In 2018, I attended a dinner where the topic was how “AI” how would impact workers. For most of the dinner, everyone focused on how blue-collar workers would be impacted by robotics and automation. Naturally, there was lots of talk about trucking being the #1 job in many states, how it was highly prone to automation, and how we should have a UBI to protect our most vulnerable.
Near the end, someone chimed in that it was low and mid-level white-collar workers who were most at risk of technological disruption. This is a somewhat obvious, but often overlooked point.
Is it easier to have a machine / “AI” (i.e., some software) replace your plumber or your paralegal? Your accountant or your construction worker?
Let’s think about how a shift to full remote work will impact salaries specifically for employees in the tech industry.
Outsourcing "white collar" jobs is not new. Tons of low-level "white collar" jobs have been outsourced to save costs around routine tasks. As an example of what happens now for business-y roles, let's look at McKinsey and Company.
McKinsey (along with all other banks, consultancies, and accounting firms) outsources the bulk of its PowerPoint development to lower cost employees in India, Poland, and LatAm (Costa Rica and Bogota, Colombia). There is nothing wrong with this. While a manager in the Bangalore office will make less than a new grad hire at McKinsey's San Francisco office, they probably make more than most of their friends and family. I bet it's a coveted job in their communities.
In Bangalore, it appears that McKinsey employs 1,500+ people in these kinds of roles. Here's a JD for a Team Leader in Bangalore. Per Glassdoor (not perfect, of course), that team leader role, at the high end, pays ₹639k ($8,528) annually. So about 1/10 of a new grad hire in San Francisco.
While many business functions have been outsourced, there has been a strong aversion to this within software engineering roles.
There are multiple reasons for this including (1) the most successful companies like Google and Facebook are well known for big campuses where proximity is viewed as a key to success (2) investors don't want to / don't like funding companies where software isn't core. An example:
So, if you combine the sentiment from Paul Graham and the assumed wisdom from companies like Google and FB, many companies felt that the best way to succeed was to have many people working together in the same room/campus.
Attitudes towards partially distributed or mostly non-HQ-based engineering were beginning to change and will be accelerated by COVID.
For several years, competition for engineering talent in the Bay Area was making it harder and more expensive to hire engineers there. As a result, companies were beginning to open HQ2, HQ3, HQ4 earlier and earlier.
I believe Stripe (an influential Bay Area-based company) helped to make remote work more acceptable by announcing that its fifth engineering hub would be "Remote". The success of companies like Gitlab and Automattic along with their habit of 'open-sourcing' their remote operating processes also did a lot to make distributed teams more accepted.
As an anecdotal counter to Paul Graham's tweets, here's a short story of how Gojek, Indonesia's most highly valued private technology company, built the early versions of their product:
In April of 2015, Sequoia Capital made the introduction to Sidu Ponnappa, the founder of C42 Engineering (“a boutique software product engineering firm based out of Bangalore, India”). He visited Indonesia on a business trip and was excited by the opportunities in Southeast Asia. By August/September 2015, Gojek had the entire staff of C42 and another outsourced engineering firm, CodeIgnition, working solely on Gojek, just as the app was beginning to explode in usage
A short while later in late 2015/early 2016, Gojek acquired both of the outsourced firms, forming the basis of their India engineering staff. The founders of both of those firms, Sidu Ponnappa and Ajey Gore would both join Gojek and are still with the company today. This seems to be a very intentional strategy for Gojek. To date, Gojek has acquired 3 Indian outsourced engineering firms along with one Indian AI-recruiting startup. These acquisitions have been the basis of their large Indian engineering offices.
The "value-add" venture capital firm of the future might be the one that owns a developer shop in India or Ukraine 🤪
In this section, I'll outline some real examples and put number behind my claims.
Here is a real anecdotal example. A friend of mine works at a 40-person startup in San Francisco where they've had trouble hiring a product manager. The product managers they wanted to hire asked for too much money and the one or two they had been able to hire didn't work out. They wound up offering a "long term contract" to a technical product manager in India with a master's degree from one of the IITs and six years of experience. They pay him $20,000/year with no benefits and no equity. They said this person has been kick-ass and is one of the harder working people on their team. They don't expect this person to go out and do customer interviews, but this person does a great job working with engineers and has brought every request from the CEO to life so far.
To add some more color to this, let’s use GitLab’s salary calculator.
Gitlab pays an "expert" Backend Engineer in a manager role in Mumbai, India between $76,815 - $82,174
For 3x as much, Gitlab could hire someone with similar credentials in San Francisco.
For about 2.5x as much as the senior engineer in India, Gitlab could hire that person in a lower cost American city like Phoenix, AZ.
Oh, and don't forget, if this hypothetical candidate lives in the United States, they get all the benefits listed below. From the Gitlab site, it appears that if they hired this person in India, they would not get these benefits (this is not at all a diss at GitLab. Just a matter-of-fact statement)
Now not all companies need to do this. For example, Facebook doesn't NEED to do this. Facebook mints money and has a very high margin business. But there are a lot of companies that have tight margins where cutting salaries by an average of 1/3 would do a lot for profitability and shareholders. The case in point is Uber.
How does one survive and thrive in this world?
- Be in top % of what you do. A mid-level engineer at Google who whines about cafeteria food might be susceptible to replacement, but Jeff Dean doesn't have anything to worry about
- As Scott Adams has said, develop your talent stack. This means, you don't have to be Jeff Dean (i.e., the top 1% at something), but you can combine being better than average at several things. For example, maybe you're an OK computer vision engineer and you also have a deep knowledge of fish, you could be uniquely qualified for a job at e.g., Aquabyte.
- Have a strong personal brand. People with large audiences who are viewed as "thought leaders" in an industry tend to be well-connected and find it easier to come across new roles. This could take the form of having a popular newsletter, a large Twitter following, or many stars on Github. A software engineer who writes a newsletter about managing engineering teams might have a CEO who reads their work and in turn would jump at the chance to hire them as a manager. A frequent open-source contributor to a project would have an easy time being hired by a company that relies on that software package.
- Cut your burn and move to a low cost of living geography. Unlike the factory worker that was put out of work, keyboard workers can move to lower CoL geographies and take lower salaries while still living comfortably. It's much easier to decrease your burn by 3x than to increase your net worth by 3x. Instead of paying $18,000/year in rent in San Francisco for a cramped studio apartment in the worst neighborhood, you could pay $6,000/year for a luxury condo rental in Bangkok or a mortgage in the midwest.
- Work at high margin company where this is less likely to happen? It makes a lot of sense why Uber needs to cut back on salaries, but as I said earlier, Facebook doesn't have the same urgency to do so. It might be the case that big companies like Facebook just keep scooping up all the high-priced California and New York-living engineering talent.