Aug 28, 2026 · 53m · saastr
Who Owns Your Data Now? Agents vs. Systems of Record on The Agents #013
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
Jason Lemkin and Amelia Ibarra explore how autonomous AI agents are disrupting legacy SaaS systems of record, illustrating this shift through SaaStr's live implementation of headless CRM architecture, automated renewal presentations, and multi-agent data enrichment workflows.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Jason holds 44.6% of the talking time here. How this is scored →
Jason as informed peer 6.3 Guest teaching 5.6 Guest disagreement 0.0 Jason pushing back 0.3
speaking balance: gold is Jason, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 47:20 Amelia recounting friction with Clay pricing and Replit connector failures
In an otherwise collaborative episode, Amelia describes her frustration when Clay changed pricing overnight and broke internal agent capabilities, leading to tool fights.
Hardest push from Jason ▶ 20:00 Jason grilling why existing vendor agents were bypassed for renewalsJason interrupts Amelia to drill into why proven tools like Agentforce, Monaco, and Artisan could not be used for renewal decks instead of custom builds.
Biggest teaching moment ▶ 48:20 Amelia educating Jason on multi-layered Claude browser Co-work enrichmentAmelia explains how she used Claude Co-work combined with ZoomInfo nested inside Clay to bypass slow MCP connectors and build automated champion tracking.
Jason holds their own ▶ 11:30 Jason breaking down the macro economics of agent data generation vs CRM API pricingJason demonstrates domain expertise by analyzing how legacy SaaS per-seat and API overage models fail when agentic workloads 100x customer data volume.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Jason as informed peer | Guest teaching | Guest disagreement | Jason pushing back | Why |
|---|---|---|---|---|---|---|
| The Battle Over Systems of Record: ServiceTitan vs. Podium | 7 | 1 | 0 | 0 | Jason sets up the core theme by analyzing the battle between ServiceTitan and Podium, detailing ServiceTitan's $1.5B ARR scale and explaining the structural threat agents pose to systems of record. Amelia validates the thesis by introducing the real-world storage cost implications. | |
| Sponsor Break: Northwest Registered Agent and SaaStr AI Annual 2027 | 7 | 5 | 0 | 1 | Amelia reveals that Claude flagged SaaStr's 40GB data footprint in Salesforce as overage territory, prompting a deep dive into database costs. Jason demonstrates substantial operational expertise, contrasting Postgres/Neon pricing with legacy CRM database pricing and criticizing toxic SaaS API overage pricing. | |
| Inside SaaStr's Headless Salesforce Stack and the Renewal Agent | 5 | 7 | 0 | 1 | Amelia screenshares and walks Jason through the 10K headless dashboard, detailing six core integrations including Agentforce, Qualified, Momentum, and custom quote-to-cash apps. Jason probes into why existing third-party sales agents could not handle the renewal workflow. | |
| Building Custom Collateral and Democratizing Renewals via Gamma | 6 | 7 | 0 | 0 | Amelia explains how her renewal agent pulls multi-channel proprietary data across WordPress, podcasts, social APIs, and Gmail to feed Gamma's deck generator. Jason articulates how this democratizes custom tier-one collateral for lower-tier Silver sponsors. | |
| Walkthrough of the Replit Deck and the Collaborative Narrative Step | 6 | 7 | 0 | 0 | Amelia walks through the customized Replit deck, demonstrating how she trained the agent to prioritize high-impact metrics and introduced a collaborative narrative step before generation. Jason analyzes why current AI SDRs fail due to lack of historical context and reasoning. | |
| SaaStr Connect, Hyper-Context Outbound, and Inbound Proposal Workflows | 7 | 5 | 0 | 0 | Jason details his work building SaaStr Connect with deep-reasoning LinkedIn profile analysis to generate hyper-contextual outbound openers. Amelia complements this by describing how 10K suggested flipping the deck generator to instantly produce custom inbound proposals. | |
| Agent-Driven Data Enrichment with Clay, Claude, and ZoomInfo | 6 | 7 | 0 | 0 | Amelia details how Claude bypassed slow native MCP connectors by using browser Co-work to orchestrate ZoomInfo inside Clay for champion tracking. Jason synthesizes the strategic takeaway: vendor discovery in the agentic era depends on API agent-readiness rather than SEO/AEO ranking. |
Statements from this episode (15)
Lemkin: SaaStr migrated off two of its three primary systems of record
“In fact, until recently at Sastr, we basically had two and a half, three systems of record. We had Salesforce, we had Marketo, and we had Notion. We left two.”
Lemkin: ServiceTitan terminated Podium integration impacting 1,000 shared customers
“That worked great for years until Service Titan shut them off the other day, shut them off after all these years with a thousand joint customers.”
Lemkin: Podium generated over $100M in AI agent revenue
“When Podium went, became agentic about a year half ago and crossed a hundred million of agent revenue, the agents for the same customers, for the same HVAC guys and other trades, the agent all of a sudden could do a lot more than get a lead.”
Lemkin: Customer AI agents will eliminate the need for underlying CRMs
“In CX and support if I'm talking to an agent on someone's website as a potential customer and that agent keeps me forever, I might not need, there may be no need for the CRM or CX system or SFA that undermined it. Used to underlie that interaction because the …”
Lemkin: Raising API prices to offset lost seats will accelerate CRM churn
“I think it is a self-inflicted wound for a lot of systems of record and SaaS centers as they jack up the prices because seats are going down and agents are going up and they jack up the price for access. It's incense us finally to put more and more of our data…”
Lemkin: Headless architecture lets AI agents pull data across multiple CRMs
“The thing about headless is it also makes it much easier to bifurcate our data and put some into outside of the system of record and some in because the agent, the headless version of your CRM or SFA or system of record, well, it has no problem up to a point p…”
Ibarra: SaaStr runs its revenue cycle through headless Salesforce
“It shows you what he's hooked up to headless within Salesforce too, because there's multiple things running into this. So he's got class. I call it classic Salesforce for sales, right? So Salesforce CRM that our sales team has always used for like 10 years. He…”
Ibarra: SaaStr built an AI renewal agent using Salesforce and Gamma
“We created a mini renewal agent. So let me scroll down. I'll show you guys this. So we've got this mini renewal agent now going where, you know, we had our AI day, I think last month or something. And previously it'd be a manual thing where we have a digital e…”
Lemkin: AI slide maker Gamma has reached $150M in revenue
“This is all Gamma does at a hundred fifty million revenue.”
Ibarra: AI renewal workflows take half a day to build
“This is one of those things where in the age of AI, ah, this took me an extra, let's say, half a day with 10 K to build. Like, you gotta, sure, you might have to set up these sources the first time, but then you can output something like this for literally eve…”
Lemkin: AI SDR emails beat 99% of humans but lack deep context
“The artisan Monaco agent force emails. They're all better than a human, which is all that really matters today as we record this, right? They're better than 99% of humans. But they're not great like that RepliDeck. None of them are great.”
Lemkin: SaaStr banned small sub-$20k sponsors in 2022 to avoid stress
“For a while, when things were crazy, like in 20, 22, we like banned small sponsors because we didn't want the stress. We just didn't, we just like you, if that's so much, if 20,000 dollars is so much money to you, like you're betting your entire startup. Come …”
Ibarra: Companies must automate champion tracking because manual work gets neglected
“You should all do this champion tracking thing. I think not. I think so many people forget to do it, right? It's like, oh yeah, duh, makes sense. And then you forget to do it because it's not automated or whatever. It's too much work. Yeah, it's too much.”