Politics in the feed 2025
This analysis shows how concentrated attention is, where the peaks landed, and how TikTok and Reels differ. We look at the short‑form output of Bundestag accounts across the full year.
We tracked 32,443 short‑form videos across TikTok and Instagram Reels in 2025 from verified political accounts. The year produced 1.010B views, 58.6M likes, 3.0M comments, and 4.74M shares. The top 1% of videos generated 49.3% of all views, and the top 0.1% still accounted for 16.0%.
The essentials at a glance
TikTok has more upside, Reels have more baseline
Concentration at the strongest accounts
Election week as the attention peak
Election week pulled the curve upward and explains a large share of the peaks. February became the dominant month of the year — not just for volume, but for engagement as well.
The election week (Feb 17–23, 2025) delivered 104.4M views — 10.3% of the year’s total — on 2,044 videos (6.3% of annual output). The single biggest day was February 22, 2025, with 31.7M views (3.1% of the entire year) across 399 videos.
This wasn’t a single viral spike. It was a stacked, multi‑day wave where posting volume and engagement rose together.
The power‑law year
Short‑form attention is brutally concentrated. The top 1% of videos captured 49.3% of all views; the top 0.1% (roughly 33 videos) still accounted for 16.0%. On the creator side, the top 10 accounts produced 42.0% of all views (top 20: 57.2%).
Takeaway: Short‑form isn’t a volume game. It’s a breakout game.
TikTok vs Instagram: ceiling vs floor
The platform split shows where attention is steady and where breakouts happen.
Instagram Reels = the floor. It carries more volume and steadier output:
TikTok = the ceiling. Fewer videos, higher upside:
Two patterns explain the gap:
- Low floor, high ceiling. Median views are lower, but TikTok’s top‑decile performance is stronger (p90 ~36.9K vs ~27.3K on Instagram).
- Power‑law distribution. The biggest single video hit 11.39M views.
If you want the framework behind these metrics, start with our guide to social media engagement metrics.
The campaign wasn’t evenly distributed
Not every party (or politician) played the same platform game. Presence is uneven — and that changes the leaderboard.
That pattern matches external research. The Friedrich‑Ebert‑Stiftung’s Swipe, Like, Vote study analyzes 1,207 official TikTok videos (Jan 1–Feb 23, 2025) and shows why remixability beats raw reach: audio‑memes, edits, and “vibes” invite participation and secondary distribution. See the FES study page and the open‑access PDF.
Takeaway: Don’t copy TV. Ship native short‑form, or the algorithm won’t reward you.
Notably, Die Linke has the fewest accounts in the dataset, yet leads in total views. That’s why we separate total attention from efficiency (views per post, engagement per post). Volume alone doesn’t tell you who is winning — and account count doesn’t automatically explain the ranking.
The party leaderboard
At the top, it’s not just reach — it’s consistency and platform mix. Some parties dominate on views, others win on engagement, and that trade‑off reshapes the ranking.
Ranked by total 2025 views and average engagement rate across short‑form.
Attention vs persuasion
The most‑viewed clip is not always the most persuasive clip. We separate raw reach (views) from engagement quality (interaction per view) to avoid confusing visibility with impact.
Top 10 videos by views (2025)
Political influencers: top accounts by platform
Who pulls the most reach differs by platform. The tables show the top accounts by 2025 views, with engagement rate and video count included.
TikTok
What the year says about 2026
- Election windows amplify everything. The two weeks before election day were the highest‑intensity period of the year.
- TikTok is the leverage channel. Higher engagement with fewer posts.
- Instagram is the stability channel. More volume, steadier output.
- Performance is unequal. A small number of posts and accounts explain a large share of attention.
For a deeper benchmarking template, see our social media analytics report template and guide to social media campaign reporting.
Actionable recommendation for parties
If attention is this concentrated, winning requires systematic learning, not just posting. Track formats, watch competitors, adapt what works, and run experiments with clear learnings each cycle. That’s what viral.app is built for: making formats, accounts, and platforms comparable so teams can iterate faster.
- Track formats: Which hooks, lengths, topics, and tones produce breakouts?
- Monitor competitors: Which formats work for other parties — and why?
- Copy‑and‑adapt, not copy‑paste: Translate patterns into your own message.
- Run experiments as a loop: hypothesis → test → readout → next version.
Methodology & scope
- Short‑form only: TikTok + Instagram Reels (no long‑form, no YouTube).
- Account coverage: built from the Bundestag member directory and verified in our backend as of late November 2025.
- Not a full census: some profiles were missing or newly created after that date.
- Best‑effort verified dataset of short‑form political accounts at the time.
- Limitation: platform visibility and algorithmic distribution can shift week‑to‑week, so results reflect observed performance, not guaranteed reach.
