The instinct is to hire someone. And it almost always comes too soon.
When a director looks at a report and does not trust the number in front of them, the first thought is nearly always the same: I need someone on the data side, in-house. It makes perfect sense. Making decisions in the dark is uncomfortable, and hiring a person feels like the most direct way to make that discomfort go away.
The problem is that the hire usually lands before the company knows exactly what it will ask of that person. And an analyst without clear questions in front of them does not bring order to the mess. They just draw it more neatly.
What fails is not a shortage of hands
Here is the figure I find most revealing. Gartner estimates that 80% of data governance initiatives will fail by 2027, and not for lack of technology or people, but because they are not tied to a real business decision. In the same vein, only 48% of digital initiatives actually meet the business outcome they set out to hit.
Read those two numbers again. Neither one says analysts were missing. They describe projects that produced reports nobody used, because nobody had connected them to a concrete decision. The bottleneck is rarely who builds the chart. It is who decides which question deserves a chart.
The craft is not making charts. It is translating a decision into a number.
Anyone with a three-month Power BI course can build you a dashboard with twenty metrics. That is cheap and abundant. What is scarce is the person who sits down with the sales director, understands what they actually decide every week, and works backward to design the number they need to decide well. That is the craft.
And a junior profile at 25,000 euros a year rarely has it. Not because they lack talent, but because that judgment is built after years spent in front of very different businesses. If you hire that junior to "sort out the data", you are asking for the easy part (the reports) without handing over the hard part already solved (knowing which report matters). The result is predictable: beautiful dashboards that no one looks at in the leadership meeting.
An analyst without the business question in front of them does not give you judgment. They give you charts faster than you have time to ignore them.
The expensive part is not the salary. It is what comes after.
Let us do round numbers. A data analyst in Spain earns on average around 35,000 euros a year, and a senior profile with real judgment moves up to 50,000 or 60,000. But the salary is the part you can actually forecast. What almost no one factors in is the rest.
First, finding them. Eight out of ten Spanish small and mid-sized companies report difficulty filling qualified vacancies, and among tech companies only 34% find valid candidates in today's market. You will spend months.
Second, keeping them. Data profiles turn over quickly: in tech, tenures tend to be short, just a few years. Translate that to your day to day. Just as the person starts to truly understand your business, a better offer arrives and they leave. And you are left with dashboards only that person knew how to maintain.
A premature hire does not bring the mess down. It turns it into a fixed salary plus a turnover risk hanging over you.
When an in-house analyst does make sense
I am not against hiring. I work with companies of different sizes, and some of them do very well to have their own data team. The difference is the order.
Bringing it in-house makes sense when you already know which business questions matter to you, when those questions repeat every week with enough volume, and when there is someone with senior judgment who can direct that analyst so they are not working blind. In other words: when the hard work, defining what to measure and why, is already done, and what you need are steady hands for the recurring work.
Hiring before that is hiring someone to discover your problem for you. It runs expensive and it frustrates both sides. If you are not sure where your company stands, start by placing what level of data maturity you are at and by defining the metrics that actually help you decide. That is worth more than any new payroll entry.
How I see it
When I step into a company that "needs someone on the data side", the first thing I do is not build reports. It is sit down with leadership and dig out the three or four decisions that truly move the business, and work out which numbers would let them make those decisions with confidence. That judgment work is the part almost no one wants to pay for, and the part that changes everything. From there, the company decides with data whether an in-house hire, an external specialist on a project, or a mix of the two pays off. That is how I approach every project.
Judgment first. Then, if needed, the hands. Never the other way around.
So the next time you feel you need to hire an analyst, try changing the question. It is not "who do I hire?". It is "which three decisions do I want to make better, and what number am I missing for each one?". If you can answer that, the hire can be a great idea. If you cannot, no hire will save you from having to answer it first.