It’s late May, and baseball is on my mind. So, how was last night’s release of Adobe Analytics Standard like a baseball team? Just as there are nine batters in a starting lineup, there are nine categories of feature updates, each which brings its own new benefits to the broader team. In this post, I’ll cover those nine categories. My colleagues in Adobe Analytics Product Management will be following up with more detailed posts covering much of this new functionality (and when those posts go live, I will also link to them from here).
All of these features are now currently available within Adobe Analytics, as of 23 May 2013. This is going to be a long one, so buckle up. Here we go.
Current (Low-Latency) Data Integration
Leadoff hitters are usually great at getting on base and they’re fast enough to steal bases and score from first base on a double. This “leadoff hitter” introduces the concept of “current data” (low-latency data streamed into marketing reports), a quick and powerful way to check trends on your site in real-time.
You may be aware that we added a set of Current Data reports to the SiteCatalyst 15 interface late last year. These reports offered traffic data within 3-5 minutes of collection and conversion data within 20-25 minutes, putting it at parity with the real-time reporting offered in previous versions of SiteCatalyst. With this release, that separate group of reports goes away, and is integrated into most reports throughout the tool. You and your users will now have the option of including current data at any time.
If you click “Yes” to include current data, if your date range includes the current day, you should immediately see data for just the past few minutes. This is a great way to check the success of a campaign or a piece of content very soon after launch. It should also help you test implementations, ensure that key pages and measurement are always available, and more. If you refresh a report while including current data, you’re likely to see more data added on the fly. Additionally, you can mouse over the clock icon at any time to view the actual latency for each metric in your report, so you’ll know exactly what you’re looking at.
A few things to note in connection with current data:
- Some metrics and reports do not support current data. If you enter a report which does not support current data, you will not see the option to include current data. If you use a metric which does not support current data, you will be given the most recent data that is available, from the most recent batch upload.
- Segmentation, correlations, and subrelations do not support current data. If you apply one of these capabilities to a report while viewing current data, you will automatically be toggled to exclude current data. Of course, this only matters if your date range includes the current day.
- As an admin, if you don’t want to use current data, you can remove yourself (and others) from the Current Data Users group in the Admin Console
Current data is also available via our Reporting API and in ReportBuilder.
- Smaller: new version is 21k uncompressed & 8k gzipped (whereas our previous code version, H.25, is 33k uncompressed & 13k gzipped)
- Faster: 3-7x faster than the H.25 code
- Single JS: new version is light/fast enough to be used with mobile sites, but robust enough to be used on the full desktop web, allowing you to leverage a single JS file across all web measurement environments
- Native Utilities: native utilities providing product-level support for several common plugin use cases (getQueryParam, linkHandler, cookies, urlencoding)
- Head Tag Deployment: deploying s_code and page code inside <head></head> tags is supported
My colleague Ed Hewett blogged about this in greater depth here.
Classification Rule Builder
With our number three hitter, we start to shift toward power—guys who can hit it out of the park at any time, always ready to go deep. You’ve probably know for quite a while that classifications—the ability to add meta-data to the dimensions you’re collecting—is one of the most popular features in expanding the value of your data set, but it’s about to get a whole lot more powerful (even without using steroids!). Those of you who were at Adobe Summit, either in Salt Lake City or London, saw this exciting new enhancement on the main stage during our “Sneaks” session. It’s also the top ranked idea in our Idea Exchange: rule-based classifications.
In the past, building out classification tables and then uploading them into Adobe Analytics was an onerous task requiring a lot of manual labor or at the very least some nifty Excel wizardry. And that was if you remembered to keep your classification tables updated.
This feature introduces the ability for Adobe Analytics to examine your “key” values (the data collected from your site or app in the variables that you are classifying) and fill out the “table” automatically and apply the classifications based on what it finds in those key values. As such, it works best when you have a reliable schema for key values. For a very simplistic example, let’s say you have the following campaign tracking codes coming in from your site:
Clearly, in these two tracking codes we can see that we’ve got components of year, season, product category, and channel embedded in there. Using the classification rule builder, you can use regular expressions to train Adobe Analytics to automatically create classifications for these and other tracking codes, so that ultimately you get something like the following:
|Key||Season||Product Category||Campaign Channel|
|2013:spring:womens:search||Spring 2013||Women’s Apparel||Paid Search|
|2013:spring:mens:display||Spring 2013||Men’s Apparel||Display Ads|
And then this gets applied quickly and automatically so that your data is always up to date with the latest classifications critical to the analysis and reporting you’re doing for your business.
My colleague Matt Freestone has written a short blog series on how to use the classification rule builder, so be sure to check those out soon.
Sequential Segmentation (Discover)
I’m batting sequential segmentation fourth because it’s extremely powerful and also one of my favorite things we’ve done for analysts in recent memory. The segment builder in Discover now allows you to define a sequence of visitor interactions.
