How to mine your data

Written by Peter H. Diamandis | Feb 13, 2015

The most successful companies on the planet are data-driven companies… Google, Facebook, Uber, AirBnb, etc.

In fact, to survive this decade you need to be data-driven.

We all collect data – Our emails, sales performance metrics, LinkedIn connections, customer profile information, tweets, survey results, flight and travel history, etc. etc. are all datasets ripe with useful information.

Collecting the data isn’t the challenge.

The challenge is how you write an algorithm to analyze that data and turn it into simple, powerful, actionable insights that change the way you do business.

In this blog, I want to share with you an example of how I did it and the rather simple platform I used – that you can use as well.

[ Click to Tweet about this (you can edit before sending): http://ctt.ec/F39aa ]

Start with a challenge and objective

I host a lot of CEO-level events, for XPRIZE, Singularity University & Abundance 360.

One of the benefits of attending these events is meeting people that could be important to you – future business partners, customers, investors, etc.

Usually, these connections arise by random chance, or through a personal introduction.

But what if a computer algorithm could instead mine everyone’s biographical information and identify for you, in advance, “the top 10 people in the room you would be most likely to do business with?”

This is exactly the data-mining/machine learning challenge I attempted to implement for the 250 CEOs, entrepreneurs, executives, and investors who attended my Abundance 360 Executive Summit this past January.

The data I collected…

When a new member is accepted into A360, I ask them to fill out an extensive profile.

The profile includes basic information, but also asks for an additional 30 data points including such other things as:

  • Your company’s revenues
  • Your Industry
  • Grand Challenges of interest to you
  • Crowd Tools of greatest interest to you
  • Exponential Technologies of greatest interest to you
  • Your angel investment profile

Then we called in the experts…

The next logical step is to bring in a data scientist, to analyze the data and build an us an algorithm…

The problem is that data scientists are in high demand. A recent McKinsey report estimates that by 2018 the U.S. alone could face a shortage of 140,000 to 190,000 data scientists.

So, a few months ago, a friend of mine recommended that I try a new online service called Experfy.

Experfy is a consulting marketplace where you can hire data scientists and machine learning experts to build algorithms, solve business intelligence problems, and mine big data. (Note: I have no ownership or relationship with Experfy, I’m just sharing a good experience with you).

It’s pretty simple actually.

Step 1: You sign up, post a job (much like Freelancer, Elance or oDesk, but specifically for data science), and a crowd of curated, qualified data professionals submit detailed, technical proposals that outline how they would solve the problem.

Step 2: A member of my team went onto Experfy and created a job called “Matching Algorithm for Executive Mastermind Group.”

Step 3: He outlined our objectives, the insights we wanted to make, the types of data we had, the format of our desired outputs.

Step 4: We proposed a price range of $1,500 - $2,500.
Within two days, we had 17 proposals. They ranged from professors to chief data officers of companies to new college graduates.

Step 5: A consultant from Experfy helped us narrow down the proposals to the top three candidates (a difficult process if you don’t know how to do technical evaluations).

Step 6: We interviewed all three, picked one, agreed on a price, and he began the work.

The Results

Two weeks after posting the project, signing NDAs, and sending over our data, we had a working algorithm.

We refined it for a few more weeks, and just like that we had a robust recommendation engine powering our event.

Every member attending A360 received a list of the “Top Ten People They Were Most Likely to do Business With”, and the experience of meeting these matches was both fascinating and very valuable.

How are you leveraging your data?

I wanted to share this story as a fun case study to illustrate how easy it is to build an algorithm and begin to mine your data.

So, this month, consider running a data science experiment.

If you want to go into greater depth, I talk more about how to do this in my new book BOLD (boldbook.com) and in my digital mastermind group Abundance 360 Digital (find out more information here).

P.S. Have you purchased a copy of BOLD? I’d love it if you could share a review here.

P.P.S. Every weekend I send out a “Tech Blog” like this one. If you want to sign up, go to PeterDiamandis.com and sign up for this and my Abundance blogs.

P.P.P.S. Please forward this to your best clients, colleagues and friends — especially those who could use some encouragement as they pursue big, bold dreams.