New episode of The Agents is out: #004
What Amelia and Jason covered:
→ AI PR is broken
The terrible AI-generated pitches flooding our inboxes prove most B2B founders still don't know how to deploy agents that actually work.
→ Vibe Coding at SaaStr AI Annual
10+ live Replit classes, the Vibe Lab, main stage builds with Replit and Vercel, Amelia building 10K live in 45 minutes. Deepest vibe coding track at any B2B event this year. Come Tuesday May 12 if that's all you can spare.
→ Six months of 10K in production
Three hard lessons.** Not a launch story. An operating story. What we've actually learned running an AI VP of Marketing day to day.
→ The API Report Card. 106 B2B APIs graded, 1,600+ runs in week one
Stripe gets the onluy and is where agent budget is going. Outreach, ZoomInfo, Marketo are where it's leaving. The bottom of the list = the budget categories most under threat over the next 24 months.
→ SDR Sprawl
Do we actually need 4+ AI SDRs?
Honest answer: not sure yet.
→ Notion vs Salesforce as system of recor
The strategic call every B2B founder is going to have to make.
Sales reps start their day with a prioritized brief from Claude, pulling context from every tool. Proposals are drafted, policy-checked, and sent in minutes, transforming deal cycles. #SalesTech#AIinBusiness
Sales teams struggle to keep up with product updates and competitor info. Claude's new AI features generate dynamic battle cards and custom collateral instantly, empowering AEs to win deals and deliver exceptional customer experiences on every call. #AI#SalesTech
The thrill of a closed deal is addictive, but the admin behind it can be a nightmare. Ensuring smooth onboarding, invoicing, and org mapping is crucial.
5 Things Anthropic Did To Reboot Their GTM Motion in 2026:
1/ Launched enterprise self-serve
MVP in January, production in February. Built the funnel with Intercom Fin guiding buyers through real enterprise plans with real ACV and terms of service. 54% of new enterprise logos in 2026 came through this funnel.
2/ Threaded Claude through their existing 6-tool stack instead of replacing anything
Clay for enrichment. LeanData for routing. Salesforce as system of record. Gong for call coaching. Ironclad for contracts. Slack for coordination. Claude became the connective tissue between them.
3/ Made Slack the front door for every support function
Deal desk, legal, RevOps, billing, compliance. Slack ticket in, Claude triages, Jira ticket out. Resolves inline if it matches precedent. Escalates with full context if it doesn’t.
4/ Codified their best reps as 5 Skills every new hire gets day one
Morning brief, call prep, customer follow-up, competitive intel, create-an-asset. AEs summon them with a / shortcut. Replaces the 6-week onboarding curve with a sales plug-in.
5/ Killed quotas
Switched to “shadow targets” because model intelligence drives adoption faster than any quota cycle can keep up with. Kelly Loftus, Head of Startup Sales: “It’s really hard to predict what is happening. The adoption is fast.”
54% of Anthropic's new enterprise logos in 2026 came through self-serve.
Self-serve enterprise. Real ACV. Real terms of service. No AE in the loop.
Anthropic's Head of Industries Eleanor Dorfman walked through at SaaStr AI 2026 last week how they rebuilt the entire sales org in 30 days after Claude Opus 4.6 broke their demand curve in December.
👉The constraint: couldn't 3x or 4x the sales team fast enough without lowering the recruiting bar.
The thesis: don't buy a new stack. Thread Claude through the one you already have.
What they kept:
1⃣ Clay for enrichment
2⃣LeanData for routing
3⃣ @salesforce as system of record
4⃣@Gong_io for call coaching
5⃣Ironclad for contracts
6⃣@slackhq for everything else
What they added: Claude as the connective tissue between all six.
The four moves:
1/ Killed the PLG vs SLG orthodoxy. Launched enterprise self-serve in January. Intercom Fin guides the buyer through the journey. Now 54% of new enterprise logos.
2/ Threaded Claude through the existing stack. Every AE starts the day with a "morning brief" Skill that pulls context from Gmail, Gong, Slack, Salesforce, @intercom, Greenhouse.
