Use case
Share of voice you can defend in the meeting
Who owns the category conversation, computed from daily data across mentions, posts, and ads, with the receipts one click away.
The short answer
Waldo tracks share of voice by collecting mentions, owned posts, and active ads for every brand in a category on the same daily cycle. Because earned and paid live on one surface, the SOV number and the paid picture come from the same pull, and every count traces back to real posts.
The job to be done
Every share of voice report faces the same two attacks: the definition is arbitrary and the data is partial. Both usually land, because earned lives in one tool, paid in another, and the analyst hand-blends them in a spreadsheet nobody else can audit. When the client asks why the number dropped four points, the honest answer is often a methodology change nobody documented.
The number that survives the meeting is computed the same way every month, decomposable by platform and theme, spanning earned and paid, with every count traceable to actual posts. That means the inputs need to come from one surface, and the computation needs to live in a script, not in someone’s muscle memory.
How Waldo does it
Define the set once
Activate every brand in the category. From then on mentions, owned posts, and active ads for all of them refresh on the same 24-hour cycle, so each brand is measured against the same day.
Pull volumes per brand
MCP: brand_overview and the conversation endpoints return per-brand activity as raw counts or scored analysis. Your script decides the SOV formula; the inputs stay consistent underneath it.
Cut it by theme and platform
Clustered conversation data shows where each brand actually wins: a rival can own Reddit and be absent on TikTok, and a flat topline can hide a lost theme. The cuts are the story.
Script the monthly number
A cron pulls the counts, computes the shares, and publishes with source links attached. When the number moves, the diff shows exactly which conversations moved it.
The endpoints
MCP: brand_overviewPer-brand activity counts across earned, owned, and paidMCP: category_trendsThe themes each brand’s share decomposes into/audiences/{id}/conversationsThe mentions behind the count, clustered and linkedDiscover endpointsLive queries for brands outside the tracked setQuestions teams ask
How does Waldo calculate share of voice?
Your way. Waldo returns the underlying counts and conversation data; the formula lives in your script where it can be audited and held constant. That transparency is the point: no black-box index between you and the number.
Does it cover paid share too?
Yes. The Meta and Google ad libraries are scanned daily per tracked brand, so paid presence sits on the same surface as earned mentions, and the two halves of the story stop living in different tools.
What happens when someone challenges the number?
You click through. Every output links to its sources, so a disputed month decomposes into the actual posts and ads behind the count instead of ending in a methodology argument.
Keep reading
Put Waldo behind your agents
Brand, category, and audience intelligence over 200+ API and MCP endpoints. Sign up, mint a key, and run it against the brands you actually track.