the death of the data analyst
it's no longer enough to build a dashboard
Just a few months ago, I actually looked down on data analysts.
So, in the Lord’s year of 2025, it’s rather ironic and gun-in-cheek of me to be positioning myself as one.
And especially more so, rather poetically and suicidal of me, to be publishing an article called ‘the death of the data analyst’, just a few months after loosely becoming one.
In March 2025, I transitioned into a data analyst in the sales, distribution and partnerships team of a multinational insurance company, based in Australia. I have some other roles and responsibilities, but one of my main ones is to facilitate reliable reporting/dashboards and interrogate data and derive business insights; all in the name of revenue generation.
This was a move from pricing or actuarial, which is also a form of data analyst but significantly more specialised, dedicated to calculating insurance risk.
A key difference between the two is that actuarial analyst requires a degree to do (which I have) while sales/data analyst, as harsh as it is, can be done by a reasonably smart monkey with a laptop.
***
In a recent catch-up with my manager who is decided not an analyst, I expressed some recent concerns about “what next” lied in my career.
It was a bit early to ask that question considering I haven’t even passed my probation yet, but it’s one of the lingering worries that had settled at the back of my head; the concern that by moving from a degree-required field to a non-degree-required field, I had actively chosen to stagnate my career.
I’ve got the prefix ‘Senior’ attached at the front of my title. I don’t know what I’ve done so far to warrant getting a corporate walking stick, but it comes with more respect and higher pay, so who am I to complain?
There are some cool-sounding prefixes that cane come next, I reckon. Principal Analyst or Lead Analyst comes to mind, where you adopt more technical expertise and responsibility or start taking charge of projects.
There is of course, donning the ‘Manager’ suffix. It’s the natural progression of worlds that I’ve never really understood because being good at [Thing] does not make you good at managing people — along with the irony that once you become [Thing] Manager, you start doing significantly less of [Thing] and simply manage people who do [Thing].
***
This could be a limiting belief but my worry is that the data analyst field is 1) nothing more than a steppingstone into a more managerial field and 2) the entry-to-intermediate level data analyst is becoming increasingly unnecessary.
In the next ten years, I think only two forms of data analyst will survive:
The analysts who use data to forecast the future and tell stories around it. These people use their technical skill to transform scattered bits of information and spin them together into actionable business strategy and predictive algorithms, and their soft skills to tell a compelling narrative and demonstrate the strength of their forecasts. What makes them unique is not the order in which they hit keys on their keyboard, but their ability to look up from the computer and speak: To clients, to customers, to business, to managers, to executives, to the team member across the road. They bridge the gap between technical knowledge and business requirements and run the basis that if they speak, people will listen. Their edge lies in being both technically-sound and client-facing.
The analysts who barely do analysis or presentation, but instead comb through systems and facilitate processes, maintain the architecture and integrity of all things data. Not to be dramatic, but these guys never see the light of day and you ideally shouldn’t talk with them unless something had fallen over. They are plumbers, getting all the pipes, tanks and fixtures flowing to the right spot, giving the previous analyst — along with the accountant and salesperson and CEO — a reliable, correct stream of information to do their job. By facilitating end-to-end overnight processes like schedulers and ETLs, they are the warehouse men with 24/7 forklifts that provide that allow data to exist in the first place, that become increasingly important in an economy where we expect everything to work seamlessly.
The first kind is sometimes called Data Scientist, though they’re often called Data Analyst as an excuse to pay them less. The second kind is decidedly a Data Engineer, also sometimes called Data Analyst, in order to — you guessed it — pay them less.
I like to say that while my background is numbers and math (see: actuarial degree), my heart belongs to words and stories (see: this blog!).
As such, I’m much more of the first type. It’s where my university and work experience lies and is more in-line with my personality.
Though, perhaps because it’s something that comes fairly naturally to me, I’m afraid I don’t think as highly of it.
I run on the belief that anyone who can operate a computer can run a spreadsheet and run some off-the-napkin calculations, that anyone with a university degree can write a decent report and learn to present to management. With the advent of AI, anyone can ask ChatGPT very nicely to spit out an SQL and Python script that more or less caters to their needs, so long as you develop the fundamentals of code, which with enough effort an say an $80k salary on the line — anyone can do.
Heck, if it weren’t for ChatGPT/Copilot, I’d have to learn SAS and SQL and Python and whatever the stupid language of the year is which would seriously hamper my ability to provide value.
And I acknowledge that this isn’t everyone’s experience and could just be how my brain is wired — but I just feel the playing field is levelling out, and it’s rewarding people not who have technical ability, but with the agency and initiative to try.
For better or for worse, it’s no longer enough to have technical knowledge. A few decades ago, you could get away with putting together a nice Tableau/Power BI dashboard or an interactive FP&A model, chuck a few talking points together and call yourself an analyst. It’s enough to get your foot in the door mind you, especially in slow-moving industries like my own (see: insurance), but it’s enough to make you flourish, stand out, and progress.
And just a few years later, I fear it’s not even enough to get you started.
The best analysts are no longer the ones with technical prowess, and dare I say, it has never been that way.
The best analysts have self-agency, ask question after question, and wield initiative + enthusiasm as their greatest weapon. They know how to tell a story and bring the audience for a ride while knowing when narratives are NOT the way to go. They can distill complicated topics into simple, digestible insights — and transform scattered bullet points and anecdotes into hypothesis, experiment, and finding.
Well, it’s either that or the analyst who builds an AI model powerful enough to innovate himself/herself out, but let’s hope we never accidentally get there.
posts that i’ve loved this week:
- 💥
Fashion Propaganda Im Not Falling For - by
👠Ranking the glasses in my cupboard - by
🥃“I don’t know anyone who isn’t in debt unless they’re a trust-fund kid or married rich.” - by
💳