01 · The Plays

The work your team does every week. Already shaped to run.

47 in production
12 in beta
PLAY-001 PR & Outreach

PPS Savant

Drop in a study or dataset. Get back the 10 most newsworthy stats, ranked by impact and emotional hook, with 10 subject-line variants and a structured pitch ready to send.

OUTPUT · PRE-PITCH STRATEGY10 RANKED · 10 SUBJECTS
01 "1 in 4 Thanksgiving flights delay 2hr+" Top hook
02 "Atlanta and Chicago lead worst-airport list" Local
03 "Tuesday now beats Wednesday for delays" Contrarian
~8 MIN · IN: STUDY OR DATASET · OUT: FULL PITCH PACKAGE
PLAY-002 PR & Outreach

Newsjacking Pitch Writer

Find the unfinished stories in this morning's news cycle. Pitch your client's data into the right gap inside the 24-72 hour window. Before the story moves on.

OUTPUT · NEWSJACK PITCHWINDOW: 36HR
● WSJ · Grocery prices hit 18mo high · 2hr ago Live
→ Angle: Local Cost-of-Living Service
~6 MIN · IN: CLIENT DATA + CYCLE
PLAY-003 PR & Outreach

Personal Connector

Replace generic openers with real connections. Pulls a journalist's last 6 to 12 months of coverage, finds the through-line, builds the transition into your pitch.

OUTPUT · CONNECTOR LINE3 ARTICLES CITED
"Your reviews of workplace tech are always insightful, especially your Apple vs. Android coverage. Our latest study shows..."
CITES: 3 ARTICLES · PAST 8 MO
~4 MIN · IN: JOURNALIST + ANGLE
PLAY-004 Content / Concept

Idea Printer

Raw concept in. Client-ready campaign mission out, with the press-friendly headline, the data sources to back it, and the meta-rankings the press will run.

OUTPUT · CAMPAIGN MISSIONBRIEF #023
The Sleepcation Index
"X% of Americans booking 'sleepcations' to combat record-high burnout"
SOURCES: 2 INTERNAL · 5 EXTERNAL · RANKINGS: YELP + GOOGLE
~7 MIN · IN: CONCEPT + AUDIENCE
PLAY-005 Research / Survey

Survey Writer

Census-quality surveys built backwards from the headline. Each question scored for newsworthiness. Question-by-question rationale included so your team can defend every choice.

OUTPUT · SURVEY DRAFT23 QUESTIONS · CENSUS GRADE
Q1 "Do you book travel solely to rest?" Headline
Q2 "Rate last week's sleep, 1-10" Stat
HEADLINE PREVIEW: "X% OF AMERICANS BOOKING SLEEPCATIONS..."
~14 MIN · IN: TOPIC + AUDIENCE · OUT: FULL SURVEY + RATIONALE
PLAY-006 PR & Outreach

Journalist Pitch Writer

Adapts a single campaign into the right voice for each outlet. Standard Exclusive for Tier 1. Formal Media for finance and business. Regional for state press. Value-Add Follow-up that never says "just checking in."

OUTPUT · MULTI-FORMAT4 MODES READY
● Standard Exclusive Tier 1
● Formal Media · CNBC, HBR voice Polished
● Regional · state ranking lead Local
~5 MIN · IN: CAMPAIGN + OUTLET LIST
Don't see your work here? Build your own Play, or tell us what we should build next. See all Plays →
What's behind every Play

Each expert is filled with research the base models don't have.

A wrapped chatbot can only repeat what's already in its training data. Our experts are different — each one is built with a research pipeline that pulls real source material, tests it against what major models already know, and keeps only what's novel. Here's a sample.

Research File · Digital PR Expert Updated May 2026
Finding #047

"Slack Connect intros now outperform cold email by 3.4x for first-touch reply rates among B2B SaaS journalists, up from 1.1x in 2024."

Sources
  • The Information · Q1 2026 industry survey
  • Pressfarm outreach data · 2026-03
  • Fractl client engagement data · 2026 Q1
Novelty test
GPT-4o✓ Not in training data
Claude Sonnet 4✓ Not in training data
Gemini 2.5✓ Not in training data
Verified novel · May 12, 2026
Research File · Survey Research Expert Updated April 2026
Finding #112

"Mobile-first survey designs now produce 2.7x higher completion rates among B2B respondents under 45, even when total question count exceeds 30."

Sources
  • SurveyMonkey 2026 industry data
  • Qualtrics methodology paper · 2026
  • Fractl survey deployments · last 18 mo
Novelty test
GPT-4o✓ Not in training data
Claude Sonnet 4✓ Not in training data
Gemini 2.5✓ Not in training data
Verified novel · April 28, 2026
Research File · Newsjacking Expert Updated April 2026
Finding #023

"First-mover newsjacks placed within 18 hours of breaking news earn 4.2x more pickups than responses placed at 24+ hours, even when the latter has stronger data."

Sources
  • Cision media monitoring · Q1 2026
  • Muck Rack journalist survey · 2026
  • Fractl reactive PR engagements · 24 mo
Novelty test
GPT-4o✓ Not in training data
Claude Sonnet 4✓ Not in training data
Gemini 2.5✓ Not in training data
Verified novel · April 21, 2026
Research File · Personal Connector Expert Updated April 2026
Finding #089

"Journalist openers that cite a specific article from the past 60 days earn a 3.1x higher reply rate than openers citing brand or beat-level work."

Sources
  • Pressfarm outreach study · 2026
  • Muck Rack journalist preferences · Q1 2026
  • Fractl cold-pitch performance data
Novelty test
GPT-4o✓ Not in training data
Claude Sonnet 4✓ Not in training data
Gemini 2.5✓ Not in training data
Verified novel · April 15, 2026
Research File · Idea Printer Expert Updated April 2026
Finding #034

"Campaign concepts framed around contrarian behavioral data outperform conventional 'big number' angles by 2.3x in Tier 1 placement rates."

Sources
  • Fractl 2024-2026 campaign performance
  • Cision pickup analysis · 36 mo
  • Muck Rack tier-1 placement data
Novelty test
GPT-4o✓ Not in training data
Claude Sonnet 4✓ Not in training data
Gemini 2.5✓ Not in training data
Verified novel · April 8, 2026
Expert files grow weekly · 1,243 verified-novel findings · 3-model novelty test on every entry
The answer is the output

Output is the only thing
that actually matters.

Run your first agent

Built inside Fractl · Used every day · 47 Plays in production