wrote to the Smithsonian Establishment asking for every part they’d on human flight. He and his brother Orville devoured each paper, concept, and calculation from the main aeronautical researchers of their time. They studied Otto Lilienthal’s glider designs, Octave Chanute’s engineering rules, and Samuel Langley’s tables on raise and drag.
Then they did one thing radical: they questioned all of it.
When their gliders didn’t carry out because the printed knowledge predicted, the Wright brothers didn’t assume they had been at fault. They constructed their very own wind tunnel and examined over 200 wing designs themselves. What they discovered then shocked them. Lilienthal’s raise coefficients—the numbers your complete area had relied on—had been incorrect, and never just a bit bit. Completely, utterly incorrect.
The Wright brothers didn’t query this out of vanity; they had been using first rules pondering. And it’s what made the distinction between their first flight in 1903 and Langley—who had extra funding, status, and assets—crashing his plane into the Potomac simply 9 days earlier.
At the moment’s knowledge scientists face the same problem. We’re surrounded by playbooks, frameworks, and finest practices—A/B testing guides, metrics taxonomies, mannequin choice checklists. These instruments are invaluable. However identical to Langley’s tables, they will lead us astray when utilized blindly.
Essentially the most strategic knowledge scientists I do know don’t select between frameworks and first rules. They use each. Frameworks give velocity. First rules present readability. Collectively, they separate job executors from strategic companions.
What Is First Rules Considering?
First rules pondering means breaking an issue all the way down to its basic truths and rebuilding from the bottom up. Aristotle described it as discovering “the primary foundation from which a factor is thought.”
In follow, it means asking:
- What do we all know to be completely true?
- What are we assuming?
- Which of these assumptions can we problem?
This doesn’t imply rejecting current data: the Wright brothers studied all of the analysis out there; they only didn’t deal with it as gospel.
For knowledge scientists, the identical applies. Frameworks are maps—confirmed shortcuts by way of acquainted terrain. First rules are the compass, maintaining us oriented when the map will get a bit blurry.
Why Knowledge Scientists Want This Now
Frameworks are all over the place in knowledge science for good motive. They assist us run experiments, outline metrics, and construct fashions rapidly. However they will additionally create false confidence.
I’ve seen groups run flawless A/B exams that answered the incorrect query. I’ve seen customary metrics produce dashboards that regarded spectacular however measured nothing significant. These failures didn’t occur as a result of the frameworks had been flawed. They occurred as a result of nobody paused to ask the primary rules questions: What determination are we really attempting to tell? What worth are we actually attempting to measure? Will we even want this stage of complexity?
This issues greater than ever as a result of AI is automating the execution facet of knowledge science. Gen AI can question knowledge, generate visualizations, and apply frameworks brilliantly. But it surely can’t determine whether or not you’re asking the precise query.
First rules pondering is your differentiator. It’s the talent that retains frameworks grounded in actuality—and it’s turning into essentially the most defensible functionality of a strategic knowledge scientist.
The place First Rules Modifications All the pieces
Listed here are three examples the place the info science playbook is totally properly written, however strictly following it mindlessly would produce a poor consequence.
Past Textbook A/B Checks
The framework says: outline your speculation, randomize customers, measure your major metric, examine significance. That works completely—in case you’re asking the precise query.
However the first rules query comes earlier: What determination are we attempting to tell? What uncertainty are we attempting to resolve?
I as soon as suggested a group who was testing a brand new advice algorithm. Framework pondering stated: randomize customers, measure click-through price, run for 2 weeks. Ship the winner.
However a primary rules pause revealed one thing completely different. We weren’t unsure about clicks—early alerts instructed they might go up. We had been unsure whether or not these clicks would drive true engagement or simply noise.
So we modified what we measured. As an alternative of clicks, we targeted on return visits, session depth, and long-term engagement. The outcome? The brand new algorithm elevated clicks by 12% however decreased return visits by 8%. The usual framework would have stated “ship it.” The primary rules strategy stated “not but.”
We determined that the brand new algorithm was too “click-baity”. Frameworks gave us the methodology. First rules gave us the precise query.
What Are Our Metrics Actually Measuring?
Metrics frameworks—North Star, OKRs, HEART—are highly effective as a result of they offer construction. However they will additionally create the phantasm that we’ve measured what issues.
First rules pondering asks: What’s the elemental habits or worth we care about? Does this metric really seize it?
Take into account engagement. Many frameworks recommend DAU, session size, or actions per session. Affordable proxies—however are they proper?
- For a meditation app, longer classes might look “higher,” however the basic aim is a sustainable follow. That may imply shorter classes over time.
- For an analytics device, extra queries per person might sign deeper utilization, or it might imply customers are struggling to search out solutions. The actual worth is quicker, extra focused insights.
I as soon as began a brand new job and inherited a dashboard that proudly reported weekly energetic customers as the first success metric. However once I dug in, I spotted most “energetic” customers had been simply logging in, wanting round, and leaving with out finishing a single job. From a primary rules lens, I made a decision that the true worth was duties accomplished. And as soon as I shifted the metric, we found (as anticipated) that utilization was far decrease underneath the brand new definition, however the brand new framing offered us readability on what to give attention to in an effort to drive significant adoption.
The framework provides you a menu of metrics. First rules pondering tells you whether or not any of them really mirror your product’s worth. Typically the usual metric is ideal, however typically it’s dangerously deceptive.
When First Rules Saved a Launch
One of many clearest examples I’ve skilled got here early in my profession when my group was tasked with constructing a “person high quality rating” to assist gross sales prioritize leads.
The framework strategy was apparent: supervised studying, predict conversion chance, rank by rating. We had the info, the options, the methodology.
Two weeks in, as we had been struggling to eke out incrementally greater efficiency scores, somebody requested: What determination will gross sales really make with this rating?
We requested gross sales. The reply wasn’t “give me actual possibilities.” It was: Ought to I take the time and personalization to name this lead, or simply ship a fast standardized e-mail?
That shifted every part. We didn’t want a fancy mannequin spanning the complete chance vary. We would have liked a easy, interpretable classifier optimized round one threshold.
By reframing, we moved from an ensemble mannequin to logistic regression, dropped half our options, shipped three weeks sooner—and delivered one thing gross sales really used.
By going again to first rules, we clarified the true downside after which went again to the usual frameworks to construct an answer.
The Compass and the Map
Right here’s the lesson: strategic knowledge scientists don’t select between frameworks and first rules. They mix them.
- Frameworks are the map—they allow you to transfer quick and leverage accrued data.
- First rules are the compass—they preserve you aligned when the map isn’t clearly exhibiting your path.
The Wright brothers didn’t reject the analysis of their time. They constructed on it, however in addition they knew when to return to fundamentals.
That’s the mindset shift that separates strategic knowledge scientists from tactical ones. It’s not about figuring out extra strategies or working more durable. It’s about figuring out when to observe the map and when to examine your compass.
With the steerage of AI, frameworks will preserve getting simpler to use. However the compass—that’s yours to construct. And it’s what is going to preserve you related, strategic, and indispensable within the years forward.
This is without doubt one of the core themes I discover in my new e-book, The Strategic Knowledge Scientist: Stage Up and Thrive within the Age of AI (Amazon affiliate hyperlink). It’s about pairing foundational pondering with confirmed frameworks to create impression, affect roadmaps, and place your self as a strategic accomplice—not only a technical executor.
In case you’re questioning how your job will change as AI turns into extra succesful, or just trying to have extra impression as a knowledge scientist and make progress in direction of a promotion, please take a look at the e-book on Amazon!
