Connect with us on Facebook Follow us on Twitter Join our LinkedIn Group Subscribe to us on YouTube Share with us on Google+ Subscribe to our RSS feed

April 01, 2016

Working Smart

What Is Big Data and How Can it Help Increase Revenue?

by Mark Rawlins


Click here to order the April 2016 issue in which this article appeared or click here to download it to your mobile device.


If you ask 100 people “What is big data?” you would likely get 100 different answers. According to Wikipedia, “big data” is a broad term for data sets so large or complex that traditional data processing applications are inadequate. One of the most famous examples of big data in action is when IBM’s Watson system played Jeopardy on national TV and beat the two reigning champions. I was very impressed, but I was left saying “Show me the money. How is playing Jeopardy going to help my clients succeed?”

Fortunately, I ran across another example of big data that is more relevant. Pratt and Whitney makes jet engines, and one of the key metrics for jet engines is “downtime.” An airplane that can’t take off because of an engine problem is an airplane that is losing the airline money—lots of money. Each jet engine has numerous electronic sensors that together generate up to 500 gigabytes of data per engine, per transatlantic flight. Pratt and Whitney gave IBM 18 months’ worth of engine sensor data, and asked them to use that data to predict the maintenance each engine would require over the next six months. When they compared IBM’s predicted maintenance to the actual maintenance required over those six months, they were astonished—IBM had successfully predicted 97 percent of engine maintenance events and 100 percent of incidents where engines had to be turned off during flights.

Pratt and Whitney made millions of dollars by using IBM’s data analytics to vastly improve engine reliability, which increased sales. Simply put, they turned big data into big money.


Traditional data processing is about getting answers, whereas big data identifies probabilities.


What Can “Big Data” Mean to My Company?

From a business perspective, I believe big data is about using historical data to predict future events in a way that improves the bottom line. MLMs generate a much smaller data set than jet engines do, but the idea is the same: If you can use this data to forecast future events, creating what I call actionable information, you can increase growth and profits.

How is big data different than traditional data processing? The idea behind big data is to look across thousands or even millions of individual data points in a way that identifies useful patterns while ignoring the “noise” that is inherent to such an enormous data set. One way to look at it is that traditional data processing is about getting answers, whereas big data identifies probabilities.

How can you use data in a direct selling company? The first step is figuring out the events you would like to predict, but haven’t been able to because of the limitations of traditional data processing.


According to Wikipedia, “big data” is a broad term for data sets so large or complex that traditional data processing applications are inadequate.


What Would We Like to Predict?

To answer this question, we need to remember who creates MLM growth. MLMs grow because they have an active and growing group of customers who are being sold to by an active and growing group of sales people. These sales people are recruited, trained and motivated by an active and growing group of sales leaders. In turn, these sales leaders are typically motivated by a small group of people I call dream builders. To get an idea of the number of sales leaders and dream builders in your organization, consider the five categories of participants in a modern direct selling company:

  • Customers who buy products for themselves
  • Social enrollers who like the product and refer interested friends and relatives
  • Sales people who seek out customers
  • Sales leaders who seek out customers and recruit others
  • Dream builders who build large organizations and motivate and inspire others

As an approximate rule, for every 10 customers, you have one sales person. For every 50 sales people, you have one sales leader. And for every 50 sales leaders, you have one dream builder. Simple math shows that if you have a million customers, you have around 100,000 sales people, around 2,000 sales leaders, and only about 40 dream builders. (These are approximate averages.)

What about these distributors would you like to forecast?

Who Has Potential to Step Up?

First and foremost, you want to know which sales people have shown potential to become sales leaders, and which sales leaders might become dream builders. When a company is small, identifying the next crop of sales leaders and dream builders is easy. You know them personally, and can identify those with potential. These personal relationships allow small companies to identify and nurture their initial crop of leaders. However, when your company has grown to 10,000, or 100,000, or even a million participants, it becomes impossible to personally identify the next group of sales leaders and dream builders—and if you aren’t identifying and nurturing your next crop of leaders, growth will ultimately slow.

How valuable would it be if you could use big data along with predictive analytics to generate a list of 100 people who have the traits of a dream builder?

Who Is Slowing Down or Retiring?

Being a dream builder is a lot of work, with constant travel, nightly conference calls, frequent webinars, and numerous meetings. At some point, many of your top dream builders and sales leaders are making enough money that they begin to slow down or scale back their efforts—in other words, they retire. The challenge for your company is that they almost never announce their retirement; they simply start slowing down. At first, without the power of big data and predictive analytics, this slowdown is imperceptible. The challenge to you is that by the time you notice it’s happening, it’s almost too late to start nurturing new leaders to continue to generate growth.

I’ve watched this pattern over the last three decades and have found it to be amazingly consistent: If a dream builder starts to retire today, within a year or two the growth in their organization slows, then flattens, then declines—unless others in their organization begin stepping up to take their place. Predicting and identifying the people who are slowing down gives you time to take the steps necessary to grow your next wave of leaders.


If a dream builder starts to retire today, within a year or two the growth in their organization slows, then flattens, then declines—unless others in their organization begin stepping up to take their place.


What Data Is Available For Analysis?

