CASE STUDY · ScreenJournal AI x Starlabs

How We Turned a Broken Outbound Into 1 New Signup Every Day

AI-first outbound infrastructure for an early-stage B2B SaaS.

CAMPAIGN RESULTS

10×

Reply Rate Increase

from <1% → 2%

150×

Bounce Rate Reduction

from 15.1% → 0.1%

~1/day

New Signups

consistent, daily pipeline

Companies that trust StartLabs

LogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogoLogo

Before vs. After in Salesforge

The "BEFORE" sequence (Outreach): 5,346 contacts sent, 825 bounces (15.1%), only 4 replies (0.2%). The "AFTER" sequence (C1 ICP 1): 1,097 contacts sent, 2 bounces (0.1%), 30 replies (2.0%).

What We Fixed :

What We Fixed

Rebuilt the Sending Infrastructure

Matched sender domains to receiver domains to improve trust signals.

Removed low-deliverability domains dragging down inbox placement.

Warmed up new Outlook domains over several weeks before scaling.

Result: bounce rate dropped from 15.1% to 0.1% — a 150× improvement.

What We Fixed

Cleaned, Enriched & Curated the List

Removed all invalid and risky email addresses from the database.

Filtered contacts for genuine ICP fit — cutting volume in favour of precision.

Removed all invalid and risky email addresses from the database.

What We Fixed

Ran Speed Blitz Messaging Tests

Tested multiple short, punchy copy variations simultaneously — no multi- week debates

Structured tests around three variables: problem statements, solution statements, and CTAs

Identified winning angles within days, then doubled down on what worked.

What We Fixed

Tested Multiple Verticals to Find the Right ICP

Ran outreach across multiple industries in parallel rather than betting on one.

Tracked reply rates and downstream conversion by vertical and company profile.

Identified the verticals where ScreenJournal resonated most strongly.

Concentrated volume on proven segments — unlocking consistent daily signups

Replies Rolling In

"The product was not the issue. The distribution was.

— StartLabs team

Ready to build your GTM engine?

Get In Touch

Copyrights 2026 | StartLabs™ | Terms & Conditions

Additionally, this site is NOT endorsed by YouTube™, Google™, Facebook™, Meta™, WhatsApp™, or Instagram™ in any way. FACEBOOK™, META™, WHATSAPP™, and INSTAGRAM™ are trademarks of Meta Platforms, Inc. GOOGLE™ and YOUTUBE™ are trademarks of Google LLC.