The Politics of Engineering

Dealing with the oil and the water of the business.

There’s an old joke:

How many engineers does it take to change a lightbulb?
5 — 1 to change the bulb and 4 to say they could do a better job.

I love this joke, mostly because it’s true. When you’re a young engineer, there’s a lot of boastful claims about who can do what better than whom; theres generally a foundation of camaraderie and it’s pretty fun to poke at your friends.

But later, after we’ve entered the enterprise arena, others can sometimes feel threatened by a competing idea. Especially when the new idea is better. In this case “better” is a nebulous composit of “more efficient”, “less expensive to run”, “reduction in support hours”; IE: higher quality.

I’m going to share my method for dealing with this issue.

The Mechanics

There’s no room for opinion in the field of engineering — full stop.

There is only the issue and the solution. Then, enter the humans and the process of continuous improvement becomes unpredictable.

The Issue

Example: Our current HRIS System:

• Licensing is Expensive
• HR staff complains about data accuracy
• Applicants hate it
• Costs too much to support
• $reason-{1..n}

Net-result: the company is paying for a bad user experience and data — people don’t want to apply at our company because of this system. No people, no company.

But, equally as important, the cost of HR staff having to manually fix the issues created by the current HR system are estimated to be around 10,000 hours annually.

The Process

We have a few constants to work though.

• The internal perception is that the HRIS system needs to be replaced.
• There are 2 analysts working on the problem so the company has options.
• The company will only settle on one solution.

The New Issue

AKA: a tale of 2 analysts…

Analysts, x and y, both set out to solve the problem. But, they have both come up with different answers for the same question. I now present the solutions for the fictitious HRIS system that sucks:

This is a Counting Operation

In this case, we really only need to look at the final row. HRIS-2 only checks 5 boxes, while HRIS-1 checks 9. Here’s why no other part of this work matters:

“All work is a form of thermal expense.”
Henry Adams

In this case, underneath it all, we are really saving HR toil; the human equivalent to CPU cycles. If one solution saves 10 HR staff members 1 hour of time/week that’s a total of 100 hrs/week; multiply that times 4 weeks/month or 52 week/year and you have time savings around 5,000 hours/year.

Get Todd Thomas’s stories in your inbox

Join Medium for free to get updates from this writer.

Remember me for faster sign in

Humans can’t process anything but food, which is grown with seed, water, soil and… the sun. Now we’re back to thermal expense. Put another way…energy lost or used during a process is defined:

Politicin’

We’re just going to say that both analysts have struggled to do their best work. They both have an expense in time and believe that their solution is best solves the current issues.

But, Analyst-y golfs with the director ultimately tasked with the decision between HRIS-1 and HRIS-2. Analyst-x does not golf; but, by now we’ve gotten the first formations of an impression that Analyst-x is a more studious engineer and knows something Analyst-y does not:

You’re either playing politics or engineering — but you ain’t doin’ both.

Time is an unrecoverable expense. Time golfing with the Director can’t be recovered, then re-spent on the analysis of the solution. Once time is gone, it’s just gone.

Here’s the general solution…

Efficiency Gains

Since we’re going to reduce everything down to a common denominator of W (work), there’s a simple way to solve all issues: measure time.

Currently: we expend 10,000 hours supporting an HRIS system.

• HRIS-1 suggests we spend 5,200 hours per year.
• HRIS-2 suggests we spend 7,500 hours per year.
• Proposed State: HRIS-1

This spreadsheet calculates efficiencies gained by comparing the current state to the prosed state. There is now a 443% efficiency gain on the table, allowing the HR team to move approximately 5 times faster by not having to waste so much time on process that should be handled by a computational system.

This method of determining a number from a simple formula removes all opinion from engineering.

math = -opion

Were it more a complex formula then we could have a conversation about inherent human bias wheedling its way into data formulation; but this ain’t that.

While there is a case to be made that the proposed future state is based on estimates both HRIS 1 & 2 would then be equivalent in that way.

But, we do know the current state is based on historic data and it’s untenable; the company has to make a move.

Also, neither solution is perfect. Sometimes, the best we can do is 9/10 requirements met. 10/10 might require building our own in-house which generally doesn’t pass the simplest feasibility tests.

The Reality

HRIS-2 could still be selected. That’s just how things work out sometimes. But, here’s what’s been accomplished, we’ve set a mathematical marker, and said:

This is the new bar for success. Meet or exceed these results.

If HRIS-2 is selected, it would be automatically known that that selection criteria was based on an additional, unknown, factor and was apart from saving the company time or money.

Here’s what Analyst-x knows that Analyst-y never figured out. The number is the number.

Analyst-x doesn’t need to say anything further, ever. There is nothing more to add to this conversation.

In fact, adding anything to this discussion does two things

1. disables or, at least, diminishes the power of the objective outcome.
2. converts the results from an conversation based on objective (measurable) results to one that is policitcal in nature.

Once we’ve allowed a technical solution to become political, the game is already lost.

Knowing this, Analyst-x can move on to the next solution, go fishing or buy a Playstation. Because this is done. Analyst-x has done their best and, for whatever reason, the solution didn’t make the cut. Analyst-x is free.

But that director is now in a bind. If it ever bubbles up to the CEO that he could have trimmed costs/time, moved the companies ball forward radically — and didn’t… well, sometimes consequences are built-in.

Feel free to use the use the calculator if you think it might help.