Much has been written about the 10x – heck, 100x – engineer. The mystical creature that ships features in hours instead of months and perseveres where others back down. Corny as it may sound, I think it’s a pretty good representation of how the world works.
But does the 10x analyst exist? I believe so.
How about the 100x analyst? I’m not sure.
Take two people – Entry-level Eric and Mid-level Mary. Eric just finished a data science boot camp. He learned how to write SQL and read bestsellers on how to visualise data.
If you’ve tried to learn to ski by watching YouTube videos, you realize it doesn’t quite work.
Eric is no exception. His analysis are too long and misses key points. He takes on too many requests from data-hungry stakeholders. And his data pipelines contain janky logic, making them run slowly.
In contrast, Mid-level Mary has seen all this before. She works independently, is highly valued by her stakeholders, and delivers fast. Her experience makes her at least 2-3 times more effective. She knows how to do the things right.
Now, meet Senior Sam. A seasoned staff analyst, having worked at a FAANG and several unicorns. Sam joins the company, and it’s like the lights have been turned on.
He’s seen how to structure a dbt project ten times the size and immediately introduces best practices.
He brings in a metric tree that breaks down the cost drivers and shows that some locations disproportionally contribute to the unit economic losses.
And he does all this without being told what to do.
Sam is at least 2-3 times more effective than Mary. He knows how to do the things right at scale.
Compared to Eric, Sam is 10 times more effective. But how about the 100x analyst?
I’m sure there’s someone out there. But most have lost the analyst title along the way.
Here’s my own experience of working with one.
He had a mental model of the business. A product manager pitched a modified version of a fast-growing competing product. He said it would fit into the product roadmap and could easily be integrated. They just needed an analyst to do the market sizing and set up tracking plans first. The 100x analyst asked, “Just for my understanding, if we succeed in rolling out to all users, that’s a $10m business. In comparison, our existing product grew $100m in the MENA region alone last year? ” The project was dropped.
He acted as a filter. A sales director wanted detailed data on what % of the customer base sees a mobile-optimized website. An answer that would take weeks to get to and require raw logs processing. Instead of jumping heads in, the 100x analyst asked, “Let’s say the number is X. What would you do then?”. Pondering the answer, the sales director returned, concluding that it didn’t matter.
He thought in systems. Using data, he broke down the customer base consisting of 100,000s of customers into find, grow, keep, and used these segments to direct the company-wide business strategy, from how customer acquisition was run to onboarding and churn prevention. This placed him in the middle of most important decisions.
Instead of focusing on doing the things right, he focused on doing the right things.
You may have guessed it. Although the person above was once an analyst, he lost the title long ago and is now a VP at Google, running an entire business unit.
Look around, and you’ll see that’s how it goes. 100x analysts are all around. They are just no longer called analysts.
Terrific post. Can the same be said of the data engineers? That is, they eventually become (say) staff software engineers?