3/ Made Slack the front door for every support function. Slack ticket in, Jira ticket out. Claude triages and resolves inline if it matches precedent. Escalates with full context if not.
4/ Codified what the best reps do as Skills. Every new rep gets a sales plug-in with 5 Skills: morning brief, call prep, customer follow-up, competitive intel, create-an-asset.
Anthropic didn't replace anything. They invested in the stack they already had and let Claude be the seam between everything.
Most companies will spend 2026 evaluating AI-native sales platforms. But Anthropic did it with its current stack + Claude.
Almost none of it required new software.
Top sales reps no longer waste time prepping for calls. AI provides a succinct one-pager on stakeholders, historical context, and ideal outcomes, ensuring a tailored and personalized customer experience every time. #SalesTech#AICoPilot
Agents are increasingly part of the employee population pool.
Ghazi Masood, Replit's CRO, on the next big management question: how do you do performance reviews, promotions, and team structure when agents work alongside humans?
It's already happening.
Live at SaaStr.
10 things to steal from Anthropic’s AI-native sales playbook, per Eleanor Dorfman at SaaSt 2026:
1/ Anthropic just hit $30B in annualized revenue with ~5,000 employees. Google needed 32,000 to reach $30B. Salesforce needed 79,000. The most efficient B2B company ever built.
2/ They run on the same stack you do. Salesforce. Gong. Slack. Ironclad. Clay. LeanData. They didn’t replace anything. They threaded Claude through the seams.
3/ 18 quota-carrying enterprise AEs serving 300,000+ business customers. 16,667 customers per rep. 500+ now spend $1M+ annually, up from ~12 two years ago.
4/ 54% of new enterprise logos in 2026 came through self-serve. Real ACV. Real terms of service. No AE-led journey. Intercom Fin guides the buyer end-to-end.
5/ They killed the PLG-vs-SLG orthodoxy in January. Enterprise self-serve isn’t a consolation prize for buyers who don’t deserve a human. It’s where most logos come from now.
6/ Slack became the front door for every support function. Ticket in, Claude triages, Jira ticket out. Deal desk, legal, RevOps, billing all run through one entry point.
7/ Codified the best reps as 5 Skills every new hire gets on day 1: morning brief, call prep, customer follow-up, competitive intel, create-an-asset. No more 6-week onboarding curve.
8/ ICONIQ 2026: high AI adopters now generate 2x net new revenue per GTM FTE. The biggest gap isn’t in Sales (+20%). It’s in Post-Sales (+83%). AI CSMs are the most underleveraged role.
9/ SDRs are already at 71% daily AI adoption. RevOps jumped from 34% to 54% in one year. Median company is planning 0% RevOps headcount growth in 2026.
10/ Anthropic doesn’t even use quotas anymore. “We have shadow targets. Adoption is driven by model intelligence, which you cannot predict.” If the leading AI lab can’t tie revenue to seat capacity, your customers can’t either.
The B2B sales stack didn’t get disrupted. It got compressed.
54% of Anthropic's new enterprise logos in 2026 came through self-serve.
Self-serve enterprise. Real ACV. Real terms of service. No AE in the loop.
Anthropic's Head of Industries Eleanor Dorfman walked through at SaaStr AI 2026 last week how they rebuilt the entire sales org in 30 days after Claude Opus 4.6 broke their demand curve in December.
👉The constraint: couldn't 3x or 4x the sales team fast enough without lowering the recruiting bar.
The thesis: don't buy a new stack. Thread Claude through the one you already have.
What they kept:
1⃣ Clay for enrichment
2⃣LeanData for routing
3⃣ @salesforce as system of record
4⃣@Gong_io for call coaching
5⃣Ironclad for contracts
6⃣@slackhq for everything else
What they added: Claude as the connective tissue between all six.
The four moves:
1/ Killed the PLG vs SLG orthodoxy. Launched enterprise self-serve in January. Intercom Fin guides the buyer through the journey. Now 54% of new enterprise logos.
2/ Threaded Claude through the existing stack. Every AE starts the day with a "morning brief" Skill that pulls context from Gmail, Gong, Slack, Salesforce, @intercom, Greenhouse.
