Propaganda
Labs.
The systematic study of what actually works in B2B SaaS growth — grounded in behavioral science, published in the open.
Most growth marketing is built on anecdotes. A tactic works for one account, gets written up as a "playbook," and circulates through agencies until the entire market has overfit to it. Propaganda Labs exists to replace anecdote with evidence.
Every study we publish starts with a written hypothesis, a defined sample, a falsifiable threshold, and a commitment to publish null results alongside wins. We draw on behavioral economics, consumer psychology, and applied cognitive science — the disciplines that actually explain why buyers do what they do. Not the disciplines that sell the most ad tech.
The goal is simple: when we say "scarcity beats feature explanation in cold B2B SaaS traffic," we want to be able to point to a study with a sample size, a methodology, and a CSV you can download. Not a LinkedIn post.
The Scarcity Effect in B2B SaaS Landing Pages.
A 47-variant analysis of urgency framing across 12 B2B SaaS accounts, Q4 2025 — Q1 2026.
Across 12 B2B SaaS client accounts, we tested 47 variations of scarcity framing on landing page hero sections. Scarcity-led copy (e.g., "only 8 spots remaining this quarter") outperformed value-led copy by a median of +47% CTR and −68% cost per lead. The effect held across 9 of 12 accounts and was strongest in early-stage SaaS, where buyer urgency is structurally lowest.
Published Research.
Every study is peer-reviewed by at least one external reader, replicated where possible, and released with its raw methodology. Null results are published alongside wins.
3 PUBLISHED · 5 IN LAB
The Founder Authority Effect
Why unpolished phone-camera founder video outperforms studio-produced ads by 2.4× in B2B SaaS — a longitudinal analysis of 340+ creative variants across 18 months.
The Attribution Gap
Platform-reported conversions inflate actual closed-won pipeline by a median of 38% across B2B SaaS accounts. A reconciliation study of 4 portfolios against server-side CRM data.
Cognitive Load & Form Completion
Every additional form field reduces completion by a median of 11%. But field type matters more than field count — replacing text inputs with pill selectors lifts completion by 34%.
Currently In The Lab.
Studies move through Protocol Design → Data Collection → In Review → Published. Subscribe below to get notified when each one lands.
5 IN PROGRESS
MAY 2026
JUNE 2026
JULY 2026
Q3 2026
Q3 2026
The Rules We Work By.
Eight commitments that separate applied research from agency case studies. We publish the methodology, not just the result.
8 PRINCIPLES
Every study starts with a written hypothesis.
Before we collect any data, we write down what we expect to find and why. If the result contradicts the hypothesis, we publish the contradiction.
Sample size is disclosed up front.
We tell you how many accounts, how many variants, how long the study ran. Small-sample studies are labeled as such — not marketed as definitive.
Replication before publication.
Every finding is tested on at least one additional account before publication. Single-account "wins" are not studies — they're anecdotes.
All client data is anonymized.
Brand identities are removed by default. NDA-bound work is aggregated with other accounts and labeled as such. The numbers remain; the names don't.
External peer review.
Every study is read by at least one external reviewer — a behavioral scientist, academic, or domain expert — before it ships. We publish their affiliation when permitted.
Raw data available on request.
If you want to check our math, email us. We'll send the anonymized dataset, methodology notes, and analysis code. No gatekeeping.
Null results are published.
When a study finds no effect — or one that contradicts our hypothesis — we publish it with equal weight. Publication bias is how industries get fooled.
No sponsored studies. Ever.
Propaganda Labs does not accept funding to reach a specific conclusion. We study what we find interesting, on our own clients' budgets, with zero strings.
Accepting Research Affiliates For 2026.
Propaganda Labs is looking for 1–2 research affiliates to co-sign studies and bring academic rigor to the methodology. We're specifically interested in PhD students, postdocs, or independent researchers in consumer psychology, behavioral economics, or applied cognitive science.
Affiliates are paid a quarterly stipend, get their name on published studies, and retain the right to cite the data in their own academic work. Your rigor, our applied data.
Apply via Email →What We're Looking For
- Background in behavioral science, consumer psychology, or behavioral economics (PhD in progress or completed)
- Interest in applied research with access to real commercial data, not lab-only studies
- Rigor — willingness to call out methodology issues even when they slow down publication
- Availability for 8–12 hours per month, quarterly cadence
- Comfort working with NDA-protected datasets in commercial context
Get Each Issue The Day It Ships.
Propaganda Labs publishes one issue per quarter. New studies, raw data links, and research notes from the active accounts. One email per quarter. Unsubscribe any time with zero guilt trip.