Podcast Summary: AI SDR Learnings, Results, and ROI
Podcast: AI to ROI (fka Metrics that Measure Up)
Episode Title: AI SDR Learnings, Results, and ROI - with Jacco van der Kooij, Winning by Design
Host: Ray Rike
Guest: Jacco van der Kooij, Founder & CEO, Winning by Design
Date: March 2, 2026
Episode Overview
In this episode, Ray Rike sits down with Jacco van der Kooij to uncover the journey, challenges, and ROI of deploying an AI-powered Sales Development Representative (SDR) at Winning by Design. The conversation dives deep into why the AI SDR project was started, lessons learned, the importance of designing AI for the buyer’s journey, and concrete business results. This episode is packed with practical insights for enterprises considering AI in sales or other front-line roles.
Key Discussion Points
1. Catalyst for AI SDR Deployment
- Declining ROI in Traditional SDRs:
- Traditional SDRs saw their effectiveness plummet—from 10-20 meetings/month down to 3-4, and conversion rates dropped from ~28% to low teens (04:01–04:46).
- Rising customer acquisition costs rendered human SDR ROI questionable.
- Global Limitations & Human Constraints:
- SDR turnover, limited hours due to global time zones, and poor CRM hygiene further motivated the move (06:02–07:06).
“We were hitting what we would call our human limitations… Fatigue is a human limitation. The inability to cover multiple hours was a human limitation. The CRM hygiene issues are human limitations.”
— Jacco van der Kooij (06:45)
2. AI SDR: Approach, Hypotheses, and Systematic Mindset
- System Over Tool:
- Instead of swapping a human for a tool, they designed a system combining leading AI products and in-house expertise; the AI SDR was conceptualized as the first impactful buyer touch (07:35–08:20).
- Factory Model:
- Emphasized running revenue operations like a factory—“system thinking” over point solutions (08:41–09:37).
“When you run a revenue operation, you actually have to run it as a factory and factories run as a system.”
— Jacco van der Kooij (08:50)
- Initial Hypothesis:
- The aim wasn’t just efficiency, but a better buyer experience (10:00–10:40).
- Designed the AI SDR (“Jack”) to look and act like Jacco to project trust and familiarity (10:55–11:54).
3. Lessons Learned & Aha Moments
- Not Just a Human Clone:
- Jack, the AI SDR, quickly surpassed the usual SDR skillset; capable of deep product knowledge, spontaneous slide sharing, and handling any buyer direction (12:06–12:53).
- Training the AI:
- The initial attempt to feed Jack the whole knowledge base failed; successful onboarding required narrow, high-quality data and iterative training based on actual customer questions (13:46–15:03).
“You can't start… by pointing it at your entire database… We started slowly… giving it the first batches of insights… After two weeks [we learned] the 20 most common questions… and tuned those up with great answers.”
— Jacco van der Kooij (13:46)
- Improvements in Company Knowledge:
- Training the AI exposed gaps in their existing processes and material, leading to “WBD files”—a pristine, unified set of deep company insights (15:20–16:06).
- Beyond a Chatbot:
- Jack was a conversational learner, not a static bot, focused on creating value exchanges and accurately qualifying mutual fit, not just company benefit (16:26–17:33).
4. Buyer-Centric Design
- Value Exchange is Crucial:
- Early versions that “took” information with little value given back performed poorly; offering to share summaries, research, or slides greatly increased engagement and perceived value (18:03–19:15).
“Once we engaged it with experiences for the customer… you can see what does the customer value right from the front and how can we engage the customer with it.”
— Jacco van der Kooij (18:25)
5. Measuring ROI: Results & Financial Impact
- Initial Metrics:
- In month one:
- 831 real conversations (20:41)
- 1,200 pain points identified
- 70 qualified leads (8–9% conversion) (20:41–21:48)
- Only 8% email capture rate (21:48)
- In month one:
- Improvements through Iteration:
- Implementing richer visual aids and optimizing FAQs raised email capture from 8% to ~20%, and MQL conversion to ~36% (23:33–23:40).
“We see that our MQL conversation rate has, has gone up significantly to about 30-ish percent, 36% to be.”
— Jacco van der Kooij (23:33)
-
Cost/ROI Calculus:
- Human SDR: $60–$80K salary, often underutilized (19:43).
- AI SDR (“Jack”): Total cost ~$100K (including tools/startup), but with 24/7 performance, consistent CRM updates, and no recruitment/onboarding disruptions (24:45–26:05).
-
Human Impact:
- Freed up previous SDR (Ian) to build the company website, maximizing “human ROI” (32:05–33:02).
6. Notable Use Case: Multiplying Buyer Engagement
- Viral Internal Buy-in:
- A $200K deal advanced after one champion directed the entire buying committee—20+ team members—to independently consult Jack, accelerating consensus and reducing friction (27:35–28:41).
- Seamless Handoff:
- Customers sometimes skipped traditional discovery calls, having their expertise needs met by Jack.
“The customer said, no, we already had that call.”
— Jacco van der Kooij (28:41)
Notable Quotes & Insights
-
On Value Exchange:
“Build your AI experience for the buying experience, not for the selling experience.” (30:59)
-
On Training AI:
“Start narrow and expand. Iterate, iterate, iterate… The system can iterate very quickly.” (31:11)
-
On Process Clarity:
“An AI system doesn’t fix a bad process. It will make sure that it reveals it where your process is bad.” (31:38)
-
On Human-Centric AI Deployment:
“Let’s deploy the human beings where they can make the best benefit.” (32:52)
Practical Recommendations & Takeaways
Do’s and Don’ts for AI SDR Deployment (30:59–32:05)
- Do:
- Build for buyers, not just internal efficiency
- Think of the AI SDR as part of a system, not a single-function tool
- Start with a narrow, high-quality knowledge base
- Commit to continuous improvement and regular iteration
- Use feedback to identify and fix process or content weaknesses quickly
- Don’t:
- Don’t dump an entire knowledge base at once; curate and iterate
- Don’t expect AI to fix broken processes—use it to expose and address issues
Broader ROI Considerations
- Account for savings in recruitment, onboarding, and disruption
- Maximize human talent by redeploying skilled employees, not replacing them
Expert Advice
- For Execs Unsure About AI Replacing Human Engagement:
“Go to winningbydesign.com, Jack will pop up… at the end ask for the Spice summary… ask yourself, will 80% of your people do better or worse than what the AI just did?” (33:22) - For Early-Career Professionals:
“Open up a Gemini account, open up a Claude account, and open up a ChatGPT account… Build something… that you're good in, build it in AI and then pursue it.” (33:58)
Timestamps for Key Segments
- 04:01 — Decline in SDR effectiveness
- 06:45 — Global/human limitations in the SDR role
- 07:35 — Approach: System over tool
- 10:40 — Buyer-first mindset shift
- 13:46 — AI training pitfalls & successes
- 16:06 — Creating unified company knowledge
- 18:03 — Value exchange and engagement strategies
- 20:41 — First-month AI SDR results
- 23:33 — Improved conversion rates post-iteration
- 27:35 — Multi-stakeholder buyer use case
- 30:59 — Summary of do’s and don’ts
Conclusion
This episode provides a firsthand blueprint for deploying AI in direct customer-facing sales roles. Jacco’s experience at Winning by Design emphasizes designing for the customer (not just the seller), treating AI as a system, and using iterative, data-driven methods to quickly improve effectiveness and ROI. The practical stories and lessons highlight that AI, done right, elevates both buyer experience and business efficiency—while freeing up humans to deliver their unique value.
