Dots launched at DevDay on Sept 29 2026 as OpenAI's answer to Meta Muse — a persistent, GPT-6 Astra-powered agent with its own cloud computer. Here's the first-week practitioner read: the 5 marketing jobs I'd hand it, the 3 I wouldn't, and where it sits against the Claude/Gemini/Muse agents already in my stack.
Google shipped Gemini 3.8 Live on September 15 and Gemini 3.8 Flash TTS on September 23, 2026 — together, they rewrite the cost curve for marketing audio. A 10-minute Live voice call runs about $0.23; an hour of Flash TTS voiceover runs about $0.81. A week in production: 14 podcast clips, 6 ad voiceovers, 3 landing-page voice agents, and a brand-voice NotebookLM Audio Overview for $1.40 total. Five jobs the combo earns in a marketer's stack, three to keep elsewhere, and one routing principle that holds.
Five frontier models, five different superpowers. After running them side by side on real campaigns for six months, here's which one I open for which marketing job — and the brand-safety landmine that makes Grok-Imagine unusable for most brands.
ChatGPT, Perplexity, and Gemini Deep Research compared — plus a 4-layer market analysis framework, a 3-step competitor teardown, and real prompts I use in client work.
A practical walkthrough of using Google's Nano Banana image models for ad creatives, e-commerce visuals, and social media content, with detailed pitfall warnings from real production use.
Most marketing teams learn their CAC doubled on Tuesday by reading the Monday dashboard next week. The 90-day pipeline that catches it within hours: GA4 + Meta + Google Ads into a Sheet, nightly Gemini call with a 14-day baseline and a name-the-cause prompt, red/yellow severity to Slack. Includes the actual prompt, the n8n JSON, and a real Tuesday alert.
Most lead scoring projects die in week three. The 30-minute version that actually ships: 5 dimensions, a one-paragraph rubric, and one =AI() formula in Google Sheets powered by Gemini. Sales filters by green, you re-score weekly.
A 20-asset RSA starter pool (5 headline themes × 4 description angles) explodes to 1,000 strategically-curated combinations. Gemini does the combinatorial math and pre-flight validation, Google Ads Editor ships the bulk upload. A workflow that lifted one ecom account's CTR by 27%.
A real funnel that turned 2,400 candidate keywords into 87 ranked targets — 22 of them top-10 within four months. The five filters, the Gemini classification prompt with explicit JSON output, the funnel math (2,400 → 480 → 280 → 140 → 100 → 87), and why trusting KD score alone is the most common way to waste six months.
A one-page Google Sheets dashboard with 8 marketing KPIs and a Gemini =AI() cell that writes the week-over-week narrative for each metric. Built in 90 minutes, fits in a Slack screenshot, kills the 20-minute Monday meeting.
Export your RSA asset report, ask Gemini to grade every pin on 4 dimensions Google doesn't even measure, flag the bottom 30, then rewrite them with a structured prompt. The full audit takes 2.5 hours, saves about 6 versus doing it by hand, and the rewritten assets routinely lift impression volume and ad strength within 30 days.
Gemini Deep Research's 1M+ token context and multi-step browsing make it a different beast for technical SEO audits. Here's the 3-step workflow I use, the prompts that actually return useful output, and the limits you should know about.
