Let’s Get Real (Time): Using the New Real-Time Reports in Adobe Analytics
It’s hard to have a digital marketing conversation these days that doesn’t touch on real-time analytics. Whether you’re for or against the proliferation of real-time data (and there are many who fall into each camp), the increasing speed of marketing is a reality that we see at every turn. On-site personalization, remarketing campaigns, and content optimization are just a few examples of emergent tactics that require data to be made available in new and faster ways. While most data-informed decisions should still take days or sometimes weeks to make, there are increasing opportunities for marketers and others based on live data that can make a significant impact on a business.
This is why, in our 17 October 2013 release of Adobe Analytics 1.4, we are introducing a set of true real-time reports—low-latency (a few seconds!), high granularity, and constant streaming—in the Adobe Analytics reports & analytics (formerly SiteCatalyst) UI. In this post, I will describe the report at a high level and share some insights we’ve gained over the past few months of beta testing so that you can take advantage of this new set of reports in your own marketing efforts.
What Can I Do with Real-Time Reports in Adobe Analytics?
As described above, we define real-time reporting as:
- Current: data latency in the real-time reports is as low as a few seconds. There is almost no lag between data collection and data availability in these reports, and you can choose from among hundreds of variables and events to view in real-time.
- Granular: while most reports in Adobe Analytics are limited to hourly granularity, the real-time reports offer minute granularity so that you can identify and respond to trends much more quickly than in traditional reports in the tool.
- Alive: the real-time report streams data into it as that data is collected. You should never need to refresh your browser to get new data when viewing the real-time report. New data becomes available automatically every 3-5 seconds, and visual animations make it clear how your data is changing.
Each report suite in Adobe Analytics is automatically enabled with access to three real-time reports. These reports are easily created, configured, and saved by any admin user, and then become available to all users in your company. Each report requires one metric and three dimensions, typically based on your existing implementation. Available dimensions include all traffic variables, Pages, Site Sections, Referring Domains, Tracking Code, Products, and more; available metrics include all custom events and shopping cart events, Revenue, Orders, Units, and Instances/Page Views. This means that most users can “plug and play” without making any tagging changes at all.
One of my favorite benefits of the real-time reports is that they explain “why something is happening” through breakdowns. In the screen shot above, clicking on any product name will show referring domain and GeoSegmentation U.S. DMA for that product. Publishers might report on Article Names, Site Sections, and Referring Domains so editors can see site sections and referring domains by article; this will inform the tactic they use to maximize upward trending articles.
I need to share one more of my “favorite features” of the real-time reports: if you click on the icon in the upper left corner of the report, you will enter our native full-screen mode. Early users of this tool reported wanting to display it on a big-screen TV in the newsroom or the marketing department. We made it really easy for you to do so, and when you click the icon, the report enters an attractive full-screen view perfect for display on these devices:
Real-Time Best Practices and Use Cases
The editorial use cases are fairly clear: publishers want to make sure that content is relevant, positioned appropriately to maximize traffic and engagement, and promoted in the right way at the right time. The real-time reports in Adobe Analytics allow editors and others see what is popular (and why) but also what is emerging and what is becoming stale. Each reportlet in the real-time report has three modes:
- Most Popular: which pieces of content (or products, keywords, etc.) accrued the most page views (or other metric) over the selected time period. This a raw measure of popularity, where more [metric] equates to a higher ranking on the list.
- Gainers: which pieces of content are increasing in popularity at the highest rate, defined as percentage of growth from the starting point in the time period to the most recent point. There might be a story that isn’t popular enough yet to crack my top 10 over the last 15 minutes, but at the rate it’s increasing in popularity, I can tell it’s about to blow up.
- Losers: which pieces of content are decreasing in popularity at the highest rate, defined the same way as Gainers. I always want my content to be fresh; when a story starts to lose its momentum, it’s time to get it off the home page. This view allows me to do that.
The reports are also filterable. If I’m the editor of the sports section on my site, I only want to see and drill into the popularity of sports stories. I can filter the report by searching for “sports” if that is part of my page name or article name taxonomy, and then my results will be far more relevant to me.
