Morning Brew's founder turns Claude into a content machine, ICONIQ's 2026 benchmarks show AI-sourced pipeline setting sales quotas, and open-weight models keep squeezing AI margins.
AI in GTM. Alex Lieberman of Morning Brew and Tenex walks through his Content Machine: an AI Oracle that scans Slack, Notion, Gmail, and social feeds to surface 15 ranked content ideas a day, a six-persona AI interview panel that extracts real specifics via voice-to-text, and a Markdown voice file the model drafts against. A reinforcement loop compares drafts to published posts and banks lessons over time, and employee-generated content drove 40 percent of inbound leads in one quarter at his prior company. (Lenny's Newsletter)
Why it matters: An idea engine plus a voice file plus a feedback loop beats ad hoc posting, and it reframes AI slop as a garbage-in problem, not a model problem.
SaaS benchmarks. ICONIQ's 2026 GTM benchmark report, covering more than 150 B2B and AI software companies, puts top-quartile AE quotas at $750K for SMB, $1.35M for mid-market, and $2.25M for enterprise, with elite reps hitting 85 to 90 percent attainment versus the traditional 65 to 75 percent baseline. Comp is shifting toward expansion revenue, with NRR-tied comp up from 18 percent to 23 percent, and reps now own more of the lifecycle: 65 percent own cross-sell, 55 percent own upsell, and 37 percent own renewals. (SaaStr)
Why it matters: The report's core claim is that AI-sourced pipeline is the real quota-setter, delivering a 10 to 11 point lift in lead-to-MQL conversion. Quotas are only rising because AI pipeline is doing the heavy lifting.
RevOps. The Science of Scaling surveyed more than 300 sales leaders: 23 percent name poor list quality and targeting as the single biggest limiter of cold-call ROI, ahead of rep reluctance and follow-up failure. Only 18 percent say sales and marketing are fully aligned on ICP definition, with 38 percent calling their alignment merely general, and the gap shows up worst at the ICP edge cases reps actually work. (The Science of Scaling)
Why it matters: A clean, quantified case for RevOps owning ICP and list-quality alignment before any AI SDR or outbound tooling gets layered on top. Better tools cannot fix a bad list.
GTM strategy. Ben Thompson argues that open-weight Chinese models like Kimi K3 are reintroducing zero-marginal-cost economics into AI, undermining the assumption that frontier model access is expensive and defensible. He separates COGS from R&D to show that free open weights still carry real infrastructure costs, but the competitive floor for model quality keeps rising while pricing power for closed models erodes. (Stratechery)
Why it matters: For any GTM leader buying or building AI features into the stack, this is an early signal that model costs will keep falling and that vendor lock-in on model choice is a weaker moat than it looks. Useful context for build-versus-buy conversations.
Sales leadership. a16z lays out the seven roles hardware startups most often get wrong, given how unforgiving physical-product business models and multi-stakeholder sales cycles are compared to pure software. (a16z)
Why it matters: Even for a software-only operator, the lesson generalizes: sequence GTM hires to match the buying complexity of the product rather than copying a generic SaaS org chart.
The GTM × AI brief, every morning. Free.