Scorchers, Part 1: Tools and Burn Rate

Jason SachsApril 12, 20165 comments

This is a short article about one aspect of purchasing, for engineers.

I had an engineering manager once — I’ll leave his real name out of it, but let’s call him Barney — who had a catchy response to the question “Can I buy XYZ?”, where XYZ was some piece of test equipment, like an oscilloscope or multimeter. Barney said, “Get what you need, need what you get.” We used purchase orders, which when I started in 1996 were these quaint forms on colored paper with a number on top, that you’d have to sign out from a book. Each month they’d switch the paper color; sometimes they were pretty gaudy shades of magenta or green. We’d call up Digikey or McMaster-Carr and tell the sales operator what we wanted to order, and give them the magic number on the purchase order, fill out the form, then get a signature from someone with Purchasing Authority, and put it into the Purchase Order cubbyhole by the receptionist’s desk. If you wanted to be sure you could get the signature first, you’d get a quote first and then wait until after you had a signature before actually buying something. I didn’t have Purchasing Authority, but some of the senior engineers could sign for things below \$1000, and you’d go to the engineering manager or the head of Purchasing for larger orders. I can still remember my first few POs — they were for strain gauges and some Analog Devices 1B31 strain gauge signal conditioning modules, these big blue potted things in an ultra-wide DIP package. Get what you need, need what you get. In other words, Barney the engineering manager would approve pretty much anything that went across his desk — he just wanted you to know that he expected you to get things because they were important to get the job done, and not because you just wanted to buy the latest greatest things. It wasn’t just a magic piece of paper; real money changed hands. I suppose Barney would have probably questioned a \$50,000 oscilloscope, but if you really did need it, he’d sign, and the smaller stuff was no big deal.

In later years at that company, Barney left, and the bean-counters in Accounting gained more influence, and although you could still buy things when they were necessary, for test equipment and software, you had to argue more often. (Oh, and the purchase orders changed; they were less quaint-looking, and became just another byproduct of QuickBooks or whatever software they used.)

At my current employer, I have had a couple of coworkers recently who wanted to buy software tools; one wanted to buy the Pro version of Beyond Compare 4 for its three-way merge, and another who wanted to buy WinGrep, a Windows GUI for global-search-and-replace. (Yeah, you in the Tux the Penguin T-shirt, don’t laugh, not everyone is productive when using the command line.) Both of them talked to me about it… and what struck me was their lack of confidence. I can’t remember the actual words, something like “Well, I could kind of use this software, do you think I should get it?”

Let’s stop a second. We’re going to do a little thought experiment, and crunch some numbers.

Suppose you are an electrical engineer with four or five years of experience out of college, in the USA. Suppose you’re in an office in… I don’t know, Albuquerque. You’re making… maybe \$70K per year? How much is your time worth?

$$ \frac{\$70,000}{1\, {\rm year}} \times \frac{1\, {\rm year}}{52\, {\rm weeks}} \times \frac{1\, {\rm week}}{40\, {\rm hours}} = \$33.65 / {\rm hour} $$

Wrong!

That’s your salary. The real question is, how much is it costing your employer? There’s something called the fully burdened labor rate, which is the total amount of money required to support your work. In addition to your salary, you’ve got:

  • Payroll taxes. USA has the FICA taxes funding Social Security and Medicare. As of 2015, you pay 6.2% for Social Security and 1.45% for Medicare, for a total of 7.65%, out of your gross salary, but your employer pays another 7.65%.

  • Overhead. You’re working in Albuquerque. It’s a satellite office, the headquarters are somewhere trendy in Silicon Valley. You are one of 10 engineers, along with an engineering manager and an office manager or administrative assistant. The engineering manager makes more than you, the administrative assistant makes less than you, let’s just call it even and say the company is paying the equivalent of 12 engineers’ salaries for 10 engineers worth of work. So that’s 20% extra overhead. Throw in some work from the HR department, Finance, Legal, company executive officers, maintenance staff, and maybe we’re up to 33% overhead per engineer: in other words, for every 100 engineers directly doing work, we have to have 33 of these other workers. (This is a factor we’re going to multiply at the end.)

