Skip to content
Back to Blog

Good Decisions Require Hard Work

"The difference between a good decision and a bad one is often the time you spent thinking about it." In the heat of daily operations, it is always tempting to decide quickly on the spot. Giving an immediate answer in a meeting, choosing the first technology that comes along, solving something right away to clear the desk and move on to the next pending item. The problem is that structural decisions made in a rush almost always return months later... and charge compound interest in the form of rework, bad contracts, and team burnout. A good decision takes effort at the start precisely to save headaches down the road. ## Why impulse tricks us: Kahneman and Annie Duke In the classic work Thinking, Fast and Slow (2011), psychologist and Nobel Prize winner in Economics Daniel Kahneman explained that the human mind operates in two distinct modes: - System 1 (Fast): automatic, intuitive, emotional, and with no apparent effort. It is excellent for avoiding an obstacle in traffic, but dangerous when applied without brakes to complex decisions regarding architecture, budget, or hiring. - System 2 (Slow): deliberate, analytical, logical, and laborious. It requires conscious energy to compare data, question premises, and calculate second-order consequences. Since System 2 takes effort and consumes mental energy, our brains try to answer complex problems using System 1 shortcuts. Worse still, as researcher and former world poker champion Annie Duke explains in the book Thinking in Bets (2018), in the corporate world we are frequent victims of so-called resulting, which is the habit of judging the quality of a decision solely by the final outcome while ignoring the luck factor. If someone made a reckless decision without risk analysis, and by pure luck nothing blew up, the company applauds the "agreed" agility and repeats the recklessness until the day the bill comes due. ## The conscious decision framework To force analytical thinking into the choices that truly matter, I run every relevant decision through five mandatory questions: 1. What real problem am I solving? It seems obvious, but this is where most projects stumble. Many wrong decisions are born from a brilliant answer to a poorly defined problem. Before discussing which database to use or which vendor to hire, write down in two lines what the concrete business pain is that needs to disappear. 2. What are the real options on the table? There is no true decision when someone presents only a single take-it-or-leave-it path. Force yourself and your team to outline at least three viable alternatives, even if one of them seems like the favorite at first. Comparing real options forces you to see the problem from other angles. 3. What are the explicit risks of each option? In engineering and management, no choice has only positive sides; every architecture or strategy is a choice of trade-offs (what you accept losing to gain what is a priority). List the risks and maintenance costs of each alternative without sugarcoating. 4. What happens if I do not decide right now? Sometimes, delaying for a week to run a practical test and collect missing data is the smartest decision in the room. Not deciding in the dark is also deciding with method. 5. Who else should be involved? Impactful decisions should rarely be made in isolation. Who has the context of the actual operation? Who will sustain this choice at two in the morning if things go wrong? Who has the technical repertoire to question your premises before the contract is signed? ## Beware of common biases and the Premortem technique Even when following a script, we need to watch out for three classic traps described by Kahneman: - Confirmation bias: looking for and valuing only the numbers that support the solution you have already fallen in love with. - Anchoring bias: letting the first price, deadline, or architecture mentioned in the meeting limit all the rest of the debate. - Availability bias: overvaluing an isolated event simply because it happened last week and is fresh in memory. A powerful tool to break these biases in a group was created by cognitive psychologist Gary Klein and published in the Harvard Business Review (2007): the Project Premortem. Before sealing the final deal, gather the team and say: "Imagine we are six months in the future and this decision of ours was a complete failure. Write down in five minutes exactly what went wrong." By legitimizing the search for flaws before execution, the Premortem reduces optimistic blindness and reveals weak points that no one would have the courage to point out in a regular meeting. ## Document important decisions Finally, structural decisions deserve written records (such as Architecture Decision Records, or ADRs, in software engineering). Why? Because a year from now, when the scenario changes or something turns out differently than expected, no one will need to witch-hunt or guess the past. It will be recorded: what we knew on that date, what three alternatives were evaluated, what limits existed, and why we chose that path. This has never been bureaucracy. It is protection for those who build and accumulated learning so that each new team decision is smarter than the previous one. > Continue deepening your knowledge: also read Categorization and Priority and Culture That Protects or Sabotages. Also learn about the work in Executive Engineering Leadership and the Case Studies. --- ### References and verifiable readings - Kahneman, D. (2011). Thinking, Fast and Slow. Farrar, Straus and Giroux. - Klein, G. (2007). Performing a Project Premortem. Harvard Business Review, September 2007. - Duke, A. (2018). Thinking in Bets: Making Smarter Decisions When You Don't Have All the Facts. Portfolio / Penguin Random House.

Cesar Zanis / Management & careers

Good ideas keep the conversation going.

Share this article or publish the visual deck to your Stories and Feed.

WhatsApp ↗LinkedIn ↗X ↗
Cesar Zanis
Written by·Software Architect & Technology Leader

Cesar Zanis

15+ years leading software architecture, database performance, FinOps, and applied AI across enterprise operations. Built and scaled an engineering organization from zero to 20+ developers.

15+ yrs experienceArchitecture & DataCloud & FinOps0 → 20+ Devs

Enjoyed this article?

I publish practical analyses on critical systems architecture, Cloud FinOps, data, AI, and engineering leadership directly on LinkedIn.

Continue Reading

Take this analysis further with data and sources

All visual cards of this article are ready above to share on your Stories, Feed, WhatsApp, or LinkedIn with full sources.