1. Dynamic Voter Exclusion on Connected TV

The single most technically sophisticated innovation of the 2026 primary cycle is something unglamorous but operationally significant: campaigns can now stop running ads at people who have already voted.

JamLoop, in partnership with political data firm Aristotle, launched a product called ActiveVoter in February 2026 that pulls real-time early and absentee voting data from state election files and dynamically removes already-voted households from ad delivery on connected TV. If a voter in your district drops their mail ballot on a Monday, by Tuesday morning they are off your list.

This matters because early voting has completely changed the campaign calendar. Around 60% of US voters cast ballots early or by mail in 2024, meaning a significant share of your electorate has already made their decision before your final advertising sprint begins. Continuing to spend budget on them is not just wasteful, it is a strategic distraction from the persuadable voters you still need to move. Campaigns that adopted ActiveVoter in the primaries reported meaningfully more efficient ad spend in the final stretch. This is the kind of plumbing nobody talks about at conferences but that wins close races.

The European takeaway: most European countries do not have early voting at the same scale, but the underlying principle, stop spending on already-convinced voters and redirect to the undecided, is universal. As early and postal voting expands across the continent, this type of dynamic exclusion logic will become essential.



2. Connected TV Goes Down-Ballot

For most of modern political campaigning, television advertising was the exclusive preserve of well-funded statewide and federal races. Running a city council campaign or a state legislative primary meant you advertised on Facebook, knocked on doors, and accepted that TV was simply not your game.

That has changed. Connected TV platforms have dropped their minimum spend thresholds dramatically, with some starting at $50 per campaign. A school board candidate in a competitive Texas district can now run geotargeted video ads appearing on the same Hulu and Peacock screens as a Senate campaign, reaching exactly the households in their district and nobody outside it.

The 2026 Texas primaries saw down-ballot candidates running broadcast-quality TV spots that would have been financially impossible three cycles ago. CTV also delivers something traditional broadcast never could: the data. Which households watched the ad, for how long, and whether they took any subsequent action, all trackable and optimisable in real time. The playing field between a well-resourced incumbent and a challenger who knows what they are doing has narrowed significantly.



3. AI-Generated Ad Creative at Primary Speed

The traditional production pipeline for a TV campaign ad ran roughly four to six weeks and cost between $15,000 and $50,000 at minimum. Brief the agency, develop concepts, schedule the shoot, edit, get approvals, go to air. In a primary campaign that might last six weeks in total, that timeline effectively ruled out responsive advertising entirely.

AI creative tools have collapsed that to hours. During the 2026 Texas primaries, candidates used AI to produce broadcast-quality thirty-second spots from a campaign website or social media profile in minutes, and deployed attack ads or rapid-response content the same day as a news event. Ken Paxton used AI video to mock a Republican primary opponent. Down-ballot candidates produced multiple creative variations simultaneously and A/B tested them in real time to identify what was actually moving voters.

The implication is not just speed. It is volume. A campaign that can test ten different message framings in the first two weeks of a primary, optimise based on real performance data, and pivot before a single dollar has been wasted on a message that does not land, has a structural advantage over a campaign still waiting for their agency to come back with a concept deck.



4. Streaming Audio as the Overlooked Channel

In 2022, audio represented approximately 1% of political ad spend, despite the average American spending over 21% of their total media time consuming audio content. That gap, between where voters actually are and where campaigns were spending money, was one of the most glaring inefficiencies in political advertising.

In 2026, it is closing fast. Programmatic streaming audio, through platforms like Spotify, podcast networks, and digital radio, allows campaigns to reach voters in the moments that CTV and display advertising simply cannot access: the commute, the gym, the kitchen. These are screen-free, attention-available moments when a well-placed audio message encounters almost no competition.

Podcast advertising in particular has demonstrated outsized persuasive impact, especially among younger, higher-education demographics that are notoriously difficult to reach through traditional channels. A thirty-second pre-roll on a locally relevant podcast with 8,000 listeners in your district will outperform a display banner seen by 80,000 people who never read it. The CPMs are higher in election season, but the efficiency calculation still makes audio one of the better-value channels in a contested primary.



5. Micro-Influencers Over Macro Reach

The influencer playbook in political campaigns had, until recently, focused almost entirely on large national accounts, the same logic that drives celebrity endorsements. Find someone with a million followers, have them post approvingly, hope the reach translates to votes. It mostly does not.

What the 2025 off-cycle elections, and Zohran Mamdani's successful New York City mayoral campaign in particular, demonstrated is that hyper-local micro-influencers with deeply engaged niche audiences are dramatically more valuable than macro accounts with passive followings. A local food blogger with 12,000 highly engaged followers in a specific neighbourhood, a TikTok creator who covers high school sports in your district, a Reddit moderator with authentic community trust: these voices move opinions in ways a national influencer cannot replicate because they are already embedded in the relationship of trust that political campaigns spend years trying to manufacture.

Going into 2026, specialist firms are pitching campaigns on exactly this logic, mapping local creator ecosystems by district and identifying the accounts whose audiences overlap most precisely with target voter segments. The ceiling is lower. The conversion rate is higher. And the cost is a fraction of national influencer deals.



