Device and Client Counting Methods in Marketo

Until Email Insights was released in June, (and you may still not have it), Marketo provided scanty device and email client data on your leads. This data can be helpful when making decisions on how much tweaking to put into email or landing page code. Ideally, our stunning visuals work across all devices, OSs, and browsers. Of course, we know this isn’t true (ahem Outlook). So if you knew your audience breakdown by these metrics, you might put a lot less work into Outlook 120dpi or that Blackberry no one is using.

Marketo on its own does not produce a report displaying the types of browsers, email viewers, devices, or operating systems used by Leads. Marketo does collect this information, and it is possible to extract it in a directional way for deciding on layouts or even offers. Even ESPs won’t be 100% accurate on these reports.

The good news is that there are two ways to obtain this data.

• Email Insights
• Smart Lists

Marketo collects the following data:

• Device – available in smart lists and Email Insights.
• Browser – available in smart lists.
• OS – available in smart lists and Email Insights.

What you still won’t get is a perfect view of your audience’s device preferences. You also won’t be able to test email or page designs against each possibility out there. For that, use Litmus or EmailOnAcid. If you really need Browser data, Google Analytics or other tools will be much more accurate. What they won’t do, though, is tie that data back to the Lead.

Using Email Insights

This is your first, best option, if you have it on your instance. The challenge, however, is that it may not have much data for you. Device data is there for the data that is loaded and can be calculated across dimensions. Here’s a quick peek at the menu selection:

When I’ve tested this feature in real life, I found that I had much better control and data using the smart lists. Also, this feature doesn’t really tell us the percentage of the total database that uses the platform. So if you did a 30 day window on this report and were fairly confident you emailed most of the database, then you could make a reliable extrapolation.

Using Smart Lists

The email data you seek is also available if you are willing to spend a couple of hours building out the smart lists.

Filters and Constraints you will use:

• Clicked Link in Email
• Opened Email

Group by Device, OS, Desktop vs. Mobile, Viewer.

There a myriad of devices, operating systems, and form factors. The best thing to do here is group them into large categories that can be analyzed more efficiently with a smart list. Here is a list I built that should save you time [XLSX]. This list could be different for your audience, or as new devices and systems come out, so you may need to update this list periodically.

Once you have these lists created, you will build a Program with a set of smart lists like this:

When you create your smart lists, consider the implications of the smart list logic.

• Method 1: Clicked Link in Email in Past X Days
  • Only looks at the most active leads.
  • Most reliable data.
  • More than 7 days of data could take a really long time.
  • You will need to run this once for Platform, Device, and OS.
• Method 2: Opened OR Clicked
  • This is a wider range of data and would likely help you understand mobile engagement (many won’t click).
  • Will depend on people downloading images.
  • Should be viewed as Directional.

It is best to use a Date constraint like IN PAST 7 days, otherwise you can end up waiting forever for the system to count the Lead. You can try 14 or 30 days as well if your database is smaller. Besides, we want recent history, not what happened a year ago with viewers that no longer matter.

There is a caveat with this method. This smart list will pull leads that may have hit your email or page from multiple browsers or clients. Each Open or Click or View will be recorded separately and may cause overlap between smart lists. Keep this in mind because this method will not give you a 100% unique list. This is directional.

The reason to do this at all is to better understand how to serve your audience. It is easy to make assumptions about the audience with general stats on the use of Internet Explorer, Chrome, Apple Mail, etc. If you are selling to other marketers, there’s a good chance iPhone rates are higher than average, while IT buyers often use a combination of Android, iOS, and Outlook 20XX on Windows. Get some benchmarks here:

• Email Monday Stats
• Litmus Stats

There’s no sense in wasting five days getting an email or page to look perfect on every device and viewer; it’s just not possible. But if you can get it perfect on the top five devices and viewers that cover 90% of your audience, then this smart list technique will help you stay focused on what matters for your leads.

Note – in the image above, I show a set of batch campaigns that could use these smart lists to Add to List various leads every X days. You will then need to clear those lists every X days to keep the data clean. This system may be a giant waste of processing time, however, it may run faster than the smart lists.

Interpreting the Statistics

Please be careful in interpreting the data. For example:

In this report, can can say that most people are using Windows desktops. About 30% appear to be opening and clicking the emails from an Apple Mobile device (iPhone or iPad). Very few people in our audience use Blackberry. (these numbers are totally made up).

If instead we used Opened only or Clicked only, we might find completely different data. We might find nearly everyone clicks from a desktop, but opens or reads email on a mobile. If your firm is local dependent or you like to send email at 6am, this data will make sense. People will commute to work and read the email on the train, but wait to click at work.

Again, this only tells us about the actively engaged email addresses, not the entire database. You can infer this, however, I would not totally bet on it. The good news is that if you use Litmus and are trying to get Lotus Notes 10 to work, you can safely ignore that group of people.

Web Client and Browsing Data by Lead

Marketo does collect a fair amount of browser and URL data, but this data is scattered through the system. I would recommend using Google Analytics or another package to better understand your audience by Device. However, Marketo’s stats are tied directly to real email addresses, including Prospects and Customers. If what you are selling relies a lot on type of  site visitors by device, browser, etc…then definitely break out your database in this way.

Filters and Constraints to Use:

• Clicked Link on Web Page
• Visited Web Page
• Filled Out Form > Querystring
• Original Search Engine
• Original Search Phrase

Other Resources

• Litmus
• EmailOnAcid
• Marketo Deliverability Pack (250ok)
• Marketo Email Insights docs