Metadata is missing or stale
Assets pile up across systems over years and the tagging never keeps up. Without consistent metadata, no tool can tell you what a piece of content is actually about.
Wave · Content Intelligence
Your content library is flying blind. Wave gives it eyes.
Most teams have hundreds of assets spread across systems with metadata that is inconsistent or missing. When a rep asks what a prospect has consumed, the honest answer is usually no idea. Wave crawls every content URL, tags it against your taxonomy with an LLM, and joins that to person-level engagement.
Quick answer
Content Intelligence is a Wave feature that crawls every content URL in your connected systems, uses an LLM to extract structured metadata against your own controlled taxonomy of topics, funnel stage, persona, and format, and joins that metadata to person-level engagement. The result is a queryable answer to what a specific buyer has actually consumed and what they should be served next.
Key capabilities
- Crawls every content URL in connected systems
- LLM metadata extraction against your taxonomy
- Topic, funnel stage, persona, and format tags
- Find related content by meaning, not just keyword
- Person-level consumption history, queryable
- Auto-crawl on new content, opt-in per tenant
- Tags exportable back to your CRM
Last updated July 2026
The problem
You cannot personalize what you cannot read.
Years of content production leaves most teams with a library that is large, scattered, and inconsistently tagged. The metadata needed to act on it either never existed or has gone stale, so consumption history is just a list of clicked URLs.
Clicks are not consumption
Your stack can show a click on a URL, but not what that content covered. Topic, funnel stage, and persona are exactly the dimensions you need and exactly what is missing.
Tagging by hand does not scale
Auditing and tagging a content library consistently is what teams hire contractors to do. It is slow, it is expensive, and it is stale again the moment new content ships.
How Wave does it
Crawl every asset, extract metadata, join it to people.
Wave turns your content library into structured, queryable data, then connects each contact's engagement to what the content was actually about.
- 01
Crawl every permalink
Wave enumerates and crawls every content URL surfaced in your connected systems. HubSpot and direct website crawling are live today, with more sources on the roadmap.
- 02
Extract against your taxonomy
An LLM extracts structured metadata against your own schema, so every asset is tagged for topic, funnel stage, persona, and format on the taxonomy you control, with guaranteed structured output.
- 03
Store it queryable, with embeddings
Extracted tags are stored as structured, queryable data, so you can find related content by meaning, not just by exact keyword.
- 04
Join to person-level engagement
Every contact's engagement events reference the crawled asset and its tags, so consumption history is queryable per person, per topic, and per stage.
Where it fits
The foundation under the rest of Wave.
Content Intelligence makes your library machine-readable, which is what every downstream recommendation depends on. Buying group next-best-content and next-up suggestions only work once Wave knows what content exists and what it covers.
Your sources
Wave
Tagged library
Wave reads from your existing systems through native connectors and runs extraction per tenant, so your content and engagement data never trains on anyone else's. Every extracted tag lives in Wave's database and is exportable back to your CRM. New content detected after a connector sync can trigger automatic crawl and extraction, opt-in per tenant, so the metadata stays current as you publish.
Works alongside Buying Groups, Channel Affinity, The full Wave platform.
Why Wave is different
Automatic, consistent, and joined to the buyer.
The difference is that Wave does at machine speed what teams used to do by hand, and then connects the result to the person. Tagging is not the deliverable. Knowing what each buyer consumed is.
Most tools
Wave
Content metadata is maintained by hand and drifts out of date.
Wave extracts metadata automatically against your taxonomy and keeps it current as new content ships.
Engagement reporting shows a click on a URL with no subject.
Wave knows what the content was about, so a click becomes a topic, a stage, and a persona.
Content and contacts live in separate tools that never join.
Wave joins every engagement event to the asset's tags, so you can answer what a buyer consumed per person.
Your data trains a shared, cross-customer model.
Wave extracts per tenant. Your content and engagement data never enters a shared model.
FAQ
Questions buyers ask about Content Intelligence.
What is Content Intelligence in Wave?
Content Intelligence is a Wave feature that crawls every content URL in your connected systems, uses an LLM to extract structured metadata against your own taxonomy, and joins that metadata to person-level engagement. It lets you answer what a specific buyer has consumed and what they should see next, instead of staring at a list of clicked URLs.
Which systems can Wave crawl content from?
Wave reads from HubSpot and crawls the content URLs surfaced there, and it can also crawl your website directly from its sitemap. More sources are on the roadmap. Wave reads on its own side, so you do not need to change your content systems for the crawl to run.
How does the LLM tagging stay accurate to my taxonomy?
The extraction runs against your own controlled schema, so the LLM produces structured output that conforms to the topics, funnel stage, persona, and format categories you define. You control the taxonomy, and the output is guaranteed to match its structure rather than free text.
Does Wave keep the metadata current as I publish?
Yes. When a connector sync detects new content, Wave can automatically crawl and extract it, opt-in per tenant. That keeps your content metadata fresh as you ship, instead of going stale the way a hand-maintained taxonomy does.
How does the person-level join work?
Every contact's engagement events reference the crawled asset and its extracted tags. That makes consumption history queryable per person, per topic, and per stage, so you can see exactly what each buyer has read, watched, or downloaded and what it covered.
Can I export the extracted tags?
Yes. Every extracted tag lives in Wave's database and is exportable back to your CRM. You own the metadata Wave produces, and it can flow back into the tools where your team already works.
Does my content data train a shared model?
No. Wave runs extraction one tenant at a time. Your content and engagement data never enters a shared or cross-customer model, and Wave is tenant-isolated from the moment data lands.
How do I see Content Intelligence on my own library?
Book a 20-minute walkthrough. We will configure a tenant against your stack and show the structured tags Wave would extract from your content and how they join to your contacts' engagement.
See it on your data
See what your content is actually about.
Book a 20-minute walkthrough. We will run Content Intelligence against your stack and show the structured tags Wave would extract and how they connect to what your buyers have consumed.
Request a demo Download the one-pager (PDF)