One SaaS Founder Answered 21,623 Support Tickets Over 8 Years. Here's What Actually Reduced the Volume.
Geoff Roberts at Outseta responded to 21,623 support tickets over eight years, personally handling more than 10,000 of them.
That's not a case study. That's a confession.
And the patterns he found match everything the research shows about why SaaS support doesn't scale — and what actually works.
The Responsiveness Trap
Roberts describes a cycle that every founder-who-handles-support recognizes:
You answer quickly because you care. Customers notice. They start expecting minute-level responses. More tickets come in. You answer faster. The expectation ratchets up. You're now a prisoner of your own helpfulness.
His exact observation: being deliberately less responsive caused customers to solve their own issues. Fewer tickets, not more.
This isn't a character flaw. It's a system design problem.
The Numbers Behind the Pain
Salesforce's 2025 State of Service survey found:
77% of service reps say their workload and complexity have increased compared to a year ago.
56% report experiencing burnout.
69% of service decision-makers say agent attrition is a major challenge.
Only 26% feel they have all the tools they need to succeed.
That's the agent side. On the customer side:
61% prefer self-service for simple issues (Salesforce 2025). 81% attempt self-service before contacting support (CEB/Gartner). But only 14% of self-service journeys fully resolve the issue (Gartner 2024, n=5,728).
Customers want to help themselves. Companies fail to let them.
The Root Cause Is Onboarding
Most support tickets aren't product bugs. They're onboarding failures.
The data (UserGuiding, aggregated from multiple studies):
40-60% of SaaS users who sign up never return after their first experience.
90% churn risk if users don't engage within the first 3 days.
72% abandon apps if onboarding requires too many steps.
43% of churn stems from unclear next steps during onboarding.
But here's the flip side:
35% fewer support tickets in the first month with video onboarding.
65% fewer support tickets overall from strong onboarding programs.
The support queue is a symptom. The disease is that customers never learned how to use your product.
What Actually Worked (Outseta + Industry Data)
Roberts found four things that moved the needle. All of them are backed by broader research:
1. Strategic response time. Being slightly less responsive didn't hurt satisfaction — it reduced ticket volume. Customers who wait 20 minutes often figure it out themselves. The Outseta team consolidated to a single support channel (tickets only, no chat, no social). Simplicity won.
2. Personalized video responses. Instead of scheduling live calls for complex issues, Roberts recorded screen-share videos. Customers could replay them. No scheduling overhead. Higher satisfaction. And the videos became reusable training content.
3. Reframing support as revenue. When Roberts tracked it, customer service directly generated revenue — through retention, expansion, and referrals. This changes the psychology: you're not "handling tickets," you're building relationships that pay.
4. Building self-serve before you need it. Companies that implement good self-service see up to 70% reduction in inquiries (Gartner 2018). But "good" is the key word. Most self-service fails — that's why only 14% of journeys fully resolve.
The Scaling Wall Nobody Talks About
Gainsight, the company that practically invented the Customer Success category, published data in 2026 acknowledging the obvious: "Simply adding more CSMs is not a sustainable way to scale."
Teams that rely on effort alone — more people, more hours — see "diminishing returns, rising burnout, and uneven customer experiences."
Nearly half of Customer Success Managers now own upsells, renewals, and cross-sells on top of support. That's three jobs masquerading as one.
The "do more with less" mandate is real, and it's universal across B2B SaaS in 2025-2026.
The Pattern
Here's what the journey looks like for most SaaS teams:
Stage 1 (1-50 customers): "I can handle support myself." And you can. It even feels good — direct customer contact.
Stage 2 (50-200 customers): "I'm spending half my day answering the same three questions." The repetition starts. Product work gets displaced.
Stage 3 (200-500 customers): "We need to hire a support person." But without systems, the new hire asks the founder the same questions customers asked. Knowledge transfer fails.
Stage 4: "We need a knowledge base." So they buy Zendesk, set up a Notion wiki, or start a "customer academy" project. Content creation stalls at month 2-3. (This matches industry data: only 4% of companies have formalized, scalable customer education — Intellum 2024.)
Stage 5: "We have all these tools but tickets keep coming." The tools aren't the problem. The content and structure are.
Roberts lived this entire arc across 21,623 tickets and eight years.
The Nervous System Connection
Every unresolved ticket is a cortisol hit for the person who has to answer it.
We've covered the neuroscience before: chronic stress physically shrinks the prefrontal cortex (Savic et al., 2018, Cerebral Cortex, n=128, first longitudinal MRI study). Working 70 hours produces the same output as 56 (Pencavel, 2014, Economic Journal). Support burnout (56% of agents) isn't separate from founder burnout — it's the same physiological process.
The support queue isn't just a business problem. It's a health problem. And it's fixable — not with more willpower, but with better systems.
What This Means
The founder who answered 21,623 tickets isn't unusual. He's just the one who counted.
If you're building a B2B SaaS product, you already know the pattern. The question is whether you build the education infrastructure before Stage 4 or after you've already burned through your support team.
The data suggests earlier is better: 65% fewer tickets from strong onboarding. But only 4% of companies get there.
The other 96% are answering the same questions. Again.
Sources
1. Roberts, G. "Lessons in SaaS Customer Service from Responding to 21,623 Support Tickets." Outseta. First-party data.
2. Salesforce. "State of Service 2025." Industry survey.
3. Dixon, M., Toman, N., & DeLisi, R. (2017). "Kick-Ass Customer Service." Harvard Business Review / CEB-Gartner. 81% self-service attempt rate.
4. Gartner. "Customer Service and Support Survey 2024." n=5,728. 14% self-service resolution rate.
5. UserGuiding. "100+ User Onboarding Statistics 2026." Aggregated from multiple studies.
6. Gartner (2018). Virtual customer assistant data. Up to 70% inquiry reduction.
7. Gainsight. "What Customer Success Teams Are Prioritizing in 2026." January 2026.
8. Savic, I. et al. (2018). "Structural Changes of the Brain in Relation to Occupational Stress." Cerebral Cortex, 28(5), 1534-1543. n=128.
9. Pencavel, J. (2014). "The Productivity of Working Hours." Economic Journal, 124(589), 2052-2076.
10. Intellum/Forrester (2024). Customer education maturity data. 4% formalized and scalable.