Workspace because we’ve got tons of flexibility in terms of how we drill into all of these different sets of reports and dimensions and metrics, segmentations, update the dates, etc.

The first thing that I’d like to show you is something relatively simple, but somewhat hidden because it’s towards the bottom here on the left. On the left, you’ve got your left panel, which contains all of your components. It’s very easy to just focus on dimensions, metrics, and segments and forget that at the bottom here, you’ve got time. I personally happen to be a big fan of leveraging the different dimensions that are within time here. Obviously, you’ve got your calendar, which allows you to set the calendar for your particular panel that you’re working on, but I also find it quite interesting and quite valuable to also be able to break things down by day and hour and month and more.

The first thing I’m going to do is show you how easily and quickly you can do this. If I were to select email for example, if I want to see that maybe over time within my table here, I can take a look at the day and I can drag and drop that in there. Now it’s going to pull all of that data based on my time frame here per day, which is really nice. It’s good to be able to do that side-by side comparison of my two metrics broken up by day specifically for email. That’s nice, and it’s certainly helpful to be able to do that, and of course we could drill in further.

That’s one of the beautiful things about Analysis Workspace is you’ve got an unlimited amount of dimensions that you can keep drilling and drilling and drilling into. But at some point, it becomes a little bit of a mess to be looking at all that data at the same time. So what we’ve done is, if I back those two items out, we’ve made it very easy to trend virtually and dimensional value, or any metric tied to a dimension, within the platform. The way that we’ve done that is very simple, you just click the little graph here, click “visualize”, and within a second it will pull up the trend for any metrics that are tied to that. It’s a good little time saver.

A recent feature that we’ve added just a couple weeks ago is around that anomaly detection that you’re seeing there. In that you can see, based on predictive modeling, were there any anomalies with the data that you’re currently looking at? So what you’ll see it we’ve got revenue here, we’ve got orders on the right, but—since revenue is so much higher than orders because we’ve got more than just the dollar per order getting pulled into this time frame—it’s difficult to see orders. So what I can do in terms of interacting there, is I can click revenue and it will remove that from my graph. Now, I can focus in just on orders, and once again, you’ll see the anomaly detection is popping up there as well. If I want to put back in revenue again, I can click that one. Again, just a couple of little time savers there, a little tip and trick for adding and removing those different items.

Of course you can have that graph which will be live linked to the table below, so you can see as I switch to referring domains, now that changes also within the graph. As I switch to social networks, same thing. Then, what’s quite powerful, is you’ve got the ability to maybe lock one of those items. So if we say email if the one that we want to focus on, we can left click on this little icon next to the word “line” within my line graph, and we can say we want to lock that selected item. That way, regardless of what I’m clicking on down here, it’s not going to update what’s up here, but I still have the ability to interact with the graph. Just one of the fun ways that you can utilize the platform to make the most of your time spent within Analysis Workspace, not accidentally clicking around and disturbing graphs that you have linked. Another really nice feature that we’ve built into the tool is the ability to copy and paste. Now I’m going to switch over to another report suite here.

Demo Screen 2:

What I’ve done is, I’ve got a prop that I’m pulling a ton of different values in. It looks like I’ve got 77 different values getting pulled into this report. What I wanted to do was use the classification rule builder in order to automatically and dynamically classify the information within that report. What I’ve done is created a relatively simple regular expression here that’s looking for the word “dead”. What that’s helping me find, is if there’s any content that’s being clicked that isn’t actually linked. I was able to capture that information very easily using Adobe’s Dynamic Tag Manager, and all I’m doing is on-click of any dead images and sending on what page that specific image is being clicked. I also have the same thing with dead headers as well. If someone is clicking a header that doesn’t have a link on it, then once again, I’m able to capture that information in this report and better understand how users are expecting items to be clickable that simply are not.

I’m a big fan of making the most of my props when I’m working within Adobe Analytics, so I’m not only focusing on those dead images and dead headers, but I’m also focusing in on all pages and all actions. That’s going to include clicks on nav menus, views of particular pages, or literally every page within the site, so I’m combining into a single variable a combination of page views, custom links, exit links and download links. I find that report to be one of the most interesting reports that I have. And what I’ve done, is I’ve built out this classification of it, so I can more easily focus in on what I’m more interested in. Of course, I could very easily click my filter button, click “dead”, and I could apply it, and now I can see a list of all my dead images and headers.

