Guide · 6 min read
Screening is done in seconds, by people and increasingly by software. What gets a data or climate CV through, and what quietly disqualifies it.
The first third of the page decides everything. Open with a three-line summary that states your field, your years of experience, and two concrete results: "Built the seasonal forecast pipeline used by X for 2024 planning", "Delivered dashboards used across 4 country programmes". Claims that a reader could check are worth more than adjectives.
Every role bullet should carry a number where one exists: stations covered, records processed, users served, budget managed, improvement measured. Follow the number with the tools, using the exact names employers search for: Python, R, SQL, QGIS, ArcGIS, Google Earth Engine, Power BI, Stata, WRF, CHIRPS. Recruiters filter on these words.
If the posting says "monitoring, evaluation and learning", do not write "M&E" only. If it says "geospatial", say geospatial, not just "maps". Software screening looks for the posting's own words; human screeners do the same unconsciously.
Two pages for most professionals, three for senior researchers with long publication lists. Put publications and trainings in a compact final section; put languages with a level (working, fluent, native) because development employers care.
Maintain one long master CV with every fact, and cut a tailored two-page version per application. Tools that check each claim against your master evidence catch the mistake where a rushed tailored CV says something the master never did.