Analysis of the Stack Overflow 2024 Developer Survey (~18,845 respondents) to identify which programming languages and databases developers use today, which they want to use next, and where the gap between the two signals a shift.
Built as the capstone for the IBM Data Analyst Professional Certificate.
Technology choices have a lag. What developers use reflects decisions made years ago; what they want to use hints at where the ecosystem is heading. This project asks:
- Which languages and databases dominate current usage?
- Which ones do developers want to work with next year?
- Where is the largest gap between current use and future intent, and what does that gap suggest?
| Source | Stack Overflow Annual Developer Survey, 2024 |
| Respondents | ~18,845 (after filtering) |
| Key fields | LanguageHaveWorkedWith, LanguageWantToWorkWith, DatabaseHaveWorkedWith, DatabaseWantToWorkWith, plus platforms, web frameworks, age, country, and education |
| Format | Wide CSV; the eight technology columns are semicolon-delimited multi-select fields |
The main cleaning challenge: each technology cell holds multiple values separated by ;, so respondent counts had to be derived by splitting each field and counting individual technologies rather than reading the column directly.
Python (Pandas, Matplotlib), IBM Cognos Analytics, python-pptx
- Clean. Handled missing values, removed duplicate respondent IDs, and normalized inconsistent category labels.
- Reshape. Split the semicolon-delimited technology fields into individual entries and counted respondents per technology.
- Analyze. Built top-10 rankings for current versus desired use across languages and databases, then compared rank and share between the two.
- Visualize. Charted the comparisons and built a 3-tab interactive Cognos dashboard.
- Automate. Generated the summary presentation with
python-pptxso the deck can be rebuilt from a script instead of manual copy and paste.
1. SQL is the most durable skill in the survey. SQL ranks #2 in current use (12,602 respondents) and holds #2 in future intent (10,944). Almost every other language shifts position. Nothing displaces SQL, which makes it the safest single investment for anyone entering data work.
2. Rust and Go are the fastest growing languages. Rust grows 145% (2,284 to 5,597) and Go 71% (3,305 to 5,661), the largest gains among widely used languages. Both sit outside the current top 10, at ranks 14 and 11, which is exactly why a top-10-only comparison misses them and makes them look like they appeared from nowhere.
3. Three databases grow against a falling baseline. Of the ten most-used databases, only Redis (+9.8%), PostgreSQL (+5.9%), and Elasticsearch (+5.0%) rise from current use to future intent. PostgreSQL adds the most people (+679) from an already dominant base; Redis grows fastest by rate and climbs four ranks. MySQL falls 27.5% and Oracle 41.6%.
4. The stack that wins today also wins tomorrow. JavaScript leads languages, PostgreSQL leads databases, AWS dominates cloud platforms, and Node.js and React lead web frameworks. All four hold their top spot in the "want to work with" data too. The churn happens beneath the leaders, not at the top.
Two traps produce confidently wrong answers here, and the analysis notebook demonstrates both.
First, respondents name fewer technologies when asked what they want than what they have used. Language picks fall from 6.19 to 5.64 per respondent, databases from 3.69 to 3.50. Every count drops on the "want" side by default, so an absolute decline does not by itself mean a technology is dying. Rank movement and relative change are the trustworthy signals.
Second, comparing the two top-10 lists directly makes technologies just outside one list look like they have zero presence. Rust appears to come from nowhere until you check the full distribution and find 2,284 current users at rank 14. Always look past the truncation.
A 3-tab interactive dashboard built in IBM Cognos Analytics.
Tab 1: Current Technology Usage covers the top 10 languages, databases, platforms, and web frameworks developers work with today.
Tab 2: Future Technology Trends covers the same four categories, filtered to what developers want to work with next year.
Tab 3: Demographics covers respondent age, country, and education distribution.
Respondents skew 25 to 34 years old (41%), concentrated in the United States, Germany, India, and the United Kingdom, with a Bachelor's degree as the most common education level, followed by a Master's.
.
├── README.md
├── notebooks/
│ └── 01_technology_rankings.ipynb # splitting, counting, and ranking the tech fields
├── images/
│ ├── dashboard-tab1-current-usage.png
│ ├── dashboard-tab2-future-trends.png
│ └── dashboard-tab3-demographics.png
└── reports/
└── capstone_report.pdf # full 18-slide analysis deck
notebooks/01_technology_rankings.ipynb contains the core technique: splitting the semicolon-delimited multi-select fields, counting individual technologies, and comparing current use against future intent. It also works through the two ways this data misleads you if you take the top-10 lists at face value. Outputs are saved, so it reads without running anything.
The complete analysis, including methodology, both trend chart pairs, all three dashboard tabs, and the discussion, is in reports/capstone_report.pdf.
Raw survey data is not committed here. It can be downloaded from the Stack Overflow Developer Survey archive.
- Segment technology preferences by years of experience to test whether Rust and Go interest is concentrated among newer or more senior developers.
- Compare against the 2023 survey to check whether PostgreSQL's rise is a trend or a single-year artifact.
- Weight results by country to correct for the survey's heavy US and European response bias.
Daniel Aderibigbe · LinkedIn