  • Office space. You’re in Albuquerque so your company pays the bargain rent of \$15/square foot per year for a 2400 square feet satellite office. That may sound like a lot of space (200 square feet per person! That’s about a 14 foot square office) but you just get a 50 square foot cubicle and a lot of space goes to hallways and conference rooms and the telcom closet and the engineering lab. Okay, total cost = \$36000 per year divided by 12 people = \$3000 per year, so another 4.3% overhead. In the trendy Silicon Valley office it might be 3-5 times as much.

  • Health insurance. You pay part of your health insurance premiums, but your employer most likely pays the bulk of it. Let’s call it \$5200/year since you’re only a few years out of college, so you’re single and you don’t have a family. That’s another 7.4% overhead.

There’s also retirement plan (pension plan or 401K matching), utilities costs (electricity, water, heat, telephone, Internet), office supplies, the annual company picnic, et cetera. You get the picture. You work in a thrifty company so maybe it costs 25% more to cover payroll taxes, office space, health insurance, etc., and then we need to multiply by 1.25 to cover the overhead of all those people like managers and HR and Legal who don’t directly do the engineering work. Total overhead: 1.25 × 1.33 = 1.66 → 66% extra for fully burdened labor rate. It might be 150% extra in a trendy company office with lots of perks, paid family leave, free lunches, foosball tables, and so on.

So your \$70K/year that you see on your pay stub turns into a fully burdened labor rate of \$70K × 1.66 = \$116K/year.

$$ \frac{\$116,000}{1\, {\rm year}} \times \frac{1\, {\rm year}}{52\, {\rm weeks}} \times \frac{1\, {\rm week}}{40\, {\rm hours}} = \$55.95 / {\rm hour} $$

Still wrong!

The \$116K/year is right, but you’ve got 10 paid holidays and 10 paid vacation days per year. (Your colleagues at Snazzy.com in San Francisco with a few years experience get 20 paid vacation days per year. They also get 3 paid sick days per year because that’s the mandatory minimum in California. Sorry.) So that’s down to 48 weeks.

$$ \frac{\$116,000}{1\, {\rm year}} \times \frac{1\, {\rm year}}{48\, {\rm weeks}} \times \frac{1\, {\rm week}}{40\, {\rm hours}} = \$60.61 / {\rm hour} $$

Almost there.

You don’t spend 100% of your time actually making progress on your deliverables. Some of your workday gets taken up by time walking from one office to another, using the bathroom, and going to meetings, training programs, team-building exercises, and so forth. Oh, and surfing the web. That’s important; Hacker News and reddit/r/programming help you stay current on the latest techniques. Yep. Let’s be generous and say that 75% of your time actually makes progress towards deliverables, so that’s 30 hours a week.

$$ \frac{\$116,000}{1\, {\rm year}} \times \frac{1\, {\rm year}}{48\, {\rm weeks}} \times \frac{1\, {\rm week}}{30\, {\rm hours}} = \$80.82 / {\rm hour} $$

There. \$60.61/hour to be at work for 40 hours, \$80.82/hour for the time that you’re actually making progress on something.

Now, let’s see… Beyond Compare Pro costs \$60 for a single license; WinGrep costs \$30 for a single license. You could buy both and it would be equivalent to the cost of 67 minutes of your productive work. So if they save you a little more than an hour of your time and you never use them again, they’re still worth the cost for your employer.

Item number 9 on the Joel Test is Do you use the best tools money can buy?, and this is why. Productive engineering time is a precious resource! If you can save some of that time by purchasing tools or test equipment, it’s worth it!

So next time, if you see some software tool or test equipment item that’s under \$100 and there’s a strong chance it will save you a few hours of effort, JUST BUY IT. (For the engineer with the numbers I calculated above, just wavering back and forth for ten minutes whether to buy something already wastes \$13.50!) If your manager gives you a hard time, ask him or her what factor you should use to calculate the value of one hour of your productive time to your company.

