Creating Process is a Dangerous and Expensive Option. Anecdotes and Guidance for Leaders.
Building processes is a mechanism for resolving issues. However, processes are also bottlenecks which slow your organization down. How do you choose when to add process?
Hey all! This newsletter is being sent late this week. I intend to send them on Thursdays on average, but my wife and I were traveling in New York, and I decided to wait until we were back before sending. I figured everyone could be patient.
I've been repeatedly asked for my thoughts on Amazon, advice on layoffs, and how individuals should handle their careers. It's far too personal of a topic for me to give specific advice. A few quick general things.
Companies are emotionless money gathering entities. I like Amazon's leadership principles. I think they're a great mechanism for building a strong money gathering entity. I don't think 'Amazon' gives a damn about me, or about you. Don't ever forget that you have a completely pragmatic relationship with your company. They'll fire you when it's convenient, and you should fire your company when it's convenient for you.
Worry about layoffs starts with risk management. When people write to me worried about layoffs at their company, what is their #1 concern? It's always money. They're worried about losing their income. If you want to stress less about losing your job, you only have one option. Save more. How do you save more? Spend less than you make. Oh yes, there are a thousand caveats and excuses and privilege wrapped up in that statement, but it's a mathematical truth. The way you reduce risk to yourself and your family is that you save more money. I know the tone goes overboard, but MMM's article on how to think about debt has always struck me as insightful.
If you've recently been hit by layoffs and need some advice on interviewing, I'd suggest you chat with Jake Casey. He was also hit by the Amazon layoffs, but as an Amazon Bar Raiser and long time recruiting leader, he has opened up a new interview training business. If you'd like to brush up your interview skills, it could be well worth your time to talk to him.
I hope you're having a lovely 2023, and I'll be sending you another newsletter in just a few days!
This anecdote takes me back a few years. Our broader organization made it through years of increases in traffic without major incident. At Amazon, that means that each service handled doubled or tripled traffic repeatedly over the years. Then one Q4, a critical team hadn't properly tested a new service. The new service didn't handle the increased load, and it took days before the related system could be upgraded enough to handle the load.
It was a painful business event. Customers were impacted, engineers stayed up late, there was a financial impact, and messages explaining the situation had to be sent to many departments. Not big enough to be noticed in an earnings report, but there was metaphorical egg on the face of the organization's business leaders. They needed to do something.
This article is about the need for leaders to do something, the cost of that something, and what you should consider when making these decisions.
Doing something! Anything really.
Perhaps the worst thing a leader can do for their career is shrug, and say that there was nothing they could have done to avoid the event. So, being leaders who didn't feel like killing their careers, they turned to their teams, and demanded something be done. "How could this be prevented in the future?"
The responsible team's leadership knew how to respond to events like this. Amazonians in particular love to point at missing mechanisms. And mechanisms thankfully are rarely about a specific person or team. Mechanisms are scalable ways you can resolve a problem forever. It's a way of taking ownership over an issue, without suggesting that a human made an error. It wasn't an error, of course, it was simply a missing process. We were all at fault.
The team said that the missing mechanism was a process for preparing for Q4 load. Teams did their preparation ad-hoc for Q4. Ad-hoc is a coded way to say randomly, which is a bit like waving a red flag in front of your leaders. You're daring them to intervene. The team proposed that the only way to avoid inevitable painful misses in the future was to create a centralized Q4 operational planning process. This team gracefully volunteered to lead the effort to create a process.
To make a long and costly story short, the organization created a Q4 operational planning process. As the next Q4 approached, every team in the organization had to prepare a document, run a set of load tests, run other availability tests, review their alarms, give a presentation in front of their leadership teams, and so on.
What were the results of doing something?
Did it save the overall organization from another painful (and embarrassing) outage? Hard to tell. Maybe? Perhaps not. I don't think there was a major outage that next year, but there were in future years. It's difficult to compare the results of a process with a lack of process when the situation is rapidly changing.
Regardless of the actual impact, I assure you that the widely broadcast statement was that our new organizational process was the best mechanism around, and everyone should probably join the crowd.
Why is that? Because in large companies, you're often rewarded for the perception of your actions, not necessarily for actual results.
How are you rewarded for a new process? Your new process clearly saved you from a major outage this year (you can't hear the sarcasm in my voice). You were thinking ahead. You were leading strategically.
How are you punished for the added cost of this new process? You're usually not. It's the cost of doing business. It's hard to measure, and no one would think to complain about it.
What are the real results of doing 'something'?
Large companies tend to reward perception of action.
Large companies tend to ignore the cost of process.
