Token
A token is the chunk of text an LLM processes - roughly four characters or three-quarters of a word. Models charge by tokens, cap context by tokens, and measure cost in tokens per million. Every piece of AI-assisted work - research, drafting, analysis - runs on tokens. When people say AI made content 'nearly free', tokens are the unit of that freedom.
Why It Matters
Tokens changed the economics of SEO work. The grunt work - drafting, summarising, data synthesis - now costs fractions of a cent per unit. That's why agencies selling content volume became obsolete: volume is cheap now. What tokens can't do is decide what's worth writing, whether it's any good, or whether it ranks. The judgment is the human part, and it's now the only expensive part.
Tokens also set the practical limits. Context windows cap how much of a site a model can read at once, which is why concise, structured, retrieval-friendly content wins over sprawling documents. And cost still matters at scale - token budgets are a real operating number for any serious AI workflow.
In Practice
Think of tokens as a budget in your SEO operations: what you spend them on is what you value. Spend them on research and analysis that inform judgment, not on generating volume nobody reads. Write content that respects context windows - the answer early, the fluff gone - because models and readers both reward it.
Understand the arithmetic when you evaluate AI tooling: per-million-token pricing is the unit price of your production line. It's the same argument as the agency overhead debate - the cost per unit of genuinely useful work is what matters, not the raw volume.
Common Mistakes
Measuring AI success in tokens generated instead of value produced. A million cheap tokens that say nothing are still worth nothing - just cheaper nothing.
Sources & Further Reading
Where to verify this yourself - Google's own documentation, industry reporting, and how we apply it at Underdog. Don't take our word for it; check the source.
Related Terms
Glossary
Large Language Model (LLM)
The AI technology behind ChatGPT, Claude, Gemini and the search engines that now read your content. Understanding it is the new SEO literacy.
Glossary
AI Content Generation
Using AI models to draft and scale content - a multiplier, not a substitute for judgment. The tool that made volume cheap and taste expensive.
Glossary
Generative Engine Optimisation (GEO)
The practice of getting your brand cited by AI answers instead of just ranked on Google. Optimising for being the source, not just the link.
Glossary
Thin Content
Pages with little useful content - Google devalues them and they drag your site down.
Know the Words.
Now See Them in Action.
Free teardown. No jargon. Just what's broken and how to fix it.
Get The Teardown