The business commentariat have long warned of a “skills gap” in enterprise data analytics and claim the solution is to upskill sales, marketers, HR, and the rest. We believe they’ve identified a real problem but have misdiagnosed it. There is a technology gap, not a skills gap, and expensive training programs are not the most efficient way to make business users proficient at data analysis.
In a matter of minutes, someone without any technical knowledge could open an account with a conversational analytics platform (like DataChat), connect to their database, ask a business question in plain English, and get the answer in seconds. Individuals and enterprises invest heavily in data skills when, for a fraction of the cost and time, they could give business users tools to get data insights—the thing they actually need.
Why isn’t that more obvious? Because the messaging around skills development conflates abilities with goals and job roles with tasks.
The Skills Gap Industry
The “skills gap” crowd insist that individuals, companies, and governments need to make massive investments in data science training that takes months, if not years, before it delivers a return on investment. A growing industry has formed to meet this supposed need.
Global consultancies are launching data science upskilling programs as a service to clients. Analytics platforms have partnered with online learning platforms to offer courses on using their software. Even the Defense Advanced Research Projects Agency (DARPA) is funding education startups that can help adults become proficient in data science.
This isn’t just a U.S. phenomenon. The education platform Great Learning surveyed 1,000 professionals in India, a vibrant talent market, and found that 85% plan to “upskill” in 2025. Of those, almost 30% plan to upskill in “Data Science and AIML,” making it far and away the most popular option. With the average data science bootcamp costing $9K to $16K, the training and education industries are happy to oblige.
Skills v. Underlying Goals
There are cases where upskilling makes sense. It can be a path to a new career or a way to remain competitive in the job market. Sometimes, though, enterprises invest in upskilling at the expense of the underlying goal.
Think about it this way: when a company wants to build a new headquarters, they don’t “upskill” their marketers and salespeople in construction. With enough training, a company could turn a marketer into a carpenter or a salesperson into a plumber, but that sounds absurd, doesn’t it? After all, the objective is to build a headquarters, not to train a team in building headquarters.
Training a marketer or salesperson to be a data analyst is similar because the objective isn’t to have a bigger data analytics team. Rather, the objective is to simplify and automate data analysis so that marketers, salespeople, HR, etc. can get the insights they need to make better decisions and improve performance, however it’s measured. There are skills more relevant to marketing, sales, and HR that would be a better upskilling investment.
Those Who Analyze v. Analysts
At DataChat, it strikes us that enterprises spend money on developing skillsets that conversational analytics can already perform on behalf of non-technical business users. After all, business users already do some of that data work. They look at dashboards, they generate reports from their CRM, and they scan monthly performance metrics in spreadsheets.
Data literacy is widespread, but Python and SQL coding are not. Someone who wants to build game-changing data science software probably needs those coding skills and an advanced degree in the field.
However, the average enterprise “business user” who is trying to make data-driven decisions does not need either. They don’t want a certificate in data analysis, but rather insights from data that will help them do their job better. They want to ask questions of data that only an experienced marketer, salesperson, or HR leader would think to ask—questions that pre-canned dashboard or report won’t address.
When a Tool Solves the Problem
Remember, there was a time when business users didn’t type anything, relying instead on specialists to operate a typewriter efficiently, without wasting time, ink, or paper. The solution wasn’t to train everyone as typists—it was to create computers with keyboards that made typing more accessible, more forgiving of errors, and less expensive.
Today, if someone is data literate, they don’t necessarily need a data analysis degree or bootcamp. They just need the right tool to extract value from plentiful, untapped data.
The underlying goal isn’t to make business users do data science all day. It is to help them get the insights they need, quickly and cost-efficiently, so they do their work better. That is what conversational analytics promise to do.