6. ACR Targeting to Find Broadcast Viewers on Digital

Automatic Content Recognition is a technology built into smart TVs that identifies what is playing on screen, whether broadcast, streaming, or cable, and logs that data at the household level. For political campaigns, it enables something previously impossible: you can identify households that watched a specific competitor's ad on broadcast TV and then serve your response ad to those exact households on digital channels.

If your opponent is running a heavy broadcast television campaign attacking you, you can serve counter-messaging directly to the households that saw the attack, rather than blanketing an entire media market. You can also identify households watching news programming, political content, or issue-specific shows associated with your target voters, and reach them with far greater precision than any broadcast buy could achieve.

The 2026 Texas primaries saw campaigns using ACR to build custom audiences from broadcast viewership data and activate them across CTV, programmatic display, and social simultaneously. It is one of the cleaner examples of a technique that genuinely crosses from sophisticated to essential in a competitive race.



7. AI Chatbots for Personalised Voter Contact

Research published before the 2026 cycle produced a striking finding: brief AI chatbot conversations shifted US voters' opinions on political issues by up to 10 percentage points, a movement nearly four times larger than traditional 2016 and 2020 political ads achieved. The persuasive ceiling of a well-designed conversational AI interaction appears to be significantly higher than a passive ad impression.

Campaigns are now deploying AI chatbots in text message and web chat formats to answer voter questions, deliver personalised issue messaging, and qualify supporters for follow-up contact. The chatbot knows which issues the voter has indicated they care about, serves relevant policy information, and routes engaged contacts to volunteer or donation funnels. Utah Senate campaigns used AI chatbots to deliver candidate talking points in private text message conversations with voters at scale, with the AI prepared for individual conversations in ways that human volunteers cannot sustain across thousands of simultaneous interactions.

The ethical and transparency questions here are real and not resolved. Several states are introducing disclosure requirements. The technology is ahead of the regulation, which is precisely the moment at which understanding it matters most.



8. AI Opposition Research and Rapid Response

Opposition research has traditionally been a labour-intensive, time-consuming process: researchers digging through voting records, financial disclosures, old news articles, and social media archives to build attack profiles. Compressing that from weeks to hours changes the tactical rhythm of a primary campaign entirely.

In the 2026 primaries, campaigns used AI tools to scan decades of public records, flag inconsistencies in opponents' stated positions, and surface old quotes or votes that contradicted current campaign messaging. The same tools monitored opponents' social media in real time and flagged moments that warranted a response within the news cycle. What previously required a research team of six working for a month can now be accomplished by one researcher with the right AI tools working for a week.

The speed advantage is not just about attack ads. It is about debate preparation, rapid response to breaking news, and identifying the precise moments in a campaign when your opponent is most vulnerable to a well-timed digital push.



9. Down-Ballot Campaigns Get Their Own Data Infrastructure

For most of political history, sophisticated voter data infrastructure was the exclusive province of national parties and well-funded federal campaigns. Local candidates operated on gut instinct, local knowledge, and voter rolls that had been cleaned roughly as often as a gym locker room.

AI-powered campaign tools that automate voter targeting, modelling, and audience analysis are now being offered to candidates at the state legislative, city council, and school board level at price points that make it accessible to campaigns with five-figure budgets. The basic idea is that a consultant using AI to automate the analytical work can offer services to down-ballot candidates at a fraction of the cost of a traditional data operation.

The 2026 primaries saw genuinely local races deploying micro-targeting and predictive modelling that would have been unrecognisable to a local campaign manager five years ago. The democratisation of data infrastructure is one of the most significant and least discussed shifts in American campaign politics.



10. Platform Blackout Strategy as a Campaign Discipline

This one is less a technique and more a survival skill, but it is new enough that many campaigns in the 2026 primaries got badly wrong-footed by it.

Meta shuts down new political ads seven days before Election Day. Google blocks political microtargeting based on party affiliation data and bans new election ads once polls close. TikTok bans paid political advertising entirely. X operates under fluid and unpredictable policies. These are not new rules, but the 2026 cycle saw campaigns that had structured their entire digital strategy around a final-week paid push suddenly discover they had spent their way to the finish line with no ability to launch new creative at the most volatile moment of the race.

The campaigns that understood this built their digital strategy backwards from the blackout dates. All message testing, persuasion creative, and contrast advertising was deployed and optimised before the seven-day window. The final week was treated as an execution phase for already-tested content rather than a launch phase for new messaging. It sounds obvious. It is not standard practice. And in a close primary race decided by a few hundred votes, being locked out of new messaging during the final week is the kind of operational mistake that does not get made twice.



A Final Note on Scale and Context

The US midterm environment is structurally different from European campaigning in almost every dimension: the money, the geography, the regulatory framework, the media market fragmentation, and the political culture. $10.8 billion for a midterm cycle is a number that has no European equivalent and probably should not.

But the techniques in this list do not require $10.8 billion. Most of them scale to a parliamentary constituency campaign, a regional election, or a well-funded local race. The underlying logic, reach the right voter at the right moment on the right channel with the right message, is identical whatever the budget or the flag. The US primaries run the experiment. The rest of us get to read the results.