table full of keys
All businesses want to get the most out of their analytics investment. Here are a few tips to make sure you do PHOTO: Marcin Szmigiel

Even with the best of intentions, sometimes the pressures to go faster cause us to get off on the wrong foot.  

At the time these steps look like the perfect shortcut, only to turn into the dead-end of a maze, causing initiatives to stall in their tracks. 

Analytics is no different. Here are four of the most common traps I see companies make with their analytics investment:

1. Walk, Don’t Run

In any endeavor, it’s often a good idea to understand where you are, and where you’re trying to get to before you get started.

The analytics equivalent of this is an Analytics Maturity Model, which highlights your strengths and weaknesses. The model tells you where the low hanging fruit is and what is your best opportunity. 

A simple framework is: data, reporting analytics, insight development, predictive modeling and machine learning.

Companies often make the mistake of failing to evaluate where they are on a maturity model, before they jump in. This causes them to try to skip levels.  

When learning to drive we all wanted to jump on the highway, though deep down we knew we needed to develop confidence behind the wheel first. Analytics is no different. 

Examples I commonly encounter include building reports before creating quality data or implementing complex predictive models before having basic reporting in place to inform model accuracy.

2. Talent Over Tools

Despite hours of practice, I was never the first kid picked in sports. I had great sneakers though!  

Did my sneakers help me get picked? Not really. My classmates wanted to win, so they picked the best athlete, not the nicest shoes.  

Winning at analytics is no different.

The equivalent of sneakers in analytics are the tools you use. Making hiring decisions based on the tools a candidate knows, rather than past results is the equivalent of choosing the player with the best shoes.  

Constraining on SAS rather than R is the same as choosing someone wearing Nikes rather than Adidas. Good analysts learn new tools as easily as they change shoes. Showing you’ve been an effective analyst is much more important. Focus on results rather than bling.

3. Analysts Are People Too

When watching a master perform their craft I’m often struck by how effortless they make it look — whether it’s champion athletes, star chefs or great analysts and programmers.  Somehow the best make it look easy.  

Some analysts make everything look simple, identify insights faster and in general, have mastered the tricks to find ways to get things done better.  So why do so many organizations, large and small, act as if all employees are of equal capability like pawns on a chess board? 

Granted, some situations warrant this. And I’m certainly not advocating desperate attempts to keep a strong performer who is leaving: the data doesn’t support it. I’m calling out the pernicious belief that a random employee can be easily replaced with another random hire.

The challenges here are many-fold. Your employee already knows your data. Your employee is a cultural fit. Your employee delivers results.  

The new hire doesn’t know your data, may not fit in and who knows what kind of track record they will produce. Whatever their role — Data Engineer, BI Analyst, Modeler — all analytics team members are knowledge workers who we pay to think rather than repeat tedious tasks.

4. The Right Tools Help You Do the Job Right

The final misstep is one of the most exasperating, particularly for your analytics team. If you’ve ever set out to repair something in your home, you know this misstep.  

Quite often I try jury-rigging a solution together with the tools on hand. Eight hours later I do what I should have done to begin with: drive to Home Depot to pick up the right tool for the task.  

When doing advanced analytics, such as machine learning, the computers you purchase are your tools.

Very often an organization will build up the analytics capability, investing heavily in people and software, only to balk at the relatively minor expense of super-fast hardware to do the analysis.  

With faster hardware your team will spend more time testing ideas and less time waiting on results. With faster hardware, your team will brag to their friends, giving you better access to talent. With faster hardware, your team will be happier and more productive.  

Every environment is different, so I won’t suggest what hardware you should get. Your team can tell you that.  

However, there are two classic warning signs to watch out for: only providing laptops with no servers for development and if your best analysts want to use their home computers because they are more powerful. Both cases are red flags that the IT budget needs to be increased.

Think Ahead to Avoid Dead-Ends

Avoiding these frequent mistakes when building your analytics team is possible, with a minor amount of forethought and common sense.  

Proper preparation, knowing the attributes that spell success, treating people like people and investing in effective tools will ensure avoiding these all too frequently made common mistakes.