Not just for data scientists: 4 strategies for getting your analytics practice off the ground

written by bethburgee on May 3, 2013 in Big Data and Guest Blog and Under The Radar with one Comment
29 Flares Twitter 24 Facebook 2 Google+ 1 LinkedIn 2 Email -- Email to a friend StumbleUpon 0 29 Flares ×

analytics

This is a guest blog post by Michelle Wetzler. Follow her at @michellewetzler

Keen IO is bringing cutting-edge data analytics technology to companies around the worldIn doing so, we’ve talked to hundreds of companies about their analytics projects and pain points. Data analytics can sound daunting—even to data scientists—and one of the most common problems that we hear from companies seeking to gather and analyze data is: where do I begin?

We talk a lot about what happens after a company already has data, but not very much about how to actually get to that point. The following 4 best practices will help any business as they get their data analytics framework off the ground:

1.    Start Small  - Analytics is not a project, it’s a practice. If you’re making a requirements matrix of everything you could possibly track in your business and then amassing a team to build out the infrastructure for it, you’re probably doing it wrong. Start by picking a few key drivers or pain points in your business and thinking how you can measure your success there.

2.    Riskiest Areas First – Teams tend to focus on what they know best. For example, product companies will focus on analyzing & optimizing the look and feel of every user experience (green button or aqua-marine button??). But if they have no idea how to market their product to attract users, their business will fail. They should be testing their ability to attract & convert new users. What are the areas of your business that worry you most and what can you do to test your assumptions there? The book “Lean Analytics” goes into this concept in depth.

3.    Be Agile – Analytics agility is a key skill for modern companies. Some people are scarred from year-long, staggeringly expensive BI implementations gone wrong. Don’t spend months agonizing over what tools to use to answer all of your questions. Try one tool to answer a specific question and if it fails, try another one. Tools will change over time as your priorities shift and new technologies emerge. Think of analytics tools as the means to answering whatever business questions are relevant at the time.

4.    Elevator Pitch – When a customer has a hundred ideas of what to track I ask them to give me the elevator pitch for their business. I make a note of the 5-10 verbs and 5-10 nouns that they use. You’ll be well ahead of the curve if you’re collecting rich data on those things. You can always drill in deeper once that data starts generating more questions.

Analytics isn’t just for data scientists; it’s a way of thinking about what things in your business can be measured and tested. Start small, fail fast, and build your practice as you go. You’ll get better at it over time and pretty soon, you’ll be collecting and leveraging incredibly interesting and useful data.

Michelle Keen IOMichelle Wetzler is a Director and Data Engineer at Keen IO. She teaches classes on analytics and does analytics consulting. Prior to joining Keen IO, Michelle was a Technical Architect at Accenture and managed large-scale SAP and Salesforce implementations.