Better Product Analytics With Keen IO

By Jason Thompson, co-founder, 33 Sticks   If this is your first time hearing about Keen IO, here is a quick overview.   Keen IO is an amazingly powerful and flexible analytics platform. No, not the traditional analytics platform you are used to, like Adobe Marketing Cloud or Google Analytics, but a robust framework for building your own analytics product.     Three Steps to Utilizing Keen IO Within Your Product or Application   Data Capture   The Keen IO API offers a broad range of languages you can utilize to capture event data from your application. From signups to errors to a user pushing the ‘on’ button, Keen IO captures event data and stores it in JSON format, automatically capturing any new events and properties as you define them. You might find yourself wondering, what kind of applications can I embed Keen IO in? Great question, Keen IO can be implemented into anything from a SaaS application to a smart watch.   As a simple demonstration, I’ve implemented Keen IO into the 33 Sticks website to track the effectiveness of our navigational elements.   When a visitor clicks on a navigational element, a small block of code is executed to capture the event: //Create a JavaScript object of N number of properties I can define var menuProperties = { menu: "About", action: "Click" };   // Add the event to the "linkClick" collection Keen.addEvent("linkClick", menuProperties);     Data Analysis   The Keen IO Workbench is the analysts best friend. Using the Workbench, you can quickly design queries that will best tell the story of the data you are...

Building A Successful Analytics Organization

We are often asked, “what is the best way to create an analytics organization within my company?” While there is no one right answer, we have had the opportunity to observe many highly successful organizations and have identified a few keys to building and running a successful analytics organization. Let’s start with the reporting structure. Who reports to whom? Who sets the strategy? We have observed firsthand the struggle organizations go through when trying to figure out where analytics should live within the organization — IT, Marketing, Finance, Operations, Office of the CEO. We believe that the analytics organization should report up through a “neutral party,” this avoids as much as possible the political infighting that happens when insights generated by the analytics organization reflect poorly on the performance of team leadership. Having the analytics organization report directly to the COO or CEO, can help the organization properly prioritize the vision of the business as well as be protected and empowered to deliver true insights for all business units, even if the insights reported show negative results. When is comes to structuring the team itself, there are four key roles that are critical to success. Depending on the company, one person could potentially play all four roles or there may be multiple people playing a single role. What is important is that each of the skillset is properly represented in the team.   Director of Analytics This role should report directly into the COO or CEO. The role is multi-focused, serving as both subject matter expert and mentor for the team. The director would handle all of the operational...

Tag Management and the Myth of the Magic Tag

By Jason Thompson, co-founder, 33 Sticks   In the early 2000’s, analytics vendors fighting for marketshare began an aggressive marketing campaign featuring the easy to implement, high value, “Magic Tag.”   The “Magic Tag” was perfect. The IT guy only had to deploy a simple block of code once and never had to deal with the needy marketeer again. The marketeer would instantly have access to all of the correct data needed to get the job done faster and better than his competitors.   The problem was, no matter how much effort the vendors put into maintaining the illusion of the existence of a magical tag, it was simply that, an illusion. A myth.   10 years later, it is fantastic to see the analytics vendors doing such a great job properly positioning tag management, not as a “Magic Tag” but as a powerful tool that, if used properly, can create amazing efficiencies, improve data quality, and ultimately drive higher revenues. However, even with vendors properly positioning their solutions, many businesses continue to hear what they want to hear, “put this magical tag on your site and you don’t have to worry about anything ever again.”   Having advised hundreds of businesses, from emerging startups to massive global corporations, we’ve found that those who have been most successful at using analytics as a competitive advantage have looked at data collection as a pathway to analysis, insight, and optimization. This may seem intuitive, however the majority of companies we’ve worked with initially viewed data collection as a highly technical task, often excluding the analysts and business users entirely from the...

Adobe DTM: Leveraging JavaScript Variables

By Jason Thompson, co-founder, 33 Sticks The ability of your tag management solution (TMS) to utilize existing site data can be extremely powerful. Existing site data can come in many different forms but in this post, we will focus on using JavaScript variables. The Scenario The Website Manager wants to get a better understanding of how visitors are using the new internal search feature that was launched last week. Unfortunately, when the new internal search feature was being developed, there wasn’t an analytics requirement attached to the product spec so now there is no specific tracking of this new feature (I know, this never happens, just pretend that it might). One of the members of the development team mentioned that the content management system (CMS) does send some data, stored in JavaScript variables, about how the search feature is being used and asks the Analytics Manager if Adobe Digital Tag Management (DTM) could use those variables to populate the reports that the Website Manager is asking for. The Solution The Developer provided the Analytics Manager three variables that could be useful in solving this problem: newSearch – This indicates the execution of a search count – This indicates the number of search results returned searchItem – This indicates the search term   Step 1: Create New Data Elements to Capture JavaScript Values NOTE: I prefer to use Data Elements, rather than referencing the JavaScript variables directly inside of Rules, as it reduces the probability of user error and also centralizes the management of each variable. To create a new Data Element, navigate to Rules>Data Elements. Create a separate Data...