To Build Amazing Data Science & AI Teams, Leaders Think Outside Of Their Zip Codes
How top tech companies are building high-caliber engineering teams beyond their local offices and why this is a great competitive advantage that you too can use to accelerate your products much faster and more cost-effectively.
To Build Amazing Data Science & AI Teams, Leaders Think Outside of Their Zip Codes
Last year, Stripe announced that their fifth engineering hub will be remote because they want to tap the “99.74% of talented engineers living outside their metro areas” of San Francisco, Seattle, Dublin, and Singapore (their first four physical tech hubs). This decision shines a light on how inefficient, and somewhat irrational, it is to spend several months trying to hire local engineers, only to have these engineers be poached by competitors a few months later.
After all, in this day and age, when the world enjoys the highest levels of technical talent distribution and real-time communication in human history, what is the probability that the absolute best engineers for a given project are conveniently located within commuting distance to our offices? Well, for Stripe, despite their four physical locations in some of the largest tech hubs in the world, those chances apparently still only add up to 0.26%. This is why they rightly decided to look beyond their zip code to find the best talent for the job.
The Move To Remote Engineering & Distributed Teams
Stripe joined the ranks of many top tech firms who are expanding beyond a few traditional tech hubs and embracing distributed engineering and remote teams. For decades, only the largest global tech firms like Google, Facebook, and Amazon were able to establish multiple global locations to reach new talent in other geographies. It’s no surprise that these large tech firms enjoy the scale and funding necessary to create large physical hubs in many locations.
More recently, however, an increasing number of companies, like Zapier, Gitlab, and now Stripe, are emphasizing remote engineering teams as primary solutions without the need for heavy investments in physical offices.
There are a number of factors that are accelerating this industry-wide move towards distributed and remote engineering teams. The following post and examples focus mainly on building data science, machine-learning, and AI engineering teams, as those are the primary areas of expertise for Factored, but these concepts apply to other engineering roles as well:
It’s Increasingly Difficult and Expensive to Hire Local Engineers
In traditional tech hubs like Silicon Valley, Seattle, New York, Los Angeles, Boston, and Austin, competition for data scientists and machine-learning engineers is incredibly fierce. According to recent studies published by IBM and LinkedIn, data scientists and machine-learning engineers are the fastest growing and most difficult roles to fill. It takes the average company about 5 to 6 months to fill a data science role, and some of these openings remain unfilled for over one year. Further, given the significant shortage of qualified data scientists and AI talent, the first-year cost to hire local engineers can easily top $250,000 USD if you include: salary, benefits, and recruiting costs in this calculation.
Complicating matters further is that qualified engineering talent regularly gets poached by FAANG companies (Facebook, Amazon, Apple, Netflix, Google) and the hottest unicorns, so the vast majority of talented engineers typically don’t consider job openings in other companies. In response to these growing difficulties faced by most tech companies, building a distributed team of remote data scientists and engineers makes for a very smart strategy that allows them to find top 3% talent in less traditional tech hubs. If done correctly, building a remote team can significantly reduce cost, difficulty, and timelines associated with building a highly qualified data science or AI engineering team.
Superior Quality of Data Science and Engineering Talent Abroad
It is widely accepted that raw intelligence and tech talent is broadly distributed around the world, and as such there are plenty of top 3% of engineers living outside of Silicon Valley and other traditional tech hubs. Those of us who’ve had the opportunity to live or study abroad won’t be surprised to learn that schools and universities outside of traditional USA tech hubs are often more rigorous when training local mathematicians and engineers. Even if you haven’t lived abroad, just look at our best universities in the USA and consider the high and growing proportion of international students excelling in their most challenging math, science, and engineering programs. Particularly for quantitatively focused roles in data science, machine-learning, and AI, the decades of rigorous training that students receive abroad in mathematics, statistics, and engineering creates a healthy pool of highly qualified “quants” around the world. The talent is plentiful, but that doesn’t mean it’s easy to find and qualify the best engineers for your particular needs.
Finding, Vetting, and Testing Remote Engineers Is Not Easy, But It’s A Great Investment
Finding great engineers remotely, at scale, requires building bonds with the local tech community and local universities, as well as implementing a rigorous process for vetting and testing these engineers and putting them through multiple rounds of difficult interviews. Not only are you looking for great technical talent, but you are also looking for superior communication skills and personality traits that will make them effective remote contributors.
At Factored, our recruiting process is extremely rigorous, by design. We only accept less than 3% of applicants (that is a lower acceptance rate than Harvard or Stanford). We use multiple coding interviews, cultural fit interviews, and programming and theory tests to ensure proper vetting of candidates. Tested against thousands of other engineers around the world, including Stanford University graduate students and Silicon Valley professionals, our remote engineers rank in the top 1-3% in core skills like mathematics, data science, machine-learning, deep-learning, and algorithmic programming. This rigorous vetting process guarantees that your engineers are silicon-valley ready.
