Signal Tracking: Staying Ahead Without Drowning in News
One of the quieter parts of my AI practice is signal tracking: monitoring developments in AI and EdTech and connecting them to active project decisions. Not reading news for the sake of staying informed. Reading with intent, tagging what matters, and acting on it.
How it works
I built a lightweight journal system. Each entry captures an article or development with tags for which projects it’s relevant to and a connection statement explaining why it matters. A Chrome extension handles quick capture. Claude Code handles entry creation through a /capture command.
Over 30 entries so far, tagged across four active projects.
Three signals that shaped real decisions
OECD AI Skills Gap report: Policy-level evidence of a growing workforce divide in AI readiness. This validated the target population for our entrepreneurship platform. If the gap is widening, the need for accessible AI-enhanced tools is growing with it.
LinkedIn data on rising entrepreneurship: A surge in historically underserved populations pursuing startups. Direct validation for our product’s target users. If more people from this population are starting businesses, demand for peer accountability and structured support grows.
Harvard + Perplexity research on AI agent usage: 57% of real agent usage is cognitive/learning work, not task automation. This validated the AI coaching model. If people naturally gravitate to agents for learning, an intentionally designed coaching agent has stronger product-market fit than a task-completion tool.
Why this matters
Each of these articles could have been interesting reading and nothing more. The practice of tagging them to active projects and writing a connection statement is what turns passive consumption into strategic input.
I share relevant findings with my team to spark conversation about how AI is evolving and how it might impact our work. This has become a regular part of how I contribute beyond my direct project scope, connecting dots between external developments and internal decisions.
The discipline
The system only works if you do it consistently. A weekly cadence is enough. The goal isn’t comprehensive coverage. It’s staying oriented so that when a decision point arrives, you’ve already been thinking about the relevant landscape.