Data Analytics

Highlight: Want to see a full analysis?  See an example here, or the analysis behind the graph below.

I love exploring, analyzing and explaining complex systems.  My goal is to master each step of this iterative process, with a particular emphasis on understanding business systems. I’ve built a particular specialty, leveraging my user research experience, in user-focused analytics.  More recently, I’ve focused on Data Acquisition and Exploration through SQL and visualization.

User-focused Analytics

In my work with Stever Robbins, I was given end-to-end responsibility for the business’s data projects.
This meant that, in addition to the usual challenges of analysis, I defined questions, designed acquisition mechanisms, and connected the results to our strategic goals. The first thing that I learned is how many of the tantalizing array of metrics that online systems like Google Analytics offer are useless or worse. Testing and calibrating my tools, just like a physicist in their lab, was a vital first step to taking measurements.

I also learned to go the extra mile to obtain a broader understanding. I’d take support calls, and chat about the customer’s situation and aspirations. For my own startup, I’d even link to competitor’s blog posts and products on social media (mixed in with promotions whose standard response rate was well-characterized, as a baseline) to get a sense of which themes resonated with our followers.

The ultimate goal, of course, was to slowly and steadily optimze our business based on empirical data. For Stever, that meant understanding what our customers were looking for. For another client that he recommended me to, it meant segmenting her followers into two sharply distinct user groups. Regardless of the specific goal, the same process of iterative acquisition, analysis and presentation will always apply.

 

Data Acquisition and Exploration

Much of my experience has been in startups, where I’ve found the opportunity to work on problems that I’m passionate about while rapidly learning and expanding my responsibilities.  In my most recent freelance data analytics project, I’m working directly for the Director of Online Marketing at a 6 year old company to improve forecasting.  Whereas designing effective data collection was paramount for many of my previous projects, much of the understanding of this company’s data set has come from asking the right questions within the organization.

Exploring the data set is also a different challenge, particularly when it’s spread across hundreds of PostgreSQL tables (and dozens of employees’ memories!).  Using exploratory queries and iterating on preliminary results, I was able to build an understanding of the business case, analyze the relevant factors, and design an effective presentation to guide their C-level decision making.

Summary

  • Talked to users, marketers and engineers to understand the system and business goals
  • Explored the Postgres database, Salesforce account, Google Analytics and Client Website to find relevant data sources
  • Ran SQL queries to gather data, used Python to parse it and used matplotlib (Python’s answer to MATLAB) for visualization
  • Communicated approach and results to relevant stakeholders, along with implications for their strategic decisions

 

“He wrote sales pages and sales emails that got higher response rates than the ones written by our marketing partners. He built and analyzed marketing analytics. He handled all customer-facing communication, fixed problems and built systems so they wouldn’t recur. He documented everything he did, then interviewed, hired, and trained a virtual assistant to operate those systems on a day-by-day basis when his time with me was complete.”

    – Stever Robbins (see full recommendation on LinkedIn)