Refresh: The App a Secret Agent Would Love

A few minutes before his helicopter touched down in a covert military base just outside of Kabul, Afghanistan, Tommy Thompson reached for his secret weapon. He was about to meet with top Afghan officials and he needed to ensure he hit his mark. But Thompson’s mission to the war-torn region in 2004 did not involve delivering guns and bombs. As the U.S. Secretary of Health and Human Services, the diplomat was there to win hearts and minds.

To accomplish his directive, assigned to him by the President of the United States, Thompson relied upon information delivered at exactly the right time and place. Minutes before each meeting with dignitaries, he was handed a top-secret intelligence briefing.

The CIA-prepared binder contained the most vital, and at times trivial, information on who the Secretary was going to meet. A quick glance provided the context for the meeting, notes from previous encounters, and often times contained personal information.

“Speaking to Secretary Thompson after he read his briefing gave you the feeling you were the most important person in the world,” said Bhavin Shah, who traveled with Thompson to Afghanistan. “You understood that he cared about you enough to mention the things that were on your mind.”

Thanks to his dossier, Thompson was never without a piece of information, which when used in the right context, served to ease the conversation. His meetings were never awkward, he was never dull, and somehow, it always appeared the Secretary was, in Shah’s words, “conversationally refreshed.”

Thompson’s personal soft skills, often turned into hard results. According to Shah, “You would suddenly see stern generals and skeptical officials relax when Thompson dropped important detail to show he cared about the individual.”

Years after traveling with the Secretary, the memory of the power of good conversation and those all-important briefings stuck with Shah. In 2012, Shah along with co-founder Paul Tyma, decided to see if they could build a technology inspired by Thompson’s official briefings.

“I wanted my own dossiers,” Shah said. “I interact with so many people during my day, and I often find it hard to remember everything. I had trouble keeping everything straight.” But whether it was actually possible to build a technology good enough to do the job, was still an open question.

I first met Shah during the early days of his new company nearly a year and a half ago. A venture capital friend made the introduction and over burgers in Palo Alto, I understood why he had connected us.

Shah needed to deliver relevant information the way Thompson’s dossiers did. But with no g-man handing over carefully drafted binders, the concept would need to leverage new technologies and create new user behaviors. To succeed, the app had to become a habit.

We sketched out a few ideas using a framework I developed called the Hook Model. The Hook distills the elements of a habit-forming technology into four fundamental steps: a trigger, action, variable reward, and investment and is found in all sorts of products, which keep users coming back.

Shortly after our meeting, the Refresh team built a bare-bones version to test the validity of their idea. The v.1 used SMS text messages to send early users snippets of information gathered from the open web about the person they were about to meet.

A message like, “Don’t forget to ask Michael about his recent trip to Vancouver,” could be sent based on a quick scan of my calendar and Facebook account. 15 minutes before a meeting, this kind of information proved to be a great conversation starter.

In the early days, Shah used the so-called “man behind the curtain” technique to gauge user response. Testing the idea with a small handful of investors and friends, Shah hired a woman named Colleen to send out talking points manually over SMS.

Shah believed that if users responded well to the messages Colleen was sending, it would be worth building a technology to automate the process. Shah’s first test was to figure out his users’ triggers. He needed to understand what problem he was solving for his users by providing them with pertinent information before a meeting.

After several months of testing and tweaking, Shah decided he was ready to build the real thing. His handful of users had verified the first three phases of his product’s hook. The trigger was the fear of forgetting to mention important information during a meeting. The action would be to open the app, and the variable reward would be the new information the user would find as well as the positive feedback from the conversation itself. Shah also believed the user felt another trigger directly after the meeting, when they feared forgetting important information.

However, Shah’s concept for a habit-forming technology was still missing a critical component, the investment phase. Whenever users put data into a service, they invest in it and increase the likelihood of returning, a phenomenon I call “stored value”.

I strongly advised that while pulling publically available information from social networks about who the user was about to meet was a good first step, the user needed to add new facts to sustain the habit.

Users had to input the kind of information only they knew. In order to refresh the user’s memory before their next meeting, they needed to add private notes. Informational nuggets like a person’s favorite sports team or a home improvement project they might be working on, could only be collected by the user. But could an app help make these memory joggers useful?

Shah’s team went from the SMS version to an early app prototype, which attempted to close the loop. By connecting the app to my calendar, I would receive a notification promptly after my meetings to ask, “What would you like to remember about your meeting?”

Unfortunately, results from the prompt were lackluster. Asking users to input notes with an open-ended question worked for some, but didn’t engage those who were less motivated.

That’s when the team came up with a novel solution. Instead of using the old vague prompt, they decided to test structured categories. Names of family members could be put into the “names and people” category, while notes on a big upcoming vacations could be entered under “big events.” The seemingly simple change made a massive impact. The new prompt increased the number of notes entered by users 3X over the old method. Shah’s app was finally becoming a user habit.

With their hook in place, Shah’s team decided to expand the reach of their new app, which recently made its debut in the Apple App Store. Inspired by Tommy Thompson’s endearing conversations, the new app is called Refresh. Update: Refresh was acquired by LinkedIn, and is no longer available for download.