About Teletype
Join
Daniil Polozov
@jokerdabro
October 13

«Mistakes that cost millions»... Or how irresponsibility leads to collapse

Hey, it's Danial.

Today I want to talk about a culture of responsibility, about being able to get things wrong, and about what is actually going on.

Over the last three or four years, large companies moved from “we can afford to be wrong if it buys us a real jump in growth” to “preserve, hold, and control.” We are in another turn of the crisis, not only in Russia (where I still am), everywhere. There is little wish to take risks, or to put effort into something you cannot control. Follow that logic, and you do not need a crowd of specialists for sharp pushes. You need people who run the operation well, or who grow and hold the current position in small steps.

This may sound harsh, but in most cases high-level specialists are not really needed. Companies have already put on enough middle-management fat to keep moving up. Right?

Or wait... A bit of backstory. Not a 30-minute lecture (just some local humor). Two minutes should do it.

2023. There is hope the economy will recover, slowly. Globalization still looks like a way forward. Companies start trying to simplify how work gets done. AI starts to hype. The early stage of the hype.

2024. The hope mostly does not pay off, but companies still have enough strength left for real pushes into new markets, to take share or lock in what they already hold. Global markets still look fine. At the same time, the experienced layer starts getting cut. The expensive people, expensive on payroll, get trimmed, and interns and juniors are hired instead. Experience is swapped for ambition and a willingness to do the work of four people. AI starts to rage. The “cut costs with AI” trend reaches startups and a few spots inside big companies.

2025. The makeup of teams has already partly changed. There are a lot of juniors. Processes slip out of control, because the experienced people left with both “how this should be done” and “why it works this way.” It is not very visible. COGS, the payroll part of it, covers the revenue you failed to get, and the company even feels a positive trend while pushing AI deeper and deeper.

2026. Companies start forcing AI through as hard as they can, and Uber-izing processes with it. Top management treats this as the main trend. Then the bill for hiring juniors shows up. Those juniors got quickly relabeled as mids, but the knowledge is still junior, or junior-plus. One unpleasant problem comes into the open. Losing experienced people is not made up for by ambition, or by working like three people, and it does not leave you the dataset you would need to train the very AI everyone is chasing. So you take another turn at cutting COGS, cutting payroll, to cover the missing revenue.

This is where the historical bravado should stop. The important part is next.

As we know, total revenue = what we earned minus what we spent. Looks like the right move. Cut costs, book a profit. The math works out, right?

What if I tell you the price of these changes is lost customers and worse service for the ones who stay? What does that hit? Retention, and their average check.

There is another formula for total revenue: total revenue = margin per paying customer × number of paying customers. Follow that, and worse quality means fewer paying customers. The market is brutal, and someone else offers better terms. Margin is not guaranteed to rise either. Yes, you cut the cost per user. Those costs tend to come back, because people inside the company get more expensive every 6–10 months, and a year later the cost of serving a customer is back where it was before the change.

So here is the point. We can rush to cut costs by hiring cheaper labor and by rolling out AI. Do not forget this:

  • Product quality suffers when previous experience is ignored and we hit the same mistakes again. Customers leave. The economy trends down.
  • People who were making 300–400 thousand rubles a month (or the equivalent of about $21,000 a month, and $250,000 a year) no longer put that money into the economy. The people who do not earn that yet do not have the same spending power. Read that as: the average check in the market falls.
  • Decisions made in the offices of large companies often move a country’s economy. Look at taxes, and at the cash the company generates, and the cash its employees generate.

I am not writing this because someone is a villain, and I am not looking for anyone to blame. About ten months ago I started asking about the quality of management decisions, and how long a result takes to show up at different levels of management.

At what point can you tell: I did this, and this was pure chance. The higher the job, the longer you wait to see your change come back, and the larger both the gain and the damage can be.

What I mostly saw was not reflection. It was silence, and a shift of attention, so the whole thing would be forgotten and the people doing the work would not carry a bad mark on their record. In places like that, some companies pay by cutting bonuses. Others pay with whole percentages of the staff. Does anyone carry personal responsibility for it? Most likely not.

I really like Fullmetal Alchemist. It has a simple idea: if you want to get something, you have to give something in return. The principle of equivalent exchange. It says that every action has a consequence, including the small decisions in our lives.

A few studies worth your time:

Daniil Polozov
October 13, 22:33
0 views
0 reactions
0 replies
0 reposts