When companies are conducting internal data projects, either to help decision making or automate processes, they require a number of technical skills, modulated by the amount of custom tools vs on-the shelf solutions that are chosen. On top of the pure software development, we see a need for more devops, databases, distributed file system knowledge, UX design, etc...<p>My question is: where in this whole ecosystem or data lifecycle software development (= actually writing code) has the most value or is the most present? That would be the place to focus specialisation for a freelance software developer like me.
I think your question depends on whose perspective. From the perspective of a company whose core business model is not software, there is little value in writing code. A CEO sees danger in every major software implementation: they're expensive, takes forever, often fail, expensive to maintain and may get him fired. A CEO wants to customize a little as possible, so he can upgrade and maintain his very expensive capital expense.<p>On the other hand, from the perspective of consulting company or large software vendor, the consultant wants to lock-in his customer to a long term contract as long as possible. This means he wants to implement a be-spoke customized solution with infinite levels of complexity that will make it easier for his customer to move mountains than switch to a different vendor.<p>If you are trying to determine which data technology you should specialize in, personally, I take great interest in the stock market valuation of those companies. It's not foolproof but I would argue is a better indicator than most. For example, recently the Dow Jones industrial average replaced Exxon with Salesforce. Salesforce is more than a CRM; it really is cloud platform that many business base their IT strategy on. Similarly, Snowflake's recent IPO of $3 billion was reported to be the largest ever for a software company.