That’s one way to do it, but another way to do it is, I’ve actually built—like I showed you a second ago—this classification rule builder based on regular expressions. The reason that I keep bringing that up is, it took some work to build that regular expression. That’s not a super easy regular expression. I had to make sure that it started with the word “dead” and that it had additional content in there with a colon, and some additional content afterwards, and then I’m sending that to my classification of that prop1 variable.

Before I hit “Activate” on that, I wanted to test it. So what I’ve done is, I’ve enabled the test rule set feature within rule builder, and what I can do really easily is I can say, “Okay, here’s all the data I’m capturing within this report. I’m going to shift-click and highlight all of it, and then I can copy it.” It’s going to be whatever kind of copying keyboard shortcut you’re used to. If you’re on a Mac, it’s going to be command C. if you’re on a PC, it’s control C. and when I head on over here, I can then paste all of those different items right into my sample keys here, so I can see: what will every single item within this report look like within my classification? So I can hit one and confirm that that regular expression that I set up, it’s going to be created properly.

So what I can see is that since I copied all over the 76 different items. I can see the ones that have the word “dead” in them—these dead images and dead headers—those are working correctly. They’re pulling in the fact that it’s an image, and what the URL of it was, or what the text of the header was, and anything that doesn’t have dead in it that’s unmatched. That’s working exactly the way that I wanted it to.

Demo Screen 3:

SO that’s sort of a double feature tip for you. In that, first of all, you’ve got the ability to copy values directly out of Analysis Workspace, and you can use that to confirm the data that you’re looking at based on the classification rule builder and the rules that you set-up within it.

With that, I’m going to keep pressing on. Something that you may notice here is that I’ve got a couple of different date ranges pulled into this table. The reason that I’ve done that is to show you something interesting that can be done. I’m actually going to remove those out, and what you can see is that we’re just simply looking at occurrences based on that prop. Once again, going back to our time area over here, we can see, very easily, that I can pull in, maybe the last full week, apply that date range to my metric, and then it makes the update right within the table. If I wanted to, I could also do an additional timeframe that’s maybe two months ago. If I pull that in here, then I’ve got a whole bunch of “Invalid”, and I can go back on over here and search for occurrences, and drag that in and place it carefully under “2 months ago”. And then I can see my comparisons side-by-side.

You can see that that took a few extra clicks, which I’m not crazy about doing. Not only clicks, but searching, and I’ve got a better way to do that. I’m going to back out those changes, remove and remove and remove. Keep in mind that you do have the ability to do and Edit Undo, which can come in handy. What you can now see is we’re back to the original. We’ve got occurrences, and I’m going to head back into time, and instead of putting the time frame above the metrics, I’m actually going to put it below that. The reason that I do that is right here. If I wanted to, once again, head on over—and I think I was looking at two months ago—if I wanted to do those side-by-side, now I could do that and I don’t have to close out of time, type in occurrences, drag that in, and push it into the right spot. Instead, it’s very easily side-by-side within the tool here. Again, just another little time saver for you, and hopefully that will make life a little easier for you.

Something else that’s really cool that we’ve only released just last week, is the ability to do time comparisons within Analysis Workspace. The way that we’ve done it is super, super simple.

Demo Screen 4:

I’m going to head back in over to our marketing channels recreation project that I created. And what you’ll see is that when you right click on a metric, you’ve got the ability to do a “Compare time periods” and this is brand new. Literally released last Thursday. When you click “Compare time periods”, then you can say this item here “Prior 30 days before this date range”, that’s actually dynamically populated based on what your date frame is. So if I click “Prior 30 days before this date range”, it’s now going to add two columns to my table here. The first is looking at the prior 30 days here, and then we’ve got the percent change as well. With the percent change, it’s a pre-built, auto generated conditional formatting. Of course I could change that if I wanted to, instead of auto generate I can do a custom set of it that maybe looks a little more friendly, in terms of how I want my red, green, and yellow conditional formatting to appear, but the concept is still the same. Just within a couple of clicks, you’re able to compare our July data that’s tied directly to my date selector, and my June data, specifically for revenue, and the percent change over that time.