The above numbers are just estimates for one contrived situation in the United States, but no matter where you live or what company you work for, you can still do these kinds of calculations.

My personal preference — if I had the choice — would be to get a yearly budget of 0.5% of my gross salary to put towards any productivity enhancements of my choice, with the goal of saving time that otherwise would have been spent on repetitive tasks that a software tool or test equipment item could do more inexpensively. For some reason, some companies seem to like not spending any “additional” money, and instead view engineering salaries as a sunk cost that they should just use to get things done, whether or not they could be used more efficiently. This doesn’t make sense. Salary may be a fixed and committed cost if they keep their engineers employed, but given the high cost of labor, there’s a strong incentive to get more work done for the same salary, and good tools are an easy way to accomplish that.

Get what you need, need what you get!

© 2016 Jason M. Sachs, all rights reserved.


Previous post by Jason Sachs:
   Padé Delay is Okay Today
Next post by Jason Sachs:
   Round Round Get Around: Why Fixed-Point Right-Shifts Are Just Fine

Comments:

[ - ]
Comment by SFriederichsApril 12, 2016
You took the thoughts right out of my head! I even fight this battle against myself whether at home or at work despite the fact that I'm never unsatisfied after buying a new tool.
[ - ]
Comment by microsoftware.grApril 12, 2016
I think that it is very important to include in this research of your article not only the “how much is it costing your employer? “ but and the Maslow’s hierarchy of needs pyramid!
https://en.wikipedia.org/wiki/Maslow's_hierarchy_of_needs

Employer has to motivate his employee to do a job, some times the cost per hour may be high but:
1. You , the employer ,will have a self actualization of your employee.
2. Your employee received new knowledge that will be helpful at the future for your company.
Some times it is needed to be out of :
“Get what you need, need what you get!”. It is not a permanent law I find. You, the employer, put some water in your wine .
[ - ]
Comment by Seb85April 22, 2016
Great analysis.

I think there is also a problem of a generation that expected everything to be available on the web, and everything on the web to be free. (I have seen people complaining about having to pay 0.99c for a smartphone app, regardless of how good it is. Although that may be changing now as subscription services become more ubiquitous.)

And, while there are some excellent open source tools out there, it can be worth paying for commercial tools to better quality control (e.g. the use of automated test suites and coverage tools), documentation and professional support (with guaranteed service levels, etc.)
[ - ]
Comment by sbkennJune 11, 2016
I once had to argue to buy 3 chip-clips so I could hang my scope probes while I looked at the schematic to see what should be happening, instead of risking shorting two pins of the chip while I held the thing in place !
This was often on $30k minicomuter PCBs, that is $30k per board for replacements ! Hard to imagine these days.
[ - ]
Comment by MatthewEshlemanNovember 7, 2016

I worked state-side while interfacing and working with a team of software engineers in Japan. My state-side boss was always great and fully understood how critical is was to acquire the right tools for our team. Your post brought back some memories, so inspired by this post, I wrote an entry about my boss here: "It makes engineers more productive". But another related story. Our team enjoyed a nice collection of high-end build servers. Our most complicated feature rich embedded application could take nearly an hour to build on a regular single threaded PC, yet on a 16-core build server hitting it with "make -j16", the same application could be compiled from scratch in under 5 minutes. 

With that context in mind, at some point I spent a couple of weeks in Japan working with their software engineers to bring up a new SoC and new PCB. The engineers in Japan were clearly under the thumb of a penny pincher. Even small partial rebuilds on their PCs were taking 5-15 minutes. My laptop was better! I made sure to "stir the pot" some, repeatedly bragging about our 16-core build machines....

A few weeks after returning home, I received an email from one of those engineers looking for more details on our build machines and some statistics on the build times we were enjoying, specifically so that he could present the "evidence" to his boss in Japan. :-)   Mission Accomplished!


To post reply to a comment, click on the 'reply' button attached to each comment. To post a new comment (not a reply to a comment) check out the 'Write a Comment' tab at the top of the comments.

Registering will allow you to participate to the forums on ALL the related sites and give you access to all pdf downloads.

Sign up
or Sign in