SMB Outreach Case Study: 171 Emails, Real Scores, What Worked
We sent 171 emails to local small businesses offering AI visibility scans. Here's the full breakdown — open rates, bounce rates, scan scores, and what we learned.
The Experiment
Goal: Test whether local SMBs are interested in AI visibility scans when offered via cold email.
Target: Local businesses only (cafes, salons, plumbers, dentists, etc.) — no tech companies or SaaS businesses.
Method: Automated outreach through our /api/automate endpoint, with spam protection and daily sending limits.
Total sent: 171 emails across 3 batches
Targeting Criteria
We only targeted businesses with low AI visibility scores (under 35 out of 100), based on pre-scanned data:
- Score range: 13–34 (most clustered around 20–30)
- Industries: cafes, salons, barbershops, dentists, vets, hardware stores, restaurants, gyms, bookstores, pet care, auto repair
- Geography: US-based local businesses
What We Sent
Each email included:
- Personalized greeting with business name
- Their current AI visibility score (if already scanned)
- Brief explanation of what AI visibility means
- CTA to get a free scan at
sottovox.com - Unsubscribe link (CAN-SPAM compliant)
Metrics (So Far)
| Metric | Value |
|---|---|
| Total emails sent | 171 |
| Daily sending limit | ~37/day |
| Sent per batch | 5–10 per run |
| Average score of targets | ~26 |
| Open rate | TBD (from Resend) |
| Click rate | TBD |
| Bounce rate | TBD |
Score Distribution
Most local SMBs we targeted scored between 13 and 34 out of 100. A few were filtered out (score >35) to focus on the ones with lowest visibility.
Why low scores matter: these are businesses that won't be cited by AI in answers like "best plumber near me" or "cafes open now." They need help the most.
What Worked
✅ Personalization at scale
Using real business names and industry categories made emails feel less template-y. We pulled names from Google Business Profile data.
✅ Clear value proposition
"Your AI visibility score is X — here's how to improve it" is concrete. Not vague "improve your online presence."
✅ Automation with safeguards
Our /api/automate endpoint handled:
- Redis deduplication (no spam)
- Daily sending limits
- Queue management
- Status tracking
What We're Still Measuring
- Open rates: Need to pull from Resend dashboard
- Click-throughs: How many clicked the scan link?
- Scan conversions: How many actually scanned their site after email?
- Sign-ups: How many upgraded to paid plans?
Surprising Insights
Local businesses are largely invisible to AI
Most scores under 35 mean these businesses won't appear in AI-generated answers at all. That's a massive blind spot for SMBs focused only on Google Local SEO.
The "AEO gap" is real
Businesses optimized for Google (GBP complete, reviews good, website SEO decent) still scored low because they're missing: FAQ pages, schema markup, llms.txt, and content depth.
Email still works for SMB outreach
Cold email to business owners (not consumers) has decent engagement when the offer is specific and relevant.
Next Steps
- Pull full Resend analytics (opens, clicks, bounces)
- Follow up with non-responders after 7 days
- A/B test subject lines
- Expand to next batch of 100+ SMBs
- Track scan-to-paid conversion funnel
Technical Takeaways
The automated outreach system proved reliable:
- Redis queue handled bursts and retries
- Deduplication prevented over-messaging
- Daily limit respected Resend's sending limits
- API-first design made it easy to script batches
Bottom Line
Local businesses need AI visibility help. Most are flying blind. Our outreach confirms demand — now we're measuring conversion.
📊 Raw Data Summary
Batch 1: Added 20 SMBs (mainstreetcafe.com → primecuts.com), sent 19
Batch 2: Added 20 more (acehardware.com → sunsetsalon.com), sent 18
Total: 40 added, 37 sent today, 171 all-time (as of 2026-04-24)
Score range: 13 (Corner Pharmacy) to 34 (Pet Palace) — all <35 threshold