Automate Everything

I do quite like the “automate everything” principle. Some have even accused me of being “obsessed” by it in the past! Personally, I see this as a good thing, but why? Lets explore that.

Reasons for automation

Removes hamster work

Let’s get the most obvious one out of the way. If you can automate a task, then doing so is a one-time job, whereas doing that task manually clearly means you need to continue to run your hamster wheel in the future. Yes, you might feel productive, but you are not actually making any progress.

Provides real-world learning opportunities

It is often easier to just get the job done. Automating that task, however, will help you to learn transferable skills that will help you in every area of your life, including professionally. Doing that task manually just means that you did the task manually, yet again. Nobody is going to be interested in that.

Potentially increases pay

Those transferrable skills could lead to an increase in pay and, potentially, working conditions too. The type of skills you learn by automating elements of your own life can help to develop DevOps skills, or the ability to manage both software development and IT infrastrucuture. Jobs in this field are widely viewed as among the most in-demand in the sector and, according to Glassdoor, can pay an average of £46,878 per year, with highs of over £70,000. According to CIO, some DevOps Engineers earn as much as $184,750, though this is clearly with several years under your belt.

Potentially improves working conditions

That’s pay, what about those working conditions I mentioned? Working-from-home has been steadily becoming more popular, particularly in the tech sector, since the dawn of the Internet, and especially since broadband connections became a thing. The events of 2020 accelerated its adoption, some estimate by as much as a decade, since we have collectively and irrefutably proven that working-from-home does work. I might even venture to suggest that any organisation which does not offer such options should be avoided. The requirement to go to an office, particularly in tech roles, could very will mean the organisation has a low-trust, micromanaging and ultimately toxic culture that you will probably hate.

For roles that could be done entirely from home, any organisation that demands anything more than than 3% (about 7 working days per year) of your contracted hours be spent in the office raises a big red flag in my mind. Even then, office attendance should only be for a good reason, such as a VIP meeting. No, “Because it’s Tuesday,” and “Because I like to get hold of you easily” are not a good enough reasons to visit the office. If you improve your automation skills and can land a job in tech, you might never see another ruddy office, or the inside of a stinky commuter train, ever again.

According to McKinsey,

More than 20 percent of the workforce could work remotely three to five days a week as effectively as they could if working from an office.

McKinsey & Company, “What’s next for remote work“, 2020-11

In their March 2020 report, ZenBusiness found that,

99% of remote workers report it has had at least one positive impact on their life

ZenBusiness, “Secrets of a Remote Worker“, 2020-03

Furthermore, the Royal Society for Public Health have found that,

45% of people felt working from home was better for their health and wellbeing

Royal Society for Public Health, “Survey reveals the mental and physical health impacts of home working during Covid-19“, 2021-02

What would you gain by ditching the commute?

Saves time

It is not a nice thought, however, you have but a finite amount of time on this planet and you can’t get it back. They don’t even let you reload from a checkpoint! Do you really want to spend that time doing tasks that a machine could have done for you? Think of some of the most boring tasks you do. Is there anything else you would prefer to spend that time on? Of course there is, so automate the boring stuff and get on with enjoying life.

For example, a 2 minute task that needs to be done weekly will cost you 104 minutes per year. If that task takes you 60 minutes to automate, you just saved yourself 44 minutes per year. You might find that you need to do far more laborious tasks at a much greater frequency, so even more time could be saved if you invest it in automating the process.

Human potential

The amount of time that we all spend on tasks that should be done automatically removes us from achieving our true potential. How many are wasting their talents in a sandwich factory when they could be offering so much more?

Arguments against automation

Does people out of a job

It is perhaps a sad fact that many occupations are at risk of being automated out of existence; some have already disappeared in just the last decade. This puts the individual through hardships that they might not have foreseen which can be stressful to say the least.

I have a bus driving licence and so have spent some time in the bus & coach industry. Many drivers are looking forward to a long and happy career behind the wheel, because their trainers are mostly “old boys” who are nearing the end of thirty or more years doing the same. When I tell them not to expect the profession to be around for more than five years, ten at the most because, you know, self-driving vehicles and all, I am usually met with disbelief. This response is arguably more dangerous than going into such a field in the first place, as it means that one is not able to plan for future difficulties.

