The notice ran in the legal section of the county paper. Most people found out from a neighbor's post, six weeks after the site plan was filed, three weeks before the vote.
By then two hundred residents had questions about water, and the hearing had room for eleven of them at three minutes each. The board thanked everyone for their comments. The rezoning passed.
This is the most common shape of a data center fight, and it is not a failure of turnout. It is a failure of timing. The room filled up after the decisions that mattered had already been made somewhere quieter.
You are not fighting a building. You are fighting a schedule.
A campus is not approved in one vote. It is approved through a sequence: land assembly, rezoning or a special use permit, a site plan, a water and sewer agreement, an interconnection queue position with the utility, a power purchase or substation build, a tax abatement negotiated with the county or the state, sometimes a noise or setback variance, sometimes an environmental review.
Each of those is a separate decision, with a separate decider, on a separate date, and each one is open for a window and then closed. Some of them are made by people nobody elected.
So before you organize anything, build a decision map. What exactly is proposed. Where it sits in the approval process right now. Which decisions remain genuinely open. Who makes each one, when, and under what legal standard. What has already been committed by the city, the county, the utility, the developer or the state, and whether those commitments are binding.
That map is the single highest-value thing a group can produce in its first month, and almost nobody does it first. It converts an unanswerable question, how do we stop this, into a set of answerable ones. It also tells you immediately whether you are early, which changes everything about what to do next.
This is not a fringe position anymore
It is worth knowing how much company you have, because most groups start out assuming they are three people being difficult.
Data Center Watch has identified at least 142 activist groups across 24 states, and tracks projects blocked or delayed through permit decisions, zoning, regulatory change, litigation, elections and developer withdrawal. FracTracker's US data centers tracker documents more than 150 instances of community resistance and roughly 40 outright victories. A May 2026 analysis by the Coalition for Responsible Data Center Development counted 345 local opposition groups with more than 428,000 members across 37 states. Nationally representative Annenberg polling now finds 61% of Americans oppose new data centers in their own area, up 12 points in a few months, while general attitudes toward AI barely moved.
Two things follow from those numbers.
First, local action changes outcomes. This is not aspirational. Roughly forty projects have been stopped or materially reshaped, and the mechanisms were procedural far more often than they were confrontational.
Second, the fight is genuinely local. Data Center Watch's own conclusion is that challenges succeed locally because that is where the permitting happens. Which means "raise awareness about data centers" and "change this provision of the county zoning ordinance before the September filing deadline" are not two versions of the same activity. They are different jobs, and only one of them has a date attached.
Do not lead with your reason
The instinct is to open with the reason you care. If that is climate, or AI, you will immediately shrink your coalition to people who already agree with you.
The tracking data is unusually clear on this point: what mobilizes people around data centers is electricity prices, water, grid reliability, noise, tax abatements, property values, green space, truck traffic and construction. Different concerns recruit different constituencies, and the issue crosses conventional political lines in a way very few environmental questions do. A county board that would dismiss a carbon argument out of hand will hold a hearing on ratepayer costs.
So spend the first weeks listening rather than persuading. Map who is affected, what each of them actually cares about, and what they would want to see happen. You are not looking for people who share your motive. You are looking for people who share your ask, and the honest version of this work is finding their reason rather than selling them yours.
The Henrico County case is instructive. A predominantly Black middle-income community organized around environmental justice while a predominantly White higher-income community organized around property values. They arrived at the same policy demand from two entirely different starting points, and the county's approval framework changed.
Build the group before you build the petition
Do not start with "sign this." A petition is a moment, and policy responds to constituencies rather than to moments.
Build a core group instead, and keep it small enough that people can feel their contribution counts. Give people roles that match what they are already good at, because the retired engineer reading the noise study, the parent running the phone tree and the lawyer reading the zoning text are three different jobs and one person cannot hold all of them. Develop local leaders rather than becoming the only spokesperson. Make contributions visible. Notice when somebody stops showing up, and call them.
Participation itself creates the sense of agency that keeps people going, but without a group to come back to, that energy has nowhere to land and the effort does not survive the month. This is the piece that gets skipped, and it is the piece that determines whether anything is still standing at the next hearing, eleven months from now, when the developer refiles a modified plan.
Choose the smallest winnable change
"Stop the five billion dollar development" is one demand with two possible outcomes, and one of them ends your group.
Underneath it sits a ladder of decisions that each improve the outcome and each can be won separately:
Disclosure of projected water consumption. An independent noise assessment before approval rather than after. A public hearing that was not otherwise required. Setback and buffer requirements. Ratepayer protections so residential customers do not fund the grid upgrade. Backup generator and air quality standards. A zoning text change that governs every future application, not just this one. A community benefits agreement with enforceable terms. A moratorium while standards get written.
Each win increases the group's sense of efficacy, its legitimacy with officials, and its membership. Each one also leaves something permanent behind, which a stopped project does not: a zoning amendment governs the next twenty applications, and the next developer reads it before choosing a site.
