While we don't generally think about it this way, our email addresses are like digital ID cards. When companies ask us to fill out forms for access to a system like Facebook, or a simple data download, we often tie some combination of our name, email address, birthday, phone number -- even our physical address -- together to create an online profile. This takes only a few seconds of our time and we get access to the latest version of Candy Crush, and the majority of people never think about what happens on the other side.

And, frankly, it's a mess.

As the recent Cambridge Analytica scandal showed with Facebook, even the data that is being collected isn't even really useful to the companies collecting it. It is faulty; records are incomplete and inaccurate. This leads to things like Facebook reporting they had more United States adult users for advertisers to market to than actually exist offline.

Kabir Shahani, CEO of Amperity, believes this is probably the biggest issue organizations face today.

"As I spoke to more and more companies, I realized that no one had the answer to what should have been a simple question -- 'How many customers do I have?'"

Hidden somewhere across all the CRM, marketing, SaaS and other tools that a modern organization uses to track consumers, there was a missing piece. Each tool was creating individual user profiles in their own systems -- and these records were not being connected. This led to inflated customer numbers, as someone may register for a newsletter with their personal email but sign up for an account with their work email -- and follow company social media with their Instagram or Twitter account. 

From a CRM perspective, that is four different customers, when in reality it is all one person.

Armed with the initial understanding of this problem, he set out to try to determine a way to fix it by scrubbing databases and joining user records -- ultimately proving that a person is the same person, whether behind a personal or work email address, a social media account, street address, phone number, or whatever information is available.

"I knew this was a problem that needed solving. I told our investors honestly that I wasn't sure that this could be done at all, and that maybe we weren't even the right team to do it -- and perhaps because of my authenticity they backed us."

In less than two years, they've raised over $37 million and announced partnerships with Microsoft, Alaska Airlines, GAP Inc. and others. Shahani has advice for others who want to follow in their footsteps.

1. Once you identify the problem, focus on it independent of constraints.

When he set out to start the business, he started by interviewing people to find their common problems. Once he found one, he ran with it.

"Market validation is the key first step to any business," insists Shahani. 

2. Don't give in to pressure to say yes to things you can't do.

Shahani's thought experiment started to come closer to life when he met Dr. Dan Suciu at the University of Washington. Shahani recalls: "It was a meeting that gave us the proof that what we were doing was theoretically possible."

Until they could prove it was actually possible, however, they remained transparent and authentic with investors about the realistic limits of their abilities. In turn, this openness gave their investors the faith they needed to back them.

3. Engage your customers early on and let them build with you.

Whether you are building machine learning models or creating the next flavor of Pepsi, it's always a good idea to involve your customers as early as possible.

Shahani says, "We may have only 'launched' in October 2017, but we were in stealth mode building with partners for almost two years. That gave us time to work on the pilot and have something solid."

When you're early on in the development process, it's much easier to course correct than later.

And as for their mission? Well, if it means we'll get three fewer advertisements from each company, the better off we'll all be.

Published on: Apr 30, 2018