ÃÛ¶¹ÊÓƵ

Transcript

In this video, I’ll review conversion variables in ÃÛ¶¹ÊÓƵ Analytics. These are the topics I’ll cover. I’ll start with an overview, discuss characteristics, the different types, how long they persist, the metrics that can be associated with them, and common use cases. To get started, let’s look at a multi-day weather forecast. The left column contains the days, and to the right several of the columns contain numbers associated with each day, like the high and low temperature, the chance of precipitation as a percentage, and the strength of the wind in miles per hour. These columns are referred to as metrics. They measure the different weather-related characteristics for each day in the forecast, so metrics are numbers. The metrics apply to dimensions. In this case, the days in the forecast are the dimensions. There are two types of dimensions in Analytics. This video covers conversion variables, which is one type. The other type is traffic variables, which is covered in a different video. As I review the characteristics, keep in mind that some of the details relate to implementation, while others relate to reporting. First, conversion variables can store up to 255 bytes. This isn’t necessarily characters, as some languages use multi-byte encoding. Anything passed in over 255 bytes gets truncated in reports. These can accept string values or counter values. Counter EVARs can be used for things like counting the number of internal searches made before a purchase. There’s a bit of overlap here with regards to conversion events, but just understand this concept for a counter EVAR for now. You can apply metadata to base values passed into conversion variables. This metadata is referred to as classifications in Analytics. They can be configured with custom expiration settings, which means they can persist beyond a hit or page view. A hit is a set of data sent into Analytics with an interaction on your digital property. I’ll get to this in more depth coming up. There is signed credit for success events that happen on your digital properties. For example, which search terms were used for an internal search? Which products viewed or eventually purchased? Which type of loan is a consumer applying for? Again, I’ll be covering this more soon. In reports, you can break down one conversion variable by another. It doesn’t matter if the values are sent in in the same hit or not. This capability is configured in the Admin Council. Now, values passed in aren’t case sensitive. The reports will reflect the first value passed into Analytics. So, for example, if I pass in lowercase necklace to EVAR1 the first time that variable collects a value and then subsequently pass in necklace in title case, the report will show only the lowercase value and metrics will aggregate against that value. On the implementation side, conversion variables are referred to as EVARs or S.EVARs. There are a few types of conversion variables. First, there’s a set of reserved or pre-defined ones. They’re used for things like marketing tracking codes, e-commerce products, marketing channels like page search and social media, and the state and zip code tied to conversions like purchases or enrollments. Second, list vars accept multiple delimited values in the same hit. Up to three list vars can be configured. Third, there are configurable custom conversion variables you can use according to your business requirements. These are labeled and configured in the Admin Council. Fourth, any of the custom conversion variables can be configured for counter support. Last, an EVAR can be configured as a merchandising variable. This is commonly used in e-commerce flows. For example, if you want to understand how different products purchased were discovered like in using internal search or linking from an internal promotion, you can tie this value to the product using merchandising syntax. Now I’ll cover persistence in more detail. I mentioned earlier that conversion variable values can persist beyond the hit a value was passed in with. Let’s assume that this collection of six EVARs are configured for visit expiration and the most recent value set in each EVAR will be given credit for success events that occur during that visit. Now during the fourth page view, the sign-in conversion event happens. The headwear value given event credit is no hat since that was the last value passed into that variable during page view three before the sign-in occurs. The pants value given event credit is shorts even though long pants is passed in later during that visit. It happens after the sign-in event. The same is true for the accessories EVAR. A specific value isn’t set until after sign-in. Conversion variables are designed to segment conversion metrics in custom analytics reports. Metrics like product views, card additions, orders, and revenue, as well as any other key downstream event metrics apply to conversion variables. This makes them more useful for allocating towards business requirements compared with traffic variables. If you haven’t already watched the introduction to conversion events video, I suggest you do so. There is a key correlation between conversion variables and conversion events. Alright, let’s review some use cases. This is an example of a report in Analysis Workspace that uses the reserved marketing channel conversion variable. This shows different allocation models, Last Touch, First Touch, and Linear for marketing channels like Referring Domains, Display, Social Campaigns, Page Search, and more for the metric online orders during the last 30 days. It segments the total online orders for this period of time by the various marketing channel values. This is a basic retail commerce report that segments online orders and online revenue by the product subcategories for a website’s fashion category. This also shows the percentage of each metric for each subcategory value, which is very useful to understand. Another common use case is understanding how many variants of a call to action button on a specific page and template receive the most clicks. This can also be related to other downstream events post CTA click like registrations or form completions. Another powerful conversion report comes from using list variables. This variable is useful for analyzing things like which banners were displayed and clicked and which received other key downstream event credit like event registrations in the same visit. This example involves using two list variables because two distinct expiration models are needed to achieve this type of report. In my example, SLList1 is set with the pipe delimited set of values for the impressions event. This is because multiple banners can randomly display on the page. These values expire after the impressions event. Notice that the list variables split out each delimited value into a single row in reports. Now when the banner is clicked, the name of the banner is passed into SLList2. This list variable expires after the visit. So if event registrations occur downstream, the banner clicked gets the attribution. In the case of banner J, one of the clicks didn’t result in an event registration. This wraps up the introduction to conversion variables. Hopefully you have a good foundational understanding of this type of variable and how to use them. Good luck!