I work on digital technology in the humanitarian sector, and I kept getting a sense of déjà vu. One year everyone was focused on biometrics, the next on interoperability, then blockchain. Themes seemed to come and go. I always felt we were somewhere on a hype cycle, but I couldn't prove it. I wanted data. It would help me decide what to ignore, and when to tell colleagues that something deserves their attention.
I tried a few ways of measuring this, and in 2022 I came across ReliefWeb's jobs API. I built a basic Power BI dashboard that tracked themes across roles. I was never happy with it. The tagging was crude, it was slow, and like every Power BI dashboard, it looked like Power BI.
In September 2026 I picked it up again. This time I had AI tools and a much better grasp of Python, APIs and data science.
So is this about hiring?
Not at all, though it might be for you. I use job adverts because of the kind of data they are. They're sometimes called organic data: records produced by the work itself, not collected afterwards. Data collected afterwards brings biases, gaps and delays. Job adverts aren't written to be studied. They're a by-product of organisations deciding what to spend money on, and in some respects they also show what organisations hope to do. That makes them an honest record. They have limits too, which the methodology sets out.
What did I find?
Biometrics had a boom and a bust, though the numbers are small. AI is a recent, rapid rise. And it was great to see localisation and local leadership climbing as a share of adverts.
Sign up below for the monthly trendline. I'll share the latest changes there and dig deeper into the earlier data.
If you have ideas for new features, questions, or just want to talk, email me at rorycrew@curinta.com.
Rory
Humanitarian Trendlines is an independent project by Curinta. It is not produced or endorsed by ReliefWeb or OCHA.