It’s not all doom and gloom on this front, however. Those who act now and learn skills that are harder to automate stand to gain tremendously. For example, a well paid bus driver generally earns less than a poorly paid DevOps Engineer, and has worse working conditions to boot.

This is not a new phenomenon either. In the 19th century, the Luddites protested powered looms by destroying them, as these machines were removing the need for weavers. What they could not destroy, however, was the idea, so people just made more powered looms. One can either, therefore, make the choice to cling to the status quo, or to embrace change and the opportunities that it brings.

Furthermore, automation does not necessarily lead to redundancies. Once a task is automated, many organisations move staff to areas where they can add more value, as they realise that the in-house knowledge that their people have accumulated is worth keeping.

Not always appropriate

In situations that demand human to human connection, often referred to as “high touch environments”, automation is, indeed, often not the best strategy. These scenarios are fewer than one might think; many consumers are often more inclined to interact with a well designed system than to have to deal with a person. Take the supermarket checkout, for example. I love my Bose noise cancelling headphones, so the self-service till allows me to conclude my purchase without interrupting the latest episode of 99% Invisible.

The care industry, however, is one that is widely regarded as harder to automate, as we all need human to human contact as part of a healthy and happy life. While assistance for care professionals can be automated, I doubt that the sector will vanish any time soon.

Costs money

Another argument I have heard is, “But don’t you just have to pay for the machine that does the job for you?” Yes, yes you do, but is your life worth less than the £40 you would spend on a Raspberry Pi to automate your personal drudgery?

From the commercial perspective, automation often presents a high up-front bill, which can make some accountants twitchy. As mentioned, tech-ie types are expensive people, so that often encourages businesses to stick with their existing manual processes. Unfortunately, this fails to take the hidden cost of not automating that process into account, which is almost always significantly higher. Traditional accounting can only record numbers that are actually spent, so ignores the true cost that these incur.

For example, I worked on a project where we had three members of staff being paid minimum wage to just process returned items into the company warehouse. Their manual process was as follows.

  1. Pick an item from the pallet
  2. Copy the model number into the spreadsheet
  3. Copy the serial number into the spreadsheet
  4. Put the item into the test bucket or the scrap bucket.
  5. Repeat

If you only look at the short term, the cost of £1,678.49 per person per month sounds manageable for most organisations. Over the course of five years, however, the total wage bill would stand at £302,127.60.

I worked with several different vendors to produce a proposal for a machine that would have allowed those staff to do tasks that add more value. I proposed the following process.

  1. Staff tip a pallet of returned items into a hopper, then go do something else.
  2. The machine drops an item from the hopper onto the conveyor belt
  3. Item is photographed on both sides using a top & bottom camera array
  4. Optical character recognition is used to identify the model and serial number
  5. Item information is recorded in the database
  6. Items are automatically sorted into the correct piles.

The quotes we received to design and build this system were in the £200,000 ball park, which would have had to be paid up-front as I mentioned. This was deemed too expensive to progress the project, even though this constitutes a saving of over £100k over 5 years! That’s effectively a return on investment of approximately 50%, far more than the 1% you would get by putting that £200k in a savings account or even the 7% you might expect from the stock market.

Unlike other investments, you know what your wage bill you be over five years and you know what the automation will cost to implement. This means that your RoI is almost guaranteed.

Perhaps one factor in this is the general human inability to equate money saved with money earned. For example, if you were to purchase a box from me for £300 (it’s a pretty awesome box) and then I were to give you a £100 gift, you would feel like you had won £100. But if I sold you my £300 box for £200, you might feel like you just spent £200. In both cases you end up in the same position, so this psychology is both fascinating and misleading.

Conclusion

It is almost always worth investing in automating some or all of a process, as it will save you time and money in the long run. It provides you with skills that can help in other areas and allows you to lead a more fulfilling life. What is there not to like?

Leave a Reply