# T-shaped

> Data and AI applied to marketing, written for CMOs, brand leaders, and data leaders. Every article is researched, fact-checked against its sources, and approved by the author.

Scott Zakrajsek, Chief Data Officer at Power Digital, writes every article. Each one lists its sources with dates, carries the date its claims were last verified, and is approved by the author before it publishes. Articles are at /<slug>/, topic hubs at /topics/<category>/.

## Pages
- [Home](https://t-shaped.com/): who the author is, the latest articles, the seven topics
- [Articles](https://t-shaped.com/articles/): every article, newest first
- [Topics](https://t-shaped.com/topics/): the seven category hubs, each with a reading order
- [About](https://t-shaped.com/about/): the author
- [Method](https://t-shaped.com/method/): how articles are researched, fact-checked, and approved
- [RSS](https://t-shaped.com/rss.xml): published articles as a feed

## Topics
- [Measurement](https://t-shaped.com/topics/measurement/): How a marketing team learns what worked, and why the evidence hierarchy matters more once agents optimize whatever they are given.
- [Data Foundations](https://t-shaped.com/topics/data-foundations/): The data work that decides whether an agent can be trusted: models, semantics, pipelines, identity, quality, consent, and access.
- [AI and Agents](https://t-shaped.com/topics/ai-and-agents/): How a marketing team builds, equips, governs, and connects AI agents, from the first pilot to a platform many brands share.
- [Media and Creative](https://t-shaped.com/topics/media-and-creative/): What agents can do inside ad accounts and creative pipelines today, and how review, planning, and testing change at ten times the volume.
- [Team and Operating Model](https://t-shaped.com/topics/team-and-operating-model/): Who builds, runs, and owns marketing agents, and how a team develops judgment when the practice tasks are automated.
- [Change Management](https://t-shaped.com/topics/change-management/): How a team moves from analysis to supervised action to autonomy, what evidence earns each step, and what volume does to approvers.
- [Strategy](https://t-shaped.com/topics/strategy/): Where AI-run marketing is heading: a maturity model to place a brand on, how brands get recommended by AI assistants, and what happens to agencies.

## Articles
- [The three roles it takes to build a marketing agent](https://t-shaped.com/who-builds-marketing-agents/): Your best practitioner should be your agent builder, because a new hire learns the playbook second-hand and can't always tell when the marketing agent is wrong. Published 2026-10-18, last verified 2026-10-18.

## Glossary
- [agent skill](https://t-shaped.com/glossary/agent-skill/): An agent skill is a written procedure the agent loads when a task calls for it, with the steps, the exceptions, and the checks spelled out.
- [data catalog](https://t-shaped.com/glossary/data-catalog/): A data catalog is a searchable record of what each data set holds and who owns it.
- [marketing agent](https://t-shaped.com/glossary/marketing-agent/): A marketing agent is AI software that carries out a marketing task in steps, following written instructions.
- [semantic layer](https://t-shaped.com/glossary/semantic-layer/): A semantic layer is one shared set of metric definitions that every tool reads.