Pick the one you can win in the next ninety days, and make sure people can see it happen.
Match the strategy to the barrier
This is the move that separates groups that get somewhere from groups that repeat themselves. Most campaigns run one tactic against every obstacle. Diagnose the barrier first, then choose.
| If the barrier is | Don't just | Consider |
|---|---|---|
| The public does not know | Circulate a petition | Local storytelling through trusted messengers |
| Officials do not see a constituency | Send more evidence | Build a visible resident coalition |
| The information is hidden | Protest in general terms | Push for disclosure and transparency requirements |
| The rules are inadequate | Fight each project | Change the zoning or the regulations |
| The decision is politically entrenched | Argue harder | Recruit constituencies and champions with standing |
| The developer needs local legitimacy | Say no to everything | Negotiate concrete, enforceable conditions |
| Your group lacks technical expertise | Try to become experts | Connect with technical and legal partners |
| Your leaders are exhausted | Hold more meetings | Distribute roles and build peer support |
The underlying principle is the one that makes most advocacy fail: closing a data gap rarely closes the decision gap, because the decision was never waiting on the data. If officials already know the water numbers and approved it anyway, a better water report is not the intervention. Something else is holding that decision in place, and your job is to name it correctly before you spend six months on the wrong response.
Say what you want, not only what you do not
A group that only opposes is spending down its credibility with every meeting. A group that proposes gives officials something to stand beside.
So write your own version. Responsible data center principles for your county. Resident-generated siting criteria. Model zoning language your board can adopt without drafting it themselves. A scorecard any application gets measured against. Community benefit priorities decided by residents rather than offered by a developer.
Then get local organizations, businesses, residents and sympathetic officials to endorse it publicly.
At that point the conversation changes shape. You are no longer the objection. You are the standard, and the developer is the one responding to your document. Politicians can attach their name to it. Neighboring counties can copy it. Other advocates can improve it.
About the argument that protesting is pointless
You will hear this, sometimes from people with a podium, and it is worth being precise rather than defensive about it.
Protest does real work. It establishes that an issue exists, puts a number on how many people care, generates the coverage that makes a quiet negotiation possible, and shifts what a community treats as normal. Some fights have no other route.
The narrower and more useful claim is this: as a way of changing one specific decision maker's mind on one specific vote, a public ask is the weakest instrument available, and it is the one most groups reach for first. Pressure applied to someone who has already taken a position tends to entrench the position. It also asks people for an afternoon and gives them nowhere to go afterwards.
Which is why the ridicule stings less than it is meant to. A group that shows up with a decision map, a drafted zoning amendment, four hundred residents in a standing organization and a record of three smaller wins is not protesting. It is doing the county's homework, in public, on the record. That is much harder to dismiss and much harder to wait out.
Connect the wins so local action becomes something larger
No community should have to rebuild this from scratch, and right now most of them are.
The movement is already learning fast. Data Center Watch describes an emerging opposition playbook and communities increasingly borrowing from each other. FracTracker publishes its victories specifically so other advocates can use them. What is still mostly missing is structured exchange: what did you try, what happened, what would you do differently, here is our zoning language, here is the testimony that worked, here is how we got the utility commission to open a docket.
Meet the people doing this in the next county, quarterly if you can. Peer contact is what carries leaders through the flat stretches between hearings, and it is how a local objection becomes a pattern that a state legislator or a utility regulator can actually see.
There is also a fairness question underneath all of this that is worth naming. Communities with money, political access, technical expertise and existing organizational infrastructure convert opposition into policy change more easily than communities without them. Sharing what worked is the cheapest available correction to that gap.
Where to start
Do not schedule a meeting about data centers. Spend one week building the decision map: every approval this project still needs, who grants it, on what date, under what standard.
Then find four people, give each of them one decision to follow, and pick the smallest item on the ladder above that closes within ninety days. Draft the language you want rather than the objection you have.
You will have more influence in the eighth week of that than in any three minutes of public comment.
What decision is actually still open in your county, and who makes it? And what would you want the rule to say for every application after this one, not just this one?
If it is the people rather than the plan that has you stuck, the Real Change Partner is built for that conversation. Three sessions free at realchangepartner.com.
Sources
Data Center Watch, national tracking of local opposition and affected projects. FracTracker Alliance, US Data Centers Tracker (resistance and victories datasets). Coalition for Responsible Data Center Development, local organizing analysis, May 2026. Annenberg Public Policy Center, national survey on local data center siting. Reece, Rauh and Simko, analysis of local government meeting transcripts, 2007 to 2026. Bommarito, "Why Communities Fight Data Centers," working paper, 2026. Carbon Direct, review of 46 delayed or cancelled AI data center projects. Henrico County organizing case study, 2026. NAACP, Stop Dirty Data Centers campaign materials. Next 10, California data centers, water and environmental justice.