What does that actually result in? Large companies and organizations tend to have a lot more process. They'll call it the inevitable cost of being large. They'll explain that smaller companies couldn't possibly understand the challenges of running a multinational multi-billion dollar blah blah.
Then, a year or thirty down the road, the large company finds that it gets stuff done a lot slower than those smaller companies. Why?
One reason is that processes (required steps / mechanisms that people need to follow) are by definition bottlenecks. They're institutionalized bottlenecks, purposefully put in place to slow people down so that they do the right thing. Mechanisms can be an invaluable tool, but they can also be one of the largest costs to your business.
What are the major costs? I'll walk through them.
Why process is expensive. Less ownership.
Let's keep things simple for this example. You are hired tomorrow by Andy Jassy to increase the sales of the next Kindle. That's it. You have complete control.
Ok! Cool, you have plenty of ideas. You're thinking about marketing campaigns you can run. You'll find influencers. You'll consider partnerships with some libraries. Maybe you could build some Kindles out of new materials! A pewter Kindle! A porcelain collector's edition Kindle!
Wait, he forgot to tell you the processes involved.
You need to declare your advertising budget every Monday. Then on Tuesday, they'll tell you how much you actually get. On Wednesday, you need to present exactly where you plan to spend your advertising. If you'd like to do anything apart from advertising, you'll have to present your idea in a 6-page document.
Do you now feel more or less like an owner of increasing Kindle sales? Obviously, you feel significantly less like an owner. The original scenario was wide open. The new scenario has dates, schedules, and specific expectations.
In the first scenario, I'd expect all sorts of wild ideas. In the first scenario, I'd expect the very best hires to crush it. They'd hit it out of the park with awesome ideas and great success. The worst hires would cause some damage to the brand or business through their crazy actions or excessive decisions.
In the second scenario, I'd expect most people to execute Kindle advertising. They'd have regular reports on their advertising spend. Most people would execute a boring version of advertising. The very best hires would either struggle through and present new ideas of how they'd prefer to do the job, but I imagine a good number of them would run out of patience with the red tape, and leave. The worst hires would not have a great ROI on their advertising, but that would be the extent of their impact.
A process is a purposeful bottleneck. It directly limits ownership by moving decision-making from the individual, to the organization. It limits the ownership and actions of great hires, and limits the downside of poor hires.
Why process is expensive. Standardization.
A great advantage to strong ownership and limited control is being able to take advantage of the creativity and common sense of individuals.
When a team can choose the tools and processes they use, they can pick and choose what works well for them. One development team wants a 2-week sprint, another wants a 6-week sprint. One marketing team wants a weekly design presentation meeting, another prefers a project-based approach. One manager prefers a structured one-on-one meeting with an agenda for their team members, another manager prefers unstructured organic meetings.
What is the right answer? No one knows. Thousands of decisions are being made on teams. If small teams or individuals have the ability to make these decisions, they can continually adjust based on their feedback, how they feel, and what they've learned.
If an organization or company makes a decision, they standardize that decision across many teams. Teams where that decision may be suboptimal.
When you build processes, you often need to standardize things. Standardize task and project reporting. Standardize programming languages. Standardize meeting cadence. Standardize project scoping, ingestion, and response timing. When organizing across teams, you immediately begin to see the need to take control away from those individual teams.
Standards are not a horrible thing. Having electric plugs standardized across the US is valuable. However, every standard should be taken with great care, as it essentially removes the ability and incentive to innovate in that space.
Why process is expensive. Time.
This is the simplest version of process costs. Almost every process (institutional bottleneck) requires some form of human time investment.
You need to fill in a spreadsheet. You need to attend a meeting. You need to prepare a presentation.
Imagine a standard weekly project review meeting. It is an assumed necessary evil in most organizations.
What's the cost involved? I remember an organization of 150 people. Prior to the meeting, every single of those 150 people had to ensure that their tasks were updated before the meeting, with clear (easy to understand) updates written up. Then around 25 people attend the organizational project review meeting. All 25 prepared at least an hour for the meeting. They spent an hour in the meeting.
What's the cost involved? Let's go into the weeds. Likely well over a hundred hours a week were spent on this process. If you assume people are productive 30 hours a week (after breaks, chit-chat, web surfing), that's well over 3 employees worth of time per week, thrown away. That's 2% of the organization spent on that single process. I don't think that's an unreasonable estimate.
Let me repeat again to ensure I'm clear. That organization was 2% less effective because they had a weekly project review meeting. And I'm not going to claim that this process was a bad idea. It was pretty useful to have everyone learn the status of each other's projects, ask questions, debate priorities, etc. However, it had a real cost, and you should be very hesitant to spend 2% of your organization on anything.
What should you do instead?