Now I’m going to undo those couple of items and we’re going to come back on into revenue because there’s another choice here. Actually, before I even do that, let’s take a look at one week. So we’ve got one week’s worth of data and now when I right click and go revenue, you see how it automatically includes some dynamically populated data ranges in there. It’s, first of all, going to say prior week to this date range, so instead of looking at July third to the ninth, it’s going to compare that to the previous week rather than the previous thirty days as we saw just a second ago. Or we can compare it to this same exact week last year. Lots of really nice dynamic functionality there, the goal of which is just to make your life a little bit easier when you’re working within the tool.

You also have the ability to do a custom date range. Within these custom date ranges, you can mess around with what exact days you want to do a comparison on. Really nice feature. In addition to custom date ranges here, it may also be of value to pull in some custom date ranges in your reports. So not just when you’re doing a time comparison, but we’re looking at the last three full weeks for example, maybe what we want to do is create a date range that’s the previous three weeks to these last three full weeks. The way that we can do that, is that we can create a new date range, and actually I’m going to take a look at what our dates were, so last three weeks is going to be our start. Let’s actually make that easier, let’s just say yesterday because we’re limited on time here, and so we’re going to say two days ago.

And now you can actually drill in to see what days are included, so two days from today. If I want to create a new date range, then maybe I want to compare that to three days ago. And I’m going to say: “Use rolling dates. Apply.” And I give it the title of “3 Days Ago (custom for ObservePoint lovers).” Hit save. Now I head on over here, I can now do the side-by-side of two days ago and three days ago. Really, really slick. What you saw me do earlier is the ability to drill in by day and just confirm, “Okay, everything looks like I expected it to.” Can be a really good what to save some time for you.

Some other fun stuff that I’ve got to pull in for you is some other new things that we have available. And what I’ve got is the ability to use a segment as a dimension. I talked about this recently. What you can do is start with a brand new freeform table. We can take a segment, say device type equals desktop, drag that in over here, and now we’ve got our segment is device type. We can see the total number of occurrences, I can pull in a new metric, maybe visits, and let’s say I want to compare that to device type equals tablet. The way I can do that, is I can actually drop that right on top of segments here and I can see that comparison, one on top of the other. That’s a lot easier than some of the other ways of doing comparisons. It’s a nice little way to see visits broken up by desktop or tablet.

Something else that I’ve got going on within this set of projects here is we have segment comparison. If you haven’t played around with it yet, please do. It’s been one of our fastest used new features within Analysis Workspace ever. What it allows you to do is really quickly dig into two different segments and figure out where there’s overlap, where there are metrics that are distinguishing them, where there are dimensions that are distinguishing them, and where are other segments that are distinguishing them. Really, really powerful and a great way to dig a little further into your data.

The simple way to create them is by heading over to the panels area within Analysis Workspace. You want to drag in a new one and all you need to do is find the segment you want to compare. We’ve got desktop, and what it’s automatically going to do is say: here is desktop, compare it to everyone else. If we wanted to, we could manually say desktop versus tablet, hit run, and now it will create a report very similar to the one I’ve put up with all these different reports built within it.

Keep in mind, within Analysis Workspace, you have the ability to interact with everything that’s populated within the segment comparison. Nothing is static. If I wanted to dig further into any of these metrics or any of these domains, these are live tables. So if I wanted to dig a little further into, let’s say, Referring Domain, I don’t have to worry about duplicating this and maybe creating a new project or anything like that. I can do my analysis right here. Maybe we want to see Visit Number and drill that into Referring Domain, and I’m able to do that right within the tables that’s build by segment comparison. Good little time saver.

Something else that you’ll notice within the tool is we’ve really tried to focus on making segments available within the product everywhere.

Demo Screen 5:

Let’s head back over to Marketing Channels. If I’m back over here and I want to focus on the last touch channels or referring domains, right click and now, not only can you run that in segment comparison with one click, you can also create a segment from that selection. That’s something that I maybe would suggest everyone do as you’re learning Analysis Workspace and playing around with it—always look for opportunities to right click. There’s always something interesting to be done there. A good example of that is with these creating segments and running it within segment comparison.

