What changed
The "MERL Tech" community (Linda Raftree, Wayan Vota and others, first conference 2014) pushed mobile data collection, dashboards, big data and machine learning into M&E while simultaneously building the "responsible data" agenda: Oxfam's responsible data policy, the Responsible Data Forum, GDPR from 2018.
Michael Bamberger's work with UN Global Pulse framed how satellite imagery, call-detail records and social-media data could enter evaluation; 3ie later ran geospatial impact evaluations while warning that big data would not replace surveys. Acumen's "lean data" — short phone surveys with customers or beneficiaries — spun out as 60 Decibels in 2019 and became the impact-investing sector's default.
COVID-19 forced "remote MEL" in 2020 (phone, IVR and SMS surveys, remote third-party monitoring) and permanently normalised hybrid data collection. Then came generative AI. The MERL Tech natural-language-processing community of practice launched in 2022, months before ChatGPT; by 2024 the American Evaluation Association reported that the share of conference proposals mentioning AI had jumped from around a fifth to well over half in a single year.
Guidance followed in 2024–25 from the UNEG Data Analytics and AI working group, CGIAR's evaluation function and the UK Evaluation Society's AI working group (good-practice guidelines, version 2.0 in November 2025). By 2026 the practical uses had settled into AI-assisted qualitative coding, evidence synthesis and report drafting, with the field arguing about where human judgement must remain.
Frameworks and tools
- ODK, KoboToolbox, SurveyCTO and CommCare
- lean data
- rapid-feedback MERL
- remote-sensing indicators
- responsible data policies
- the UNEG ethical principles for AI in evaluation
- UKES AI good-practice guidelines
- "human in the loop" protocols
- small and non-English language models
Vocabulary to recognise
- MERL Tech
- responsible data
- data minimisation
- lean data
- remote MEL
- real-time data
- NLP
- LLM
- generative AI
- human-in-the-loop
- AI-ready data
- hallucination
- synthetic respondents
Live controversies
Hallucinated findings and fabricated citations; bias baked into Anglophone, Global North training data; the confidentiality of respondent data fed into commercial models; deskilling of junior evaluators; the environmental footprint of AI; "AI adoption as the goal" rather than a means; and whether AI-generated syntheses of donor archives (such as the analysis of 13,000 USAID evaluations) are credible evidence or plausible-sounding noise.
Further reading
- Linda Raftree, MERL Tech State of the Field: The Evolution of MERL Tech, MERL Tech, 2020.
- Rick Davies, Integrating Big Data into the Monitoring and Evaluation of Development Programmes (on Bamberger's UN Global Pulse report), Monitoring and Evaluation News, 2017.
- 3ie, Despite the hype, do not expect big data to replace traditional surveys anytime soon and Three ways you can start using remote sensing for measuring impact.
- Sasha Dichter, Tom Adams and Alnoor Ebrahim, Leveraging Lean Data, Stanford Social Innovation Review, 2016; 60 Decibels, The Power of Lean Data.
- Wayan Vota, New USAID Guide: Remote Monitoring Processes for COVID-19 Response, ICTworks, 2020.
- Linda Raftree, How to Develop and Implement Responsible Data Policies, ICTworks, 2016.
- MERL Tech, Wrapping up: A year in the NLP-CoP, 2023.
- John Baek, A Snapshot of the State of Artificial Intelligence in Evaluation Practice, AEA365, 2024.
- UNEG, Ethical Principles for Harnessing AI in United Nations Evaluations, 2025.
- UK Evaluation Society, AI in Evaluation: Good Practice Guidelines for Practitioners, v2.0, 2025.
- Timo Wencker, Judith Borst-Graetz and Andreas Niekler, Text as data for evaluation: NLP and large language models to generate novel insights from unstructured text data, Evaluation, 2025; Sanae Bouyousfi and Moussa Ouedraogo, Artificial intelligence and big data-driven evaluation research and practices: A systematic literature review, Evaluation, 2025.
- Linda Raftree, Where is Generative AI Supporting MERL Improvement in 2026?, ICTworks, 2026.
- Antony Maikuri, Veronica Olazabal and Vidhya Shanker, Doing no harm with technology in evaluation, AEA365.