Understand, Identify, Execute

At Meta, we work on many different problems that only become more complex over time.

When I joined the company in 2005, there were one million college students using what was then “the” facebook dot com. One of my first projects was to expand our community beyond just .edu email addresses and make it so high school students could join too. This was very controversial at the time 😊.

Today, there are over 3 billion people using our family of apps (including Facebook, Instagram, Messenger and WhatsApp), our devices (Portal, Quest, Ray-Ban Stories) and new metaverse products (like Horizon Worlds). I lead the team focused on serving everyone who wants to use our products, no matter where they live, what language they speak, the quality of their mobile phone or internet access, or whether they can afford a data plan.

Of course, every problem big and small is different, but I have found that their solutions share some things in common, a general framework I call: Understand, Identify, Execute.

Understand, Identify, Execute means:

Understand the problem space
Identify the levers to achieve your goal
Execute on those levers perfectly

This sounds simple enough, but it has really helped me and my teams structure how we solve problems. Like any good framework, Understand, Identify, Execute has stood the test of time. Before I adopted it, I would go straight to throwing spaghetti at the wall to see what sticks, which is not a winning recipe. Over the years, I find myself (and many others) coming back to it.

So, let’s walk through each phase in more detail:

1. Understand

In order to solve a problem, you need to understand the problem space.

Often, the answer to a question or the solution to a problem is not a mystery. There are clues everywhere, if you know where to look. My friend and coworker Alex Schultz summarizes this sentiment well: “Why guess when you can know?”

That’s why I always start with the Understand phase. First, I collect all of the relevant information I can get. This includes quantitative data like analysis and modeling and qualitative data from interviews and focus groups across both internal and external sources.

“Understanding” the problem space can be more broadly interpreted beyond acquiring traditional kinds of data. To explore all of the dimensions and perspectives of a problem, I’ve seen teams do everything from blue sky brainstorms to pixel perfect designs; anything works, as long as it helps you literally or metaphorically gain a better understanding of the issues.

As William Gibson famously observed “the future is already here, it’s just not evenly distributed.” I take this to mean: we can learn a lot about the future if we look at the world around us, and focus our magnifying glass on the right country or demographic or product. We can try to predict the future, but again — why guess when you can know?

I cannot overstate the importance of taking your time in this first phase. I’ve seen too many projects trip at the starting gate because the people involved, however enthusiastic, did not do the work to really understand the problem space. This kind of work is never time wasted.

2. Identify

After you’ve worked to understand the problem space, step two is to identify the most important levers you can use to solve your problem or achieve your goal.

If the Understand phase is about casting as wide a net as possible, the Identify phase is about pulling out only the most important threads; finding the “signal in the noise” that allows us to focus our limited time and energy on the most impactful things.

To get started on this path, I ask questions like:

What are the opportunities you can pursue to solve your problem/achieve your goal?
For each opportunity, how does it impact your goal?

At this point, I should have a list of opportunities ranked by importance. Then I need to figure out how to address these opportunities. So then I ask the next logical questions:

For the biggest opportunities, what are the best ideas?
For the best ideas, what is their likelihood of success?

If you can understand the problem, you can begin to identify solutions. The best ones to pursue are the ones with the biggest impact on the project. And for those, not only do you need to list out your best ideas, but you should focus on the best ideas that are also most likely to be successful.

Overall, I find that the best ideas come from people brainstorming together, and often are nothing like the ones you originally had in mind.

3. Execute

Once you identify the biggest opportunities and your best ideas to address them, you need to execute these ideas. This is the “just do it” phase of the work.

In my experience, teams often fail at the Execute phase. Just like I’ve seen people get stuck when they can’t understand a problem, or can’t find the best solutions, I’ve also seen people come up with great ideas but fail to put them into practice.

That’s because: it’s never quite enough to “just do it.” You need to “just do it, perfectly.” I overcorrect by emphasizing not just execution, but perfect execution. Even if you identify the right things to do, it doesn’t matter if you don’t perfectly execute them. (More on that in a this note 😊).

Example

To illustrate the Understand, Identify, Execute framework, let’s use an example from the early days of Facebook.

The product team responsible for the Facebook registration flow was trying to solve a problem where people were not confirming their accounts. This team wanted to increase account confirmations, which is important not only to deter spam and bots but also improve account security generally.

So, taking the framework one step at a time:

Understand: We analyzed data on the registration flow by breaking it into steps. We wanted to understand what was happening at each step in the flow and where we were losing people along the way. This is called a “waterfall analysis,” where we measure the number of people who land on the registration page, fill out the required registration fields, click “sign up”, attempt to confirm their account, or successfully confirm their account.
Identify: We found that around 1/3rd of people who started registering an account did not ultimately confirm their account. 23% of them did not even try. This was a huge opportunity if we could work out the best way to address it.
Execute: After some brainstorming, we decided to proactively nudge people that hadn’t confirmed their accounts by sending an SMS with a confirmation code when they logged into Facebook.

I really like this example because it illustrates all three parts of the framework in action:

We gathered data to understand the problem (the registration waterfall);
We identified the biggest opportunities (⅓ of registrations don’t confirm)
We perfectly executed our best idea (a reminder SMS)

In the end — account confirmations went up by almost 10% and Understand, Identify, Execute worked.

I’d love to hear from anyone with a similar experience in the comments. Do you have examples of when you’ve used this or another framework to solve a problem?