Not Everyone Needs to Code – But Everyone Needs Stats

Which of these two blocks of text are more important for any marketer or business person to understand?  Block#1

You’ve rented two lists and sent the exact same mail piece to each list.  List A had a response rate of 0.36% and an average order value of $367.  List B had a response rate of 0.15% and an average order value of $265.

OR Block #2

function [theta, J_history] = gradientDescent(X, y, theta, alpha,
m = length(y); % number of training examples
J_history = zeros(num_iters, 1);

for iter = 1:num_iters
temp1 = theta(1) - (alpha/m) * sum(X*theta - y);
temp2 = theta(2) - (alpha/m) * sum((X*theta - y).*X(:,2));
theta(1) = temp1;
theta(2) = temp2;
J_history(iter) = computeCost(X, y, theta);

end end

You should be answering block #1… Code does not matter if you can’t make business sense.

Why Code Matters

Tools for Kids to Learn to Write Code

There is a real need for programmers and a real need for the average person to be computer literate (at a level that is above the current average).  However, all of the “Code is as important as reading, writing and arithmetic” hyperbole is coming from Silicon-Valley investors or employees – those that live in a bubble isolated from the rest of the world.

While the next programming star may come from Podunk, Iowa, little farm boy Clark K. shouldn’t need to struggle with overloading operators in python to graduate from High School.

Why Business Statistics Makes More Sense

Simple algebra gets you through your daily life.  Statistics helps you in your marketing day job.

Statistics are:

  • Understanding the probability that you’ll get a response from mailing list A.
  • Predicting how much profit a particular mailing will generate.
  • Knowing which variables influence response or reorders the most.

I’m willing to wager that the majority of business decisions are made without developing formal hypotheses, testing alternatives and validating the results.  As a marketer, if you start to look at decisions as testable hypotheses you’re about 10 steps above the rest.  You don’t need to code to make better business decisions.  You need stats to do that.

The Overlap or When Business Requires Code

Despite this being the era of Big Data, I think the most important calculations can be done in a relational database and with Excel.  You don’t need SAS or R to calculate response rate (block #1) – even for a million prospects.

Serving up personalized ads, using bandit-algorithms and implementing a gradient descent algorithm (block #2) all require statistics, code and business rules.  And the end results are amazing – higher response, improved experience and accurate models.

There are so many instances where learning to code will enable you to create amazing features, products or tools that will enable you to ACT on your business decisions.

You might be able to code an app by going to a “Hacker School” but will you know how to turn your one-time buyers into repeat-customers?