It's Not All About the Code: What Makes a 10x Engineer Today

Photo by Bee on Unsplash

The idea behind the “10x engineer” is that one developer can be ten times more productive than another. It goes back nearly sixty years, to a 1968 study by Sackman, Erikson, and Grant. Among the experienced programmers they studied, the gap between the best and the worst was tenfold or more.

Fred Brooks made the idea famous in The Mythical Man-Month, where he suggested building a “surgical team” around one standout engineer. The term became industry shorthand and, over time, a bit of a joke on social media.

People still ask whether 10x engineers exist, but the question means something different now. AI coding assistants have made raw coding speed much less important. These tools can draft code, test it, and clean it up faster than any person. So being able to “write code ten times faster” doesn’t count for as much.

What matters most now is judgment: knowing what to build, catching mistakes before they go live, and making architecture choices that pay off later. Many in the industry argue that the real 10x engineer is the one who makes the whole team better.

That’s why we need to rethink what top engineering talent looks like today, and how to measure it. In my experience, coding skill is only one part of the picture.

Four traits set apart the elite engineers who create lasting value:

1. Technical and coding skills
2. Understanding the engineering context
3. Collaboration and communication
4. Empathy

Here’s how to check for each one, whether you’re looking at your current team or hiring new people.

Talent Trait #1: Technical and Coding Skills

I’ll admit it: you still need a basic level of technical skill. The good news is that it’s one of the easiest things to check, even in a remote interview.

Coding exercises are easy to create and send out, and they work well as a first filter for basic skills. But AI has made them less reliable. So I rely more on live coding, where the candidate writes code while I watch, and pair programming, where we write the code together.

Working through a real-world problem with a candidate shows you things a finished take-home project can’t. You see how they approach the task and how they ask questions when the requirements are unclear. You also see whether they naturally reach for logging, testing, mocking, and serialization. When anyone can generate decent code quickly, the candidates who stand out can explain their thinking out loud.

Talent Trait #2: Understanding the Engineering Context

Technical skill only gets you so far without context. You need to know whether a candidate understands how code flows from inputs through the release pipeline to production. That’s the live system your users depend on.

Do they know what a deployment environment looks like? Do they know how to organize code so it delivers real value to the business and its users? Writing code is just one part of the job. Building the systems that turn it into working services takes more experience.

This takes more effort from the interviewer, but it helps you spot weak candidates early. You can set up a web service and ask them to use it. You can also give them an existing codebase to extend, or have them submit jobs through a build system.

Tasks like these pull the candidate out of isolation and closer to real work. One catch: this only works if the interview panel knows exactly what the role needs.

Talent Trait #3: Collaboration and Communication

Engineering is a team effort. With continuous delivery, where teams keep their software ready to release at any time, engineers have to work together at every stage. A good collaborator often adds more value than a brilliant jerk, someone talented who’s hard to work with. That lone “rockstar” genius who gets called a 10x engineer but leaves behind frustrated teammates and code nobody else can maintain. That’s a ticking time bomb with a great reputation.

Development runs on communication. Sure, that includes managers passing designs to developers. But it also means debating solutions, building agreement on tough decisions, and making sure everyone understands what’s needed. A good engineer looks beyond the list of requirements to understand the intent behind a project.

To check this on your current team, set up regular peer feedback. For candidates, create an exercise that requires collaboration. For example, ask them to update code they wrote earlier based on new input from a designer or product manager.

Pay attention to how they handle that feedback. Someone who takes it well is often more valuable than someone who does great on tests but gets defensive when challenged.

Talent Trait #4: Empathy

Empathy is easy to overlook in engineering, but it’s key to building strong teams and products. In software, empathy means understanding the real people your technology helps.

Whether you’re building a reservation app or automating marketing, your team has to think about everyone involved. That includes the operations team checking logs, customer support helping users, product managers reviewing analytics, and QA engineers testing releases. An empathetic engineer asks whether a feature adds value. They also look for ways to make life easier for everyone downstream who uses their work.

Empathy is also the hardest trait to measure. A coding test won’t show it, and neither will a checklist. The best way to spot it is to talk with the candidate once they’ve shown you the other three traits.

In that conversation, notice whether they’re interested in the wider organization, the people they’ll work with, and the customers they’ll serve. Do they feel responsible for the user experience, or do they stop caring once their code is done?

Reference checks help too. Ask whether the engineer went beyond the basics and spoke up for others, even when they didn’t have to.

Conclusion: Build a Team That Lasts

Think long-term. You’re filling today’s role, but you’re also building a team you want to keep. So hire for growth potential as much as for current skills.

Communication, collaboration, and empathy show how much someone can grow over time. And you can usually teach technical skills to people who have those traits.

Raw coding speed no longer defines a 10x engineer, because AI tools provide much of that speed for us. What’s rare now is the judgment to use those tools well and the ability to adapt as things change.

Take an average engineer who communicates well, works well with others, and cares about the people who use their work. Put them in a supportive organization, and they have the greatest potential to grow. In the age of AI, they’re more likely than the fastest coder to deliver 10x value.