← back to live analytics
The Swipe File · Benchmark Report

2026 U.S. Senate
Digital Ad Playbook

What every Senate campaign is saying, showing, and spending on Meta — and which creative tactics actually keep ads alive and in front of voters.
17,999
unique creatives
122,038
ads analyzed
234
candidates
$86M
ad spend
2.8B
impressions
Coverage 2018-11-06 – 2026-06-30 · every creative classified on 22 visual + message axes by a Sonnet multimodal pass · generated 2026-08-27

Executive summary

43%
of unique creatives are fundraising — the dominant message type. With 6.8× ad-variant churn, its share of raw ad *volume* is far higher still.
+116%
longer ad life for creatives tagged “survey poll” (cta visible) vs. the field average — the single strongest longevity lever.
-48%
shorter life for “image heavy” (text density) — the weakest-performing tactic on longevity.
25%
of creatives are video (4,480 of 17,999); the rest are static graphics, photos, and screenshots.
11.2%
of recent ads use attack framing — up from 10.8% earlier in the window. Negativity climbs as the cycle heats up.
40%
of creatives center democracy elections — the most-invoked issue across the field.

01 · The message mix

What campaigns are trying to do with each ad.

By unique creative
fundraising 7,441
persuasion 4,445
list building survey 1,931
awareness 1,852
gotv mobilization 1,589
By ad volume, month over month
2025-07 1,979
2025-08 2,688
2025-09 2,825
2025-10 2,869
2025-11 1,690
2025-12 2,136
2026-01 2,370
2026-02 3,248
2026-03 4,749
2026-04 4,029
2026-05 4,978
2026-06 4,735
Volume ≠ variety. The 122,038 ads collapse to just 17,999 unique creatives — a 6.8× churn ratio. Fundraising drives that churn: campaigns spin up dozens of near-identical variants to A/B-test subject lines and match deadlines, so it dominates ad *volume* even more than the creative count suggests.
Who churns hardest — ad-variants per creative
candidateadscreativesads / creative
OR — Jeff Merkley6,90416442.1×
SC — Annie Andrews3,24914822.0×
IA — Josh Turek4,00820119.9×
NM — Ben Ray Luján1,1786518.1×
TX — John Cornyn6,06133718.0×
WY — Cynthia Lummis8595116.8×
MI — Abdul El-Sayed6,35539416.1×
ID — Jim Risch4763115.4×

02 · What actually works

Creative traits ranked by how much longer their ads stay live than the 12-day field average — a revealed-preference signal (campaigns keep what performs).

Extends ad life ▲
traitaxislift
survey pollcta visible+116%
list building surveymessage type+92%
sign petitioncta visible+75%
social security medicareissue+73%
housingissue+67%
jobs laborissue+51%
immigrationissue+40%
ominous threateningvisual mood+30%
Dies fast ▼
traitaxislift
image heavytext density-48%
gotv mobilizationmessage type-45%
unknowncandidate portrayal-40%
votecta visible-37%
no texttext density-33%
unknownvisual format-33%
educationissue-32%
volunteercta visible-30%
How to read this: lift is observational, not causal — it shows correlation between a creative trait and ad longevity across 17,259 creatives (minimum 40 per group). Longevity is a proxy for performance because campaigns pause ads that underperform; it does not control for budget, candidate, or timing.

03 · The issue landscape

Which issues the field is advertising on, and how salience shifts month to month.

Top issues by creative
democracy elections 6,748
candidate character bio 2,577
other 2,082
economy taxes 1,186
healthcare 859
immigration 508
jobs labor 494
energy environment 471
abortion 415
foreign policy defense 405
crime safety 386
education 277
Issue salience by month (% of that month's ads)
070809101112010203040506
democracy elections 454244514856454250504551
candidate character bio 17191118175121616171819
other 58139101115121110129
economy taxes 116977851110987
healthcare 71110657423545
immigration 210111861111

04 · Visual & creative tactics

The look of the field — formats, production values, and emotional register.

Format / media mix (4480 video)
video frame 4,480
unknown 4,400
graphic text card 3,059
photo 2,434
mixed 1,539
social screenshot 1,087
news screenshot 921
illustration 64
meme 11
Production tier
semi pro 9,312
professional 1,741
diy amateur 1,624
unknown 202
Emotional appeal
urgency 6,394
pride 2,783
anger 2,494
fear 2,155
empathy 1,480
hope 1,408
neutral 385
nostalgia 159
Visual framing
positive candidate 7,456
contrast comparison 2,221
attack opponent 1,896
issue advocacy 1,061
informational 245

05 · The attack picture

Negativity over time, and which campaigns go negative most.

Attack-ad rate by month
2025-07 10.8%
2025-08 9.0%
2025-09 8.7%
2025-10 6.6%
2025-11 9.1%
2025-12 7.7%
2026-01 13.8%
2026-02 16.6%
2026-03 11.6%
2026-04 7.6%
2026-05 14.6%
2026-06 11.2%
Most attack-heavy campaigns (share of their creatives)
candidateattackshare
FL — Ashley Moody12100%
AR — Tom Cotton12251%
KY — Nate Morris550%
SC — Mark Lynch2244%
NH — Chuck Morse3343%
MN — Adam Schwarze343%
TX — Jasmine Crockett1642%
IL — S Krishnamoorthi9741%
IA — Thomas Laehn240%
TN — Robert Jones436%

06 · Proven creatives

The longest-running ads in the field — median lifespan is 5 days; these are the outliers worth studying.

advertiserdays livereach (lower–upper)
Colin Allred300d 20,000 – 24,999
Colin Allred299d 8,000 – 8,999
Bluejay Rising296d 250,000 – 299,999
Colin Allred294d 7,000 – 7,999
Colin Allred294d 7,000 – 7,999
Colin Allred294d 15,000 – 19,999
Colin Allred294d 9,000 – 9,999
Logan for Kentucky285d 8,000 – 8,999
Logan for Kentucky285d 5,000 – 5,999
Logan for Kentucky285d 8,000 – 8,999
Logan for Kentucky285d 4,000 – 4,999
Logan for Kentucky285d 4,000 – 4,999
Go deeper than the report

Get the underlying dataset

This report is the summary. The full asset — every creative with all 22 axes, ad-level spend & reach, per-candidate reports, and the images — is licensable for your own analysis.

See access & licensing →
Free sample & live analytics at adlibrary.alaskatargeting.com

Methodology & caveats

Source. Ads are pulled from Meta's public Ad Library API for U.S. Senate 2026 advertisers (234 candidates, 287 advertiser pages), covering 2018-11-06–2026-06-30. Spend and reach are Ad-Library range midpoints.

Analysis. Every ad is reduced to a unique creative by hashing its normalized copy, so near-identical variants are counted once (122,038 ads → 17,999 creatives). Each creative is classified once by a Claude Sonnet multimodal pass on 22 visual and message axes, then joined back to every ad that shares it.

Caveats. Effectiveness figures are observational, not causal — ad longevity is a revealed-preference proxy (campaigns pause ads that underperform) and is not adjusted for budget, candidate, or launch timing. Ad-Library ranges are wide. Read directional, not precise.

The Swipe File · generated 2026-08-27 · figures refresh as new ads are ingested.