The Weekly Blog is curated content from the Evolytics staff, bringing you the most interesting news in data and analysis from around the web. The Evolytics staff has proven experience and expertise in analytics strategy, tagging implementation, data engineering, and data visualization.

BAD DATA COSTS THE U.S. $3 TRILLION PER YEAR

HARVARD BUSINESS REVIEW | ARTICLE

Do you have a “hidden data factory” lurking somewhere in your company?  This article describes the concept, which represents the people, time and effort required to clean up or work around errors in bad data.  The estimate only takes the cost of fixing or working with bad data into account however.  While harder to estimate, there is certainly a risk and an opportunity cost to making decisions on bad data as well.

The first step to minimizing hidden data factories is knowing you have a problem.  The article states, “There is no mystery in reducing the costs of bad data – you have to shine a harsh light on those hidden data factories and reduce them as much as possible.”  At Evolytics, we shine a light on bad data through our comprehensive data audit process.

Much like real factories, we also automate data processes as much as possible to prevent errors.  We use technologies such as tag management and data visualization to reduce human error and increase accuracy in data capture, transformation and reporting.

WHY EVERY MARKETER NEEDS A QUARTERLY FAILURE REPORT

THINK WITH GOOGLE | BLOG

We rarely get a pat on the back for failing, but in the world of AB testing and optimization, failure can teach you just as much, and maybe more than, success.  In this blog post, Casey Carey, the director of Google Analytics Marketing, recommends quarterly “failure reports” in order to share learnings and encourage a culture of “failing – and learning – fast.”

The hard truth is that tests are “tests” because success is not guaranteed.  A certain amount of failure is inevitable, but with the right testing methodology and approach, risks are minimized and learnings are maximized that lead to future growth.  The head of Nest analytics Jesse Nichols admits that his team’s success rate is about 10%, but they “learn something from all [their] tests.”  Even if only 10% of tests “succeed” though, at Evolytics, we also believe that 100% of tests are opportunities to learn, optimize, and keep driving performance for our clients.

THE BIG DEBATE: IS CODING A MUST-HAVE SKILL FOR MARKETERS?

MARKETING WEEK | ARTICLE

Every company nowadays seems to be hunting for professional unicorns.  Those mythical, magical beings that can seemingly do it all.  Such skills sets may include creativity, analytics, marketing, research, advertising, social – and now coding?

This Marketing Week article shares several perspectives on skills that modern-day marketers and analysts need to survive and thrive and how companies should organize around these different types of technical, marketing and analytic talent.  Short answer is that there is no “right” answer.  It doesn’t hurt to hire “T shaped” talent, however, “who have a depth of knowledge in one area and a breadth of experience across a number of others.”

We take the “T shaped” approach at Evolytics, building our teams with two different types of analysts – business and technical.  Business analysts focus on understanding the client’s business, goals and analytic requirements through planning, analysis, testing and insight generation.  Technical analysts focus on collecting and managing data through tagging implementation, javascript coding, data collection, API development and data engineering.  However, all of our business and technical analysts have a broad-based knowledge of the complete data lifecycle, analytics strategy, and technical implementation in order to connect all aspects of analytic work back to client goals and objectives.