AI Moves Fast, Strategy Moves First
Developing AI capabilities before the data exists to prove it
Starting this week, we’re pivoting from weekly deep dives to shorter, bite-sized vignettes that break down the strategic moves behind the headlines.
Every Friday, I break down the decisive choices of leading companies and show you how to start seeing those patterns, learning from them, and applying them to your own products.
This week’s theme: Developing first-mover AI capabilities.
Time to read: 4 minutes
1. Adobe Bets $1.9B on Generative Engine Optimization
Adobe paid a hefty 78% premium to acquire Semrush for $1.9 billion in November, betting that AI LLMs (ChatGPT, Claude) will increasingly displace Google as the web’s front door.
When the acquisition closes, Adobe plans to integrate Semrush with both its Adobe Experience Manager (AEM) and Brand Concierge products, enabling marketers to optimize for both Google’s algorithms (SEO) and AI discoverability (GEO) in a single workflow. That 78% premium signals the existential urgency it’s feeling to keep its products relevant.
Can’t/Won’t: Competing marketing platforms will be hard-pressed to rapidly build and deliver unified SEO/GEO without making similar acquisitions at this scale. It’s unlikely creative platforms like Figma or Lovable will embrace discoverability optimization as a core platform strategy.
Potential Risk/”What Would Have to Be True?”: Acquisitions are always risky. Always. The odds are against Adobe’s ability to integrate and realize the value it’s invested. Counterintuitively, Adobe’s better acquisition play would be to understand what value it can bring to Semrush, not the other way around.
2. Target Treats ChatGPT as a Sales Channel
Target launched a full, end-to-end ChatGPT checkout in November 2025, allowing customers to complete purchases from discovery through payment directly within the AI chat interface.
All purchase intent starts with a question, but when customers complete checkout from inside ChatGPT, AI moves from passive billboard to active storefront, opening another potential revenue stream. It remains to be seen whether “conversational commerce” becomes a viable retail channel with meaningful revenue potential or just a minor, secondary sub-channel under web and mobile.
Can’t/Won’t: Amazon can’t build on OpenAI’s platform. Traditional retailers won’t redesign their checkout flows for conversational interfaces. Target’s moving first while others wait for proof before committing to develop the necessary commerce capabilities.
3. Notion Unlocks Private Slack Channels for Organizational Context
Notion released version 3.1 in November, upgrading its Slack connector to access private channels and DMs while respecting all existing permission levels.
The unlock: Most workplace AI tools are blind to private conversations. Getting into territory previously owned by Otter, Notion is giving enterprise employees access to broader AI context while directly addressing enterprise security data exposure concerns.
Can’t/Won’t: Slack AI won’t natively surface insights beyond parent Salesforce. By building permission-aware private channel access, Notion is positioning itself as an alternative knowledge layer that synthesizes all relevant workplace communications, going one step deeper than public artifacts.
4. DocuSign Opens Agreement Intelligence to AI Agents
DocuSign released its Intelligent Agreement Management (IAM) platform via Anthropic’s Model Context Protocol (MCP) in October 2025, enabling Claude, Copilot, and other agents to access agreement workflows directly via natural language.
As AI agents proliferate, whoever enables access to authoritative business data (agreements, CRM, HR) captures disproportionate value. Extending its deep agreement expertise into agentic AI, DocuSign seeks to own the source-of-truth workflows involving contract context.
Can’t/Won’t: Competitors without industry-standard MCP infrastructure can’t join the agentic ecosystem. Legal teams will be unlikely to trust custom integrations. DocuSign is capturing an early advantage over competing agreement providers before the contract trust market becomes fragmented.
5. Vijil Raises $17M to Solve AI Agent Trust
Vijil, named a Gartner Cool Vendor in Agentic AI Security, raised $17 million to accelerate its agentic AI trust infrastructure platform.
The bottleneck: enterprises lack the confidence that the agents they’ve piloted will behave reliably in production. Vijil’s platform covers the full agent lifecycle, from build, test, deploy, govern, and continuously improve via reinforcement learning using live production data. SmartRecruiters cut its time to ship agents from 6 months to 6 weeks using Vijil.
Can’t/Won’t: Neither observability vendors (Datadog, etc.) nor generic security tools have a shot at building agent-specific lifecycle management or continuous reinforcement learning. Vijil occupies a tiny but essential niche to roll out and ensure AI agents perform reliably in mission-critical environments.
📊 THIS WEEK’S SIGNAL
Five companies are making similar strategic choices to move early in AI capabilities now, before waiting for enough data to exist to tell them they needed to.
By that time, it would already be too late.
Adobe’s betting that GEO will become as critical as SEO. Target’s betting that conversational commerce is real and that it can capture significant value through the channel. Notion’s betting on grabbing a bigger slice of the enterprise market with comprehensive context while respecting permissions. DocuSign is betting on being first to own agreement signatures through agentic ecosystems. And Vijil’s betting that everyone, from startups to enterprises, will increasingly reimagine every part of their business workflows around AI capabilities with AI agents at their core.
These moves redefine what “table-stakes” capabilities look like in the age of agentic AI.
For Your Product Strategy
What new channels could AI be making available to your product?
In what ways could you use agentic AI and MCP to completely reimagine legacy human processes, rather than just accelerate them?
What features or pilots could you be running today that could scale to become infrastructure tomorrow?
What tradeoffs is your industry accepting that you could attack directly? (Not either/or, but and?)
Who’s making moves that look improbable, too early, or too expensive until you see what they see?
That’s it for this week. Next Friday: another set of strategic moves and the patterns behind them you need to understand.
— Mike
"shorter, bite-sized vignettes that break down the strategic moves" 💙
The SEO vs GEO bets are really interesting to follow. I think Google has a good response so far to the traffic migration threat with AI overviews and the great capabilities on Gemini.