The One Million Dollar Team
February 4th, 2013
Recently, I wrote about what I consider to be an optimally sized, self-sufficient software development team. To summarize, I proposed that a minimal viable product team would consist of ten people including six devs, two Product engineers, one Test engineer, and one DevOps engineer.
After writing the article, I became curious about how much such a team would cost per year. So, I decided to collect some data and find out.
Getting Salary Data
Unfortunately, accurate data for technology salaries is not easy to come by for a couple of reasons. First, some positions (like developer) are known by multiple titles (Software Engineer, Computer Programmer, etc). Second, titles like Product Engineer aren’t that commonly used, so there’s just not enough data available.
To overcome these problems, I used the following approach. First, I got raw salary data from two well known online resources Glassdoor and indeed. If the actual job titles I was looking for didn’t have enough data, I used a variety of reasonable substitutes.
Next, I took an average of all these numbers to come up with the Average Industry Salary for the position. Of course, the problem with average salaries is that they attract average people. And, unless you’re Google or Facebook, most companies have to pay a premium (in the 15% – 30% range) to attract top talent. So, I adjusted average industry salaries for a more realistic estimate.
By the way, my reasoning behind specific adjustments is this: the rarer the position, the higher the premium you need to pay. That’s why Test and DevOps engineers are adjusted more (20%) than developers (15%) and senior positions (Tech Lead and Sr. Product Engineer) are adjusted the most (30%).
Salary Data by Position
The table below contains all of the data I collected:
As it turns out, an optimal, self-sufficient, highly talented team in Los Angeles will cost you a bit over $1.1 million per year.
Some Thoughts
First off, $1.1 million is a non-trivial amount of money to spend on building software. If you were selling that software for $100, you’d need to move 11,000 units annually just to cover the development costs. Hence, you better be building something worthwhile.
Second, recruiting such a team will likely be time consuming and expensive. An industry standard recruiting fee for a technical position is about 20%. Therefore, even if you assume needing recruiting help to get half the team staffed, it’ll cost you over $110K.
Another thing to consider is productivity. Boosting this team’s productivity by 1% will save you $11K annually. Or, looking at it from a different perspective, you’ll come out ahead if you spend less than $11K to boost productivity by 1%.
Are there tools you can buy for less than $11K that’ll yield such a benefit? Probably. What about perks like free lunches? Or free dry cleaning? Or paying an admin to do the timekeeping for the entire team? Or letting people work remotely?
Finally, consider how much overhead (i.e. management) you can justify within your technology group. 10%? 20%? If you only have one product team, 20% overhead works out to about $220K per year, or just enough money to pay for a CTO (maybe). Throw in another product team and you could perhaps justify a VP of Engineering (or Infrastructure). And so on and so forth.
You may also like:
Did you love / hate / were unmoved by this post?
Then show your support / disgust / indifference by following me on
Twitter!
This post got 4 comments so far. Care to add yours?
Very interesting analysis. The point about improving productivity by 1% being worth 11K really hits home when thinking about spending vs saving for software groups. Productivity tools and other such tools at a couple hundred per are really hard to argue with.
Thank you for your comment Erik.
The amount of money that Software Engineers make over there is considerably more than what they make here. I was interested in getting into that field, but was put off by a number of people, one being my programming teacher.
Hi Mike, thank you for your comment. I’m sorry to hear that a bad teacher had such a negative impact on your interest in programming. It’s unfortunate, but not fatal. If you still have interest in writing software (professionally or not), there are many places to learn. I’d start with code.org.
Good luck!