In the screen shot above, notice that I am defining a very specific sequence: Visitors who saw the “Cabana” product page, and then no less than one week but no more than four weeks later, they viewed the product again and made an order. Never before has it been so easy to delve deeply into the sales cycle, or understand how visitors interact with key site workflows over a period of time.
There is practically no limit to the kinds of interesting questions and valuable segments that you can build with this functionality. One of my favorite uses for this powerful feature is that it can make cohort analysis a snap (provided that you’re capturing the date of the event or events that demarcate your groups into a variable, similar to what Adam Greco describes here).
It’s also worth noting that, while you can’t create sequential segments in SiteCatalyst 15 yet, the sequential segments that you build in Discover can be applied in SiteCatalyst today.
Up fifth, it’s an addition to SiteCatalyst 15 that I’ve been excited about for weeks and weeks. In Ranked reports, you can now compare two segments side by side.
To do this, in a Ranked report, click on “None” next to “Compare to Segment” in the report header. This works both with a segment previously applied, as in the screen shot below, or if you haven’t yet applied a segment, in which case the segment that you choose will be compared to all visits (unsegmented data).
At the moment, you can either apply segment comparison or apply date comparison, and you are limited to two segments at a time. Discover allows you to do combined segment/date comparison and do multiple segment comparisons already, and of course we’ve got more great stuff planned for segments in SiteCatalyst 15 as well.
Better Report Suite Settings Control
Our sixth hitter can do a little bit of everything: it can enable pathing, it can turn on participation, and it makes sure that all of your desired report suite settings are turned on automatically whenever you create a new report suite. This improvement primarily includes two elements:
- When you create a new report suite based on an existing report suite’s settings, more of those settings are now copied to the new suite, including GeoSegmentation, multi-byte character support, and Data Warehouse. You don’t need to do anything in order to copy these additional settings upon report suite creation; the Admin Console will take care of this automatically.
- Within the “Edit Settings” menu in the Admin Console’s Report Suite Settings page, there are several new settings that you can control as an admin for your report suites: Pathing, Participation, Merchandising, Event Serialization, GeoSegmentation, and Transaction ID Storage. These settings are only available on report suites that have been upgraded to SC 15.
While you’ve been able to get these things turned on for quite some time, in the past, enabling these features (either for existing report suites or for new report suites) required a phone call to ClientCare or your Account Manager. By putting this control into your hands, we hope to save you time and make it easier for you to enable features as you need them.
It’s also worth noting that pathing and participation, once enabled, will provide you with data going back to your SiteCatalyst 15 upgrade date. So if you upgraded on 1 Feb 2013, and you turn on pathing on a traffic variable today, you will instantly get data going back to 1 Feb 2013 in your pathing reports and metrics.
Classifications on List Variables
Up seventh, one of my favorite capabilities of SiteCatalyst 15 is list variables. These special variables allow you to capture a list of values so you can easily measure things like guided search usage or on-site ad impressions—things that might have a many-to-one relationship to the page. I’m pleased to report that you can now add meta-data classifications to these variable values. This extends the value that you can get out of your list variable data by allowing you to group individual values from your list into ad groups/types, guided search elements, etc.
To set up classifications on list variables, go to Edit Settings > Conversion > Conversion Classifications in the Admin Console, then classify using the Classification menu under Admin to upload or set up rules to govern your list variable classifications.
Improved search filtering
In the eight spot, we’ve got what may seem like a minor change to the way you do reporting and analysis: we have improved the way that you filter report values. In the past, certain characters, such as the pipe (“|”) and the dash (“-“) were treated as special characters. In the case of these two, the pipe was treated as an “OR” in your search, and the dash was treated as a “NOT.” This could lead to some issues with your filter results if you used these characters in your variable values (for example, as a delimiter, as in “products|mens|shoes”).
These characters are now treated literally, and you are given a number of additional options to help you design the search that you want:
In particular, note the introduction of “Starts With” and “Ends With” criteria, neither of which had been possible previously. And the introduction of “Advanced (Special Characters)” allows you to use the previous filter method if you’d like, or if you have filtering needs not satisfied by the new criteria and treatment of special characters.
A final note on this: If you have bookmarks and dashboards that include search filters, don’t worry; we’ve taken care to ensure that they won’t be impacted.
Hourly trended metrics
Batting ninth is a nice little feature which plays good defense. . . okay, so maybe this metaphor is breaking down. But the feature allows you to now view hourly granularity for all metrics in SiteCatalyst 15. In the past, metrics such as Visits had a limit of daily granularity, meaning that you could not see how trends change from hour to hour in your data. That limitation has been removed, giving you greater intra-day and day-over-day/week-over-week visibility into trends across dozens of metrics and KPIs.
To select hourly granularity, simply use the granularity drop-down menu on the calendar module (or in the report header) on any trended report.
I hope you can tell that I’m even more excited than usual about this release. I’d love to hear your feedback, so please share in comments and tweets how you’re using these features or how you’d like to see us improve them.
I’ll be updating this blog post with links to my colleagues’ more detailed posts on many of these improvements, so stay tuned for a lot more about how best to take advantage of these things.