Amazing CTO Newsletter
by Stephan Schmidt
Happy 🌞 Sunday,
today I want to talk about OKRs. They are something new and you need to use them in that way. They are not the classic number-based goal system with new paint. If you use them that way you are doing it wrong. Agile is not waterfall with dailies (most people with a long roadmap use Agile that way though - that’s why it doesn’t work). Many managers conflate a performance management system (= expectations for employees, ask yourself why not everyone has only As in school) and KPIs (numbers to see if your business is fine, like the RPM in your car) with OKRs. But OKRs are different. When you put numbers in objectives, you’re doing it wrong.
I hear people say “We achieved 80% of that objective”. You’re doing it wrong. Objectives are binary. You either achieved them or not. If not, they need to be at the top of the list for the next quarter, not tossed away. If they are important they can’t be skipped. If they are essential, 80% won’t do it. Objectives are key steps in your strategy to achieve your vision. You can only toss them away if they were not key steps in your strategy from the beginning - and you’ve used them wrong then.
Happy working on your OKRs.
On to this week’s insights
- 🦹 People can read their manager’s mind
- 🤖 How to Succeed as a CTO
- 💻 Correlation between the Use of Swearwords and Code Quality?
Good reading, have a nice Sunday ❤️ and a great week,
CTO-Coach and CTO-veteran
If you only read one thing
People can read their manager’s mind
OMG yes “People generally don’t do what they’re told, but what they expect to be rewarded for. “ Print that out and hang it over your desk. This is why OKRs fail. Or QA.
Stories I’ve enjoyed this week
How to Succeed as a CTO
“Many technical founders haven’t really recruited before. It’s important to stress that recruiting is your job as a VPE/CTO”
Finite vs Infinite mindset
There are two companies: Those that want to achieve some growth with minimal costs and those that don’t see a limit to their growth. When talking to a coachee in a company with the first mindset, I realized that eBay was the second type of company. If you didn’t spend your budget, you failed because you made the company grow less than it could. If your forecast was too low, you failed the company because eBay would have given you more money on your larger forecast to grow faster. It’s the finite vs. infinite mindset.
Is there a Correlation between the Use of
Swearwords and Code Quality[PDF]
Not going to spoil this for you.
My Fifth Year as a Bootstrapped Founder
It is amazing what you can achieve as a solo developer today. I recently discussed with a coachee what a manager is, because his co-founders wanted to press people back into the office (O’m an office guy! - but that ship has sailed). Developers in offices is an old paradigm. The new paradigm is remote work. And much smaller companies. We will see tectonic shifts. And just as most physicists couldn’t deal with Einsteins’ theories, most managers will not be able to deal with this new paradigm. Go read Kuhn btw.
The case for frameworks
Although I disagree with the conclusion, this aims to explain the usage of SPAs with game theory. I think everything that drives the understanding of our industry and why we do things (use massively complex stacks to do simple things) is good and a step forward. I still think the beginning of SPAs was imitating mobile apps and speed. As internet speeds grow and computers get faster and faster (my current SSD is twice as fast as my last one and the next PCI5 SSD is again 50% faster) the argument for SPAs goes away.
Maybe people do care about performance and reliability
A plea for quality.
“More and more I feel like software is dysfunctional because everything is dysfunctional, because complex and interlocking societal factors make excellence a pipe dream.”
WTF is PMF?
PMF=Product market Fit and it is the most important thing for an early-phase startup. Most problems I see come from the fact that a startup is pre-PMF and acts post-PMF. Where are you right now?
More than five whys and “layer eight” problems
Just keep two things in mind: Root cause analysis is like the roots of a tree, a graph not a line. A crisis never happens because of a problem, but because two problems overlap. Traffic to large for your database + failover fails. Database gone and backup doesn’t work. Second, don’t stop with “human error” - humans are never the reason. ”[…] you might discover that management is forcing people to ignore the maintenance schedule while saying things like “it’ll work, trust me”.
The McKinseys and the Deloittes have no expertise in the areas that they’re advising in
Yes. “The McKinseys and the Deloittes have no expertise in the areas that they’re advising in” - consultancies are not for that. They are either body-leasers or the bringer of the bad message you don’t want to bring - so you hire them. What they say, is what you already know.
The 6 Decision-Making Frameworks That Help Startup Leaders Tackle Tough Calls
Decisions, decisions, decisions. I now think the decision-making ability of a startup is what makes a startup succeed or fail. So go get better at decision-making.
What is Sustainable Software?
Personally, I think we should start with cars and planes, but here you go, this will become important in the future. BEFORE your CEO is going to ask you next year, become an expert and show leadership (to your CEO at least).
Technology Has Lost Its Pragmatism
The article is about when to stop coding. But it starts with:
“When I was a kid, my family had a mundane friendly argument over who threw away dad’s back-scratcher. Dad had been using it for years and suddenly it was gone. Nobody was able to answer his question. In fact, nobody was really sure what he was talking about, although the term ‘back-scratcher’ is not that ambiguous.
After a bit of childhood detective work, it was determined that mom had thrown away the spaghetti fork.”
Toolformer: Language Models Can Teach Themselves to Use Tools
Next important step to AI-is-eating-software-development.
“We introduce Toolformer, a model trained to decide which APIs to call, when to call them, what arguments to pass, and how to best incorporate the results into future token prediction.”
We need programming mentors
I’ve been arguing for a long time for a master/apprentice relationship in teams and assigning every senior developer an intern.
A CTO needs to create a culture and organization where people are supported, can learn and grow. Such a culture and a rising tide lift all boats. This is the biggest lever for productivity in your tool belt.
Why is remote work seen as a gift?
Just forget about the offices. This train has sailed. Or the ship. But recently we have seen a wave of tech companies who want to bring developers back to the office, 2, 3, or 4 days a week. This will not work and developers will leave - even if your CEO thinks differently. Or if you want to read a deeper analysis, go read the article.