Hamlet turns meeting comments into data-center siting-risk rankings

Hamlet has launched local-government issue rankings that turn comments at public meetings into a searchable signal of where data-center projects may face political resistance. The rankings cover city councils, county boards, planning commissions and school boards, drawing on video and transcripts published by those bodies.

The data-center view counted 2,291 people who took a side across 358 places: 1,947 against and 344 in favor. Hamlet’s page said most comments were against in 300 of 355 places, but did not explain why that headline denominator differed from the 358 places in its detail line. When observed, the broader service said it had read more than 83,000 meetings and held remarks for more than 1,600 places and more than 2,300 officials.

Hamlet applies the same search terms in every place. An automated reader then labels each matching remark as for, against, mixed or neutral, and a second model checks the classification. Mixed and neutral remarks are excluded. On data centers, opposition means favoring a ban, moratorium or tight limits; support means allowing construction.

The system counts people, not comments. Each person gets one position per place and topic, determined by the side that person took most often, regardless of how many times they spoke. An even split produces no position. Every speaker currently has equal weight because Hamlet says meeting roles remain automated guesses. After confirming roles against its roster, Hamlet plans to give elected and appointed officials a weight of three, while public commenters and staff would keep a weight of one.

A place’s score divides its weighted voices against by all weighted voices for and against. Before ranking places, Hamlet adds 10 hypothetical voices at the national rate for the topic. The smoothing step is designed to keep tiny groups from leading a list by chance. To qualify, a place must have at least 10 people speak at two or more meetings in the prior 24 months. Places are compared only within large, midsize and small population groups.

The observed data-center results covered meetings since October 19, 2024, and the newest included meeting was dated September 17, 2026. Data centers were also discussed in 182 places where fewer than three people spoke. Those places appeared as lighter dots and could not qualify for the rankings without meeting the 10-person and two-meeting thresholds. Hamlet said data centers appeared in 16% of tracked meetings in the third quarter, up 28% from the second quarter.

In exclusive reporting, Semafor identified two unnamed paying customers based on information from Hamlet CEO Sunil Rajaraman: a large fund financing data centers and a data-center developer, both seeking signs of moratorium risk. Rajaraman also identified real estate developers, political action committees and retailers as customers. The customer identities and contracts were not disclosed. Hamlet separately markets the service for site screening, including tracking counties considering moratoriums and finding officials who have spoken. Customers can use the product directly or access records through an API.

The rankings are not a representative opinion poll. They measure people whose comments took a side in the meetings Hamlet tracks and officials who spoke, not a probability sample of residents or the country. Hamlet also cautions that its automated reading can miss or misinterpret remarks. Among the large places shown, Corpus Christi, Texas, ranked first for opposition, with 23 people counted against and none for. Atlanta followed with 19 against and none for.