New Technology Enables Improved Real-Time Collaboration
Across most companies, Slack and other communication tools like it have already replaced in-person conversations as the preferred method for employees to communicate with each other in real-time. Yes, even if your colleague sits in the row of desks across from you, chances are you are already communicating mostly via Slack messages. Especially now that most companies have adopted open-plan office spaces, in-person conversations are significantly reduced and sometimes discouraged by “headphone” policies.
The same goes for “in-person” meetings when was the last time you went to a work meeting where every participant was physically there in person and not connected via a conferencing tool like Zoom or Google Meet? Chances are these tools are already part of your daily routine, even if participants are physically located in your neighborhood, they may be in different offices or buildings. Add other tools like Google Docs and GitHub and you already have everything you need to enable distributed teams to work effectively, sometimes more effectively than in person. You are already doing these things anyway, and your whiteboarding sessions can now be shared globally via Google Docs and other tools, and have the added benefits of video recording, sharing, comments, etc.
More Time For Productive Work and Better Quality Of Life
The average commute in most large tech hubs is bad and getting worse. For example, in the San Francisco Bay Area/Silicon Valley, the average commute is about 1 hour long, and 20-30 miles in distance per day, and a growing number of super commuters spend 3 hours (50 miles each way) daily getting to and from their physical office. Beyond the terrible effects on the environment posed by such long commutes, think about the wasted time and productivity these long commutes signify to your bottom line. That’s an average of 225 extra hours of productive work that your data science or engineering team could spend solving business problems or building products and valuable IP instead of sitting in traffic.
For these and other reasons, not to mention recent viral epidemics like Covid-19, we see the world moving faster towards accepting distributed teams of remote engineers as the norm.
More Tips For Building High-Caliber Data Science and AI Engineering Teams Remotely
Factored builds high-caliber data science and machine-learning engineering teams for ourselves and for many other top tech companies, so we have a bit of experience with the advantages and some of the common mistakes that companies make when building remote data science and AI engineering teams. While we are big believers in the benefits of distributed teams and like to promote the virtues of remote engineering for the advantages described earlier in this post, there are a number of things to consider before building your remote engineering team.
Working in the Same Time-Zone is More Productive And Will Save Your Team Costs and Headaches.
One important consideration is the issue of time zones. At Factored, we have found that while physical proximity is not as important, real-time collaboration and real-time feedback is still essential. In fact, good real-time communication becomes even more important for remote teams. Though some companies argue that asynchronous communication is fine because it allows coders to focus on coding and to not get too distracted, at Factored we have found that the right solution to ensure optimal communication is to build distributed teams of remote engineers who live and work in USA time zones, where we are headquartered and where most of our clients are located.
Too many engineering managers have dealt with time-zone difficulties when working with remote offshore service providers. The idea that “they work while you sleep” sounds great in principle, but it leads to wasted time and resources due to shortened communication windows. We think that nearshoring models, where engineers “work while you work” are much more effective. We believe this real-time presence in our local time zone is essential to making remote engineering work for us as productively as possible.
Legal and HR Complexities are Minimized
There are a number of other potential hurdles to consider if you try to build remote offices and try to structure remote employment yourself directly, so you may want to think of alternatives where third parties help you with this process. If you are a growing tech company, you probably want to focus on building your product and doing what you are good at. You are probably not in the business of learning about and complying with local regulations related to employment law, payroll, employment, taxes, healthcare, market-competitive benefits and other local requirements or ordinances This could add significant costs and complexity if you decide to tackle this directly. For that reason, we recommend working with experienced third-party companies who specialize in this area. They will save you headaches, time, and money in the short and long run. Factored, for example, handles all of these requirements when building data science and AI engineering teams for our client companies so they only need to focus on building their product, not learning about local laws and requirements.
Don’t Forget The Details That Will Make Your Team Happier And More Productive
Additionally, while options abound for coworking spaces that could serve as your office space, it also helps to have an expert company on the ground helping you select the best and safest neighborhoods close to public transportation, as well as test the internet connectivity and ensuring it is the optimal space for your team. In addition, it helps to have expert locals negotiate your rent expense to avoid paying the “gringo” or “foreigner” price. Our team successfully negotiated a rent reduction of 50%, for example, for our Medellin office. Further, making sure that your team has great perks like two meals per day, healthy snacks, transportation stipends, and great spaces to work and socialize after work, will ensure that your employee satisfaction and retention rates.
Don’t Wait, Start Building Your High Caliber Data Science & AI Team Today
It makes no sense to keep your data science and AI projects on hold because of hiring difficulties. Most leading tech companies are already tapping the superior data science and AI engineering talent outside of their local areas. The companies are building a competitive advantage over their competitors and growing their teams much faster and more cost-effectively. By implementing a rigorous vetting process, and building the right partnerships, your company could start placing great engineers on your projects in just a few days.
If building a team of top 3% data science, machine learning, and AI talent is what you are looking for, reach out to Factored and we will point you in the right direction. Click HERE to contact us. I regularly respond to email requests personally and would be glad to help.