Demo Screen 6:

But, it’s also really quite powerful when we’re doing some flow reporting. Other really recent features put into Workspace are: Fallout and Flow reports. Those are based on on paths. Really, really cool feature and you can integrate Evars and Props and events, metrics and dimensions too, as much as you like. So let’s start with a flow here, if we pull in flow, we can drag and drop anything we want. If we want to start with the very first visit number, we can drag in a value of a dimension. Then we can say that from that visit number we want to identify a specific page—user went to the news page. Then from there, we want to see that a user had a particular custom link—they zoomed. And it looks like I went a little crazy with my dragging and dropping. If you do that, don’t worry, you can always either undo or just click to close. Drag and drop your zoom change and now we can see we’re starting with all visits, now we’re limiting it to only first visits, then we’re saying they accessed the news page, and then they changed the zoom on that page.

It’s a good way to understand what your particular funnel is. Maybe yours is going to be a little more interesting around carts or checkouts. Whatever you’re doing, you’ve always got the ability to do this. And note that I’m mixed and matched—I started with visitor number, then I went to the page name variable, then I went to a custom link. If I wanted to, I also have the ability to deploy any of my metrics as well. If I wanted to focus in on, let’s see if we can pull in the fact that someone clicked the copyright date, there we go. We can see that someone was served a copyright date. Lots of good ways to break up your data here.

I was talking earlier about right clicking. This is another place that we’ve really integrated the concept of right clicking. If you right click anywhere within a fall out, you can choose to fall through this touchpoint, so you can break it down so that you’re focusing on the green section. And if I click breakdown, fall through at this touchpoint, now it’s going to focus on: someone did an all visits, then visit number, then they went to the news page. And you can see that broken down over here. If I instead did the same thing and did a fall out at this touch point, the difference is that users did not get to that news page. That’s going to be the difference between a fall through and a fall out.

Some other fun stuff that’s built right in here, again, you’ve got that right click to get to get to create segment from touchpoint. Another really valuable thing, as well as trending in the touchpoints, too, so you can see over time: how is my fallout trending?

Something that I want to focus on is the fact that not only can you mix and match within Fallout, you can actually do that within the Flow as well. This is may be one of the lesser known features within this recently released flow component. Maybe we’re just used to dragging in page, and very quickly we can see what the top page is, so this is the homepage, and I can drill in and I can say I went to the news page, then went to the leadership page, etc. that can be interesting, but it’s not doing a doing a full mix and match of all the power within Analysis Workspace. Maybe what I want to know is: how are users getting from the homepage to the news page. I can do that because I have things tagged in a custom link here. Now check this out—I can actually collapse entire column, then I want to start with SAR homepage, and I should be able to simply drag in my custom link. You’ll see when I hover over this column, I can actually either replace that column with an additional dimension or I can append to it.

So I can see how users are getting from one page to the next. I can see that from this page, going through the homepage, I can see exactly what users are doing in terms of my custom links. Once again, I can focus on starting on home, going to news, and maybe I want to see how users are interacting with the site after that. Once again, the difference here is I can collapse, head on over to home, and instead of using this page, I can go to my custom link to see what users are doing immediately after. It looks like they’re changing the zoom. Maybe from there, I want to click a little further, to see what users are doing after they zoom. It’s going to remember you were at page and then custom link, but I want to go back to page. All of these fun things are available within the tool and you may not have even known them.

Demo Screen 7:

With that, I know that we’re getting low on time, so there’s one thing that I want to make sure you’re aware of—there’s some really great sessions coming up at one o’clock. Make sure to hit up any of these three. All of them look really good. I appreciate you joining me, I look forward to the questions. Find me on Twitter and find me on the measure Slack. Talk to you soon.

Previous Video
Jason Call, ObservePoint - Dynamic Dashboards: Make Your Data Work for You
Jason Call, ObservePoint - Dynamic Dashboards: Make Your Data Work for You

This session is a tactical deep-dive with a few quick tricks that will show you how to build a dynamic anal...

Next Video
Michele Kiss, Analytics Demystified - Pairing Analytics with Qualitative Methods to Understand the WHY
Michele Kiss, Analytics Demystified - Pairing Analytics with Qualitative Methods to Understand the WHY

Learn quantitative and qualitative techniques companies can leverage to get more insight into their custome...