This is where our industry’s big data comes into play. Your company should be storing downline, commission, and sponsoring information—and should keep that data for years. For a large company it can be “big data.” A large company may store gigabytes of data from each commission run to use for future analytics.

You also can reveal interesting and useful patterns by combining your internally generated data with demographic census information. For example, if you knew that a significant portion of your successful sales people are women in their 30s with children, and who have a household income between $35,000 and $65,000 a year, that information could be a gold mine for your marketing, sales, and promotions teams.

As a side note, it goes without saying that you’ll want to be careful how you use demographic information. You can easily trigger fears of big brother if you use the results in inappropriate ways.

Teasing useful information out of massive data isn’t easy. That’s why there’s an entirely new field of data scientists who do nothing but big data analysis. But there are some things that you can do without expensive data scientists.

Who Are Your Sales People?

Sales people are the easiest group to identify and track because most companies define a minimum amount of earnings necessary to be considered a sales person. Depending on the company, this is commonly between $200 and $500 per month. Next, you have to figure out which sales to attribute to a given sales person. An easy rule of thumb is to count all sales within three levels of a distributor, excluding anyone who is also a sales person.

After you have an algorithm for attributing sales, you can start to look at retention, growth and other “health” indicators for your sales people. In my experience, without a healthy group of sales people—with a very high retention rate—a company will not be successful over the long term.

Who Are Your Sales Leaders And Dream Builders?

Before you can think about predicting growth or retirement, you have to be able to identify the people who are either growing or retiring. Obviously earnings is a critical part of this calculation, but you also have to understand who to attribute this activity to. Just because a leader’s team is growing doesn’t mean that he or she is the one doing the work. Two relatively easy things to look at are:

  • How many levels of sponsorship away from the leader is growth happening? If all of a leader’s growth is happening deep in their organization that is a concern.
  • Is the growth all happening in one leg? This is the most tell-tale sign that someone in the downline is actually generating the growth.

For example, in the graphic pictured below you can see that Leaders A, B, C, D, and E all have $1 million in downline organization volume. The question is: Which one of them created the volume?

In this simplified example, it’s pretty clear that Leader E is the one who created the volume. In actual practice, the answer is usually not this easy to ascertain. This is where a data scientist can help—by building the sophisticated algorithms that attribute growth to the correct leader.

After you have built these algorithms to correctly attribute activity, you can use your historical data to identify the traits of up-and-coming leaders. In other words, you can create algorithms that start to forecast future events—much like Pratt and Whitney used historical data to predict future engine behavior. This will require several years of historical data. Using that data, you can look at the behavioral patterns shown by your current sales leaders and your current dream builders during their first years in the business. When did you first start seeing indications that someone was going to become a sales leader or dream builder? By comparing those patterns with the actions of your current distributors, you can identify distributors with potential to become your next crop of sales leaders and dream builders—information that is critical to your continued growth and success.


After you have built algorithms to correctly attribute activity, you can use your historical data to identify the traits of up-and-coming leaders. In other words, you can create algorithms that start to forecast future events.


Turning Data into Dollars

There are several ways to monetize this information after you’ve identified it. The most obvious is to use it to identify your next potential group of leaders.

You also can use the information to answer other key questions. For example, are you paying your leaders who are retired a higher percentage of organization volume (OV) than you are paying active leaders? Because the dream of direct sales is to build an income stream that provides freedom, most compensation plans continue to pay leaders who have “slowed down.” But if your plan pays them more than it pays those who are actively working and building, that can cause real financial problems, and some adjustments might be in order.

Second, do you have a healthy and growing class of sales people? Look at how much money you’re paying active and growing workers and how well you’re retaining your sales people. Do those selling in the $200-500 range stay active? A definite sign of a sick compensation plan is when the people who get to that level either go on to be sales leaders or quit. A healthy compensation plan creates a class of sales people who continue to sell in that range long-term.

Third, who in the field are you listening to? Is your corporate feedback structure set up to listen to your active and growing sales people, sales leaders, and dream builders? Do you have a good mix of all of those people? Many companies fall into the trap of listening to the people who make the most money, or even worse, the same people they’ve always listened to. Often, the vice president of sales listens to a certain group of people when the company starts, and then continues to listen to those same people long after they have started to retire and are no longer the most relevant voices.

More importantly, you need to listen to people whose business is currently and actively growing. In other words, listen to those who are still in the trenches, not just those with a high total OV or who have a high total number of distributors and growing OV—because it may well be that their growth is actually being created by someone else.

It also is important to avoid the trap of listening only to people who continually lobby you. Remember that the people who have the most time to lobby you are people who have plenty of time on their hands—because they are retired. And like all lobbyists, they will lobby for the things that are good for them even if those things aren’t necessarily good for your business.

Conclusion

Just as Pratt and Whitney did, you can turn what may seem like routine or even mundane data into actionable information that increases your company’s revenue. Even though it’s not likely that you’ll achieve the 97 percent to 100 percent accuracy of IBM’s jet engine analysis—because people are, after all, less predictable than machines—the results of big data and predictive analytics will help you understand your business in a way that has never before been possible.


Mark RawlinsMark Rawlins is Founder and CEO of InfoTrax Systems.