3/ Made Slack the front door for every support function. Slack ticket in, Jira ticket out. Claude triages and resolves inline if it matches precedent. Escalates with full context if not.
4/ Codified what the best reps do as Skills. Every new rep gets a sales plug-in with 5 Skills: morning brief, call prep, customer follow-up, competitive intel, create-an-asset.
Anthropic didn't replace anything. They invested in the stack they already had and let Claude be the seam between everything.
Most companies will spend 2026 evaluating AI-native sales platforms. But Anthropic did it with its current stack + Claude.
Almost none of it required new software.
Forecasts are largely run by AI (Claude) and then reviewed by managers. This approach ensures Salesforce data is current, account plans are aligned, and consumption is understood.
12 things from @jasonlk’s closing AMA at SaaStr AI Annual 2026:
1.Your sales team has to be product experts now. Schmoozing on Zoom is dead. If your rep can’t answer a technical question the way Kody at Replit can, I’ll bypass them and buy the competitor.
2.Marc Benioff told me on 20VC last summer, unguarded: “I wish every Salesforce customer had an FDE and didn’t have to pay me until they were deployed.” That’s the new bar. The agent has to work on Day 30, not Year 3.
3.Stop aiming for 80% of your best human. In some slivers, agents are already at 120%. Qualified set up 682 strong meetings for us in the run-up to Annual. Better than any BDR I’ve ever worked with.
4.The right frame isn’t “is there slop.” It’s “what’s the slop-to-value ratio, and which way is it moving.” For our products, value is 50x the slop. And the slope is getting steeper every week.
5.2020-2021 taught us nothing. It was an aberration. People bought 20-year-old software like it was going out of fashion. Then it ended. Shopify said it was over in summer 2020. Throw that playbook out.
6.B-tier engineers are out of work. S-tier engineers have never been more in demand. Both true. The arms race is at the top. My son at Penn State has labs chasing him. His CS classmates? Zero offers.
7.There’s no land grab in AI. There’s an inertia grab. Customers are explicitly signing 1-year deals and re-evaluating in 10 months. Once we’re trained on a platform, the next vendor has to be 2x better, not 10% better.
8.Venture in 2026: nothing matters but growth. Not gross margins. Not NRR. Not multi-year contracts. Nothing. People are writing checks into Anthropic at $950B in 48 hours without talking to anyone.
9.Going from $10m-to-20m ARR in a year used to be a slam-dunk fundraise. Today it gets you 100 beatings. The bar is 2-to-500 like Replit. The high end has no ceiling, so the middle has no oxygen.
10.The worst buying experience in B2B still exists: clicking “Contact Sales” and getting on a call with a kid who can’t answer your questions. AI can do 110% of that role today. Kill it tomorrow.
https://t.co/9Hf2e52lyN can’t lecture your leadership into AI fluency. You can only show them. Bring them in a room. Show them the single most magical thing your team does with AI. Live. If their jaws don’t drop, you’ve got a deeper problem.
https://t.co/8lq7SU4XcF year ago we had 1 pseudo-agent. Today we have 21, with 3 humans and $1M+ in AI-driven revenue. Next year? I genuinely don’t know. But you can do everything Amelia and I do. Probably better.
https://t.co/gGvyoWh8gp
AI is now a VP of Marketing, but it's not a 'set it and forget it' solution. Unexpected challenges and constant maintenance are part of the deal. Plus, this AI is getting increasingly autonomous in managing campaigns. #AIMarketing#TechEvolution
Bizzabo's API transformed guest pass creation with a micro-app, showcasing how APIs democratize tech. The speaker's view on API potential evolved through this project. #APIDevelopment#TechInnovation#EventTech
An AI grading system is evaluating companies and products for AI agent compatibility. The focus is shifting from traditional features to essential API access, reflecting AI's evolving needs. #AIEvaluation#FutureOfTech
Introducing the AI Agent API Report Card! See how APIs stack up for building AI-powered products. Stripe shines as a top performer, proving its mettle in the AI era. #AIAgent#API#Tech