AI eats sales teams: Vercel slashes SDRs 90%, Google spends $2.4B

In H1 2024 AI firms captured half of all venture deals and 61 % of megadeal capital, even as total VC spending hit $204 billion, the third‑largest year on record. Late‑stage AI startups command double the valuations of non‑AI peers, forcing founders to rethink fundraising strategies and investors to re‑balance portfolios.

SaaStr’s latest AI stack shows agents merging rather than multiplying: the new AI VP of Finance runs inside the existing AI VP of Marketing. This consolidation solves a six‑figure collections shortfall and signals a shift toward deeper, multi‑capable agents, challenging the hype of hundreds of narrow bots.

Vercel’s new GTM engineering team built an AI lead‑qualification agent that replaced a ten‑person SDR squad with a single overseer, costing only $5,000 a year in compute and token usage. The move delivered a 32× ROI and freed the former SDRs for higher‑value, strategic work, proving AI can reshape sales pipelines at scale.

Lovable reached $75M ARR in just seven months by letting anyone type English and instantly create SaaS apps, echoing Shopify’s impact on e‑commerce. The piece probes whether its current pricing, largely usage‑based, can sustain such rapid growth as non‑technical founders scale. Insights reveal potential revenue cliffs and pricing tweaks needed for longevity.

A VP of Marketing swapped a 10,000‑lead MQL strategy for a 100‑account ABM focus, and the qualified pipeline surged. By moving BDRs under marketing and redefining metrics around pipeline instead of leads, the company gained faster, visible wins. The shift shows that tightly aligned account‑based GTM can replace noisy lead gen.

After OpenAI’s $3 bn takeover collapsed, Google struck a $2.4 bn licensing and reverse‑acquire deal, bringing Windsurf’s CEO Varun Mohan and key researchers into DeepMind. The move gives Google non‑exclusive access to Windsurf’s code‑generation tech and signals a new talent‑focused play in AI development.

Spencer Burleigh shows that businesses built on a single transaction, like his ADU venture Rent the Backyard, consume a finite market and quickly erode margins. As each sale depletes the TAM, the model collapses under rising costs and competition. Founders must design repeatable revenue or risk becoming a short‑lived gold rush.

Meta CEO Mark Zuckerberg warned staff that its AI‑agent projects have stalled, lagging behind the aggressive timelines the company set. The admission follows a wave of layoffs and reassignments aimed at fast‑tracking AI, but Zuckerberg says tangible gains won’t materialise for another three to six months.

Figma is growing revenue 46% YoY with accelerating growth, 82% margins, 139% NDR and 27% free‑cash‑flow margin, yet its stock trades around 6x forward revenue. Peers growing 30%+ command 15‑36x multiples. The gap stems from an IPO pop, AI‑fueled sector sell‑off, and misplaced fear that AI will replace design tools.

In H1 2026 SaaS firms multiplied their AI token spend fivefold, but most cannot trace any revenue boost, sparking a broad‑scale ROI crisis. The squeeze forces boardrooms to demand measurable returns and threatens future AI budgeting across B2B companies. The episode coincides with Anthropic lobbying to block Chinese open‑source models and Microsoft logging its worst month since 2000.

Instead of buying a SaaS product that encodes expert workflows, companies can now feed those playbooks to LLMs that execute them directly. This sidesteps the traditional subscription model, threatens SaaS revenue streams, and forces founders to rethink product strategy around AI‑driven orchestration.

Jersey Mike's S‑1 filing mentions "artificial intelligence" 22 times, even though the sandwich chain offers no AI product. The overuse shows how AI buzz has become a default marketing tick for IPOs, diluting genuine tech innovation signals for investors.

xAI released Grok 4, a model that now leads benchmark leaderboards, while AI rivals scramble to launch browsers aimed at dethroning Chrome. The surge signals a broader AI‑powered browser arms race that could reshape search and web traffic, putting pressure on Google’s monopoly.

Google now returns one visitor for every 18 pages it scrapes, while OpenAI and Anthropic scrape thousands of pages per visitor. As AI answers searches directly, traditional web traffic collapses, turning distribution, and the ability to surface content in AI prompts, into the only lasting competitive edge for SaaS firms.