One great benefit of the real-time reports generally is that, because you have up to three of them, you can create different views for different personae. I might create a generic “Content Real-Time Report” for my editors, but the video team needs something different. I can create a special report just for them that shows Video Name, Site Section, and Video Type, with Video Views as my metric. This will allow both teams to meet their needs using real-time data to make decisions in their own sphere.
As one publisher told me, “Christmas is coming early for our editorial staff.”
When I demoed an early version of the real-time reports at Adobe Summit 2013 (yes, this counts as another “sneak” that has been released!), I used a retailer example. The retailer was made up, but the use case was not. Among the many uses of real-time data for retailers is the ability to quickly identify mis-listed products or runaway coupons that have been picked up by sites like slickdeals.net so that the issue can be rectified before thousands of dollars are lost. This is the classic case of “unknown unknowns,” things that marketers can’t possibly know to look for on their own. Using a real-time report, I can see when a product unexpectedly shoots up the list of popular or fast-gaining sellers on my site, and then break down by referring domain to see what is driving those users to my site to purchase.
In many cases, simply knowing which products are selling in real-time allows a merchandiser to make important decisions that can be the difference between hitting a number and missing it. On Black Friday and/or Cyber Monday, retailers are often glued to their reports, either to provide real-time updates to management on the success or failure of certain promotions or the movement of online inventory. In the past, these marketers needed to know what they were looking for, but with the volume of promotions and deals being offered, and products being purchased, it’s easy to miss an opportunity (or a problem). Not only will the real-time reports keep a retailer up to date on product purchases or response to promotions, but it is great at detecting errors or out-of-stock messages and surfacing those. Simply implement a custom event for either of these two issues, and then select those custom events for use in real-time reports.
I love this quote from a marketer who was an early adopter of the real-time reports: “With the addition of some of the other dimensions it has really helped give it some utility in the marketing department. And having products in there brings our sales team in to be able to use this tool. Before, we were looking at just editorial—now we’re bringing in marketing and sales. That’s great.”
What Else Do I Need to Know?
- API Access: Is there an API already available to all customers so that I can pull this high-granularity real-time data into my own application? There sure is!
- Time Ranges: Four time ranges of real-time data are possible in the report: 15 minutes, 30 minutes, 1 hour, and 2 hours.
- Classifications: In this build of real-time reporting, classifications are not available. However, we anticipate adding them very soon, so that you can report on campaigns rather than tracking codes, and product names rather than SKUs.
- Persistence: One thing to note with the real-time reports is that there is no persistence with the variables that you choose. Conversion variables (eVars) are not available in this phase, and dimensions like Tracking Code, Referring Domain, and Search Keywords do not automatically persist and therefore cannot be tied to downstream conversion in real-time. To use those dimensions, you can “set and persist” them into a Custom Traffic variable, or pass them into a Custom Traffic variable at the time of conversion. All of these dimensions are available with the Instances metric, so traffic trends are easily obtained for any dimension; conversion can require a bit of implementation work. Revenue, orders, etc. on products is available out of the box since those metrics are always set on the same call as the products variable.
- Real-time and Adobe DTM: Adobe Dynamic Tag Management can greatly enhance real-time reporting. For example, I once wanted to track Device Type in the real-time reports. I used DTM’s built-in _satellite.browserInfo object to capture the device type and the operating system in a Custom Traffic variable, which immediately became available to me in the real-time reports.
- Enabling reports: When you first configure the real-time reports, it can take a few minutes for them to start working. However, once they start working they will remain available and current in perpetuity.
- Changing configurations: You can always change the configuration of real-time reports. This means that if you build a real-time report and then you realize you wanted a different metric, or different dimensions, it is easy to swap them out. I like this feature a lot, because you are never locked in to one configuration; it can change as your business needs do. For example, you may need to “commandeer” one of your real-time reports to use metrics and dimensions specific to a live event that you’re sponsoring or conducting. You can set up to use during the event, and then change it back to its previous settings after the fact.
I’m sure I’ll come up with more points that are worth mentioning in connection with the report, but this post is long enough already!
As I mentioned at the beginning of this post, the real-time reports in Adobe Analytics are an early step toward empowering marketers and others to use real-time data to drive real results in their campaigns, their merchandising, their content optimization, and more. You’ll definitely be hearing more about this soon, but for now, check out these real-time reports and enjoy!