MCP
Detect positioning shifts across a year of videos
Compare how a rival described their product a year ago vs today; answers cite the exact moments the pitch changed.
library_import_youtube_channeldeep_contextThe problem
Repositioning never announces itself. No competitor publishes a press release saying "we're abandoning SMBs" or "we're quietly becoming a security company." The pivot happens one word at a time: "easy" becomes "enterprise-grade," the demo persona changes from a founder to a compliance officer, the integration slide gains three logos and loses four.
By the time the repositioning is obvious, they've had months of head start in the segment you both now want. But every step of that drift was recorded — in their launch videos, webinars and demos, dated and public. Line this year's language against last year's and the pivot stops being a hunch your team argues about; it becomes a documented trajectory you can plan against.
How veridive MCP helps
library_import_youtube_channel captures the rival's full published record into your veridive library — transcripts included, upload dates preserved. That dated corpus is the raw material: positioning drift only shows up when old and new sit in the same queryable place.
deep_context does the longitudinal read. Scope one question to their older videos, the same question to their recent ones, and compare — or ask for the drift directly: "how has the way they describe their core value changed?" Answers come back with citations to the exact chunk of the exact video, so the finding isn't "our CMO feels they went upmarket" — it's a two-column story with timestamps: the words they used then, the words they use now, and the quarter the switch happened.
How to run it
- 1Connect veridive MCP
One-time setup: add the veridive MCP server to Claude, ChatGPT, Cursor or any MCP client and generate your API key. Takes about two minutes.
- 2Capture their record
Import the rival's channel — the whole arc, not just recent uploads. The old videos are the baseline the drift gets measured against.
Import <competitor>'s YouTube channel into my veridive library, tagged '<competitor>' — use veridive's library_import_youtube_channel. - 3Baseline the old pitch
Scope the question to their earlier videos: who they targeted, the promise they led with, the words they leaned on. Citations pin each observation.
From <competitor>'s older videos: how did they describe their core value and target customer? Quote the recurring phrases, with citations. - 4Diff against the current pitch
Same question, recent videos — then ask for the delta. What appeared, what vanished, what got renamed.
Now the same analysis on their recent videos — and list what changed: new language, dropped claims, different demo personas. Cite the moments on both sides.
Problems it solves
- Competitor repositioning is only noticed after it's complete — when they've already claimed the ground.
- "They feel more enterprise lately" is a vibe; without dated quotes it can't drive a strategy decision.
- Nobody rewatches a year of a rival's videos to reconstruct how the pitch evolved.
- Team debates about a competitor's direction run on memory, and memory smooths out exactly the drift you're looking for.
Why it's valuable
- Repositioning becomes visible mid-move — while there's still time to respond in your own messaging.
- The finding is a documented trajectory: then-quotes, now-quotes, and roughly when the switch happened.
- Works retroactively: a year of their published record is importable today; the baseline already exists.
- The same corpus keeps answering — every quarter's diff costs one more question, not another research sprint.
Example prompts
Import <competitor>'s channel into my library — use veridive's library_import_youtube_channel.From their older videos: how did <competitor> pitch their core value? Recurring phrases, with citations.What changed in <competitor>'s messaging between their older and recent videos? Cite both sides of each change.FAQ
How far back can the comparison go?
As far as their channel does. Everything importable comes in with upload dates and transcripts, so a multi-year arc is as queryable as last quarter.
How is the old-vs-new comparison scoped?
deep_context runs on a selection of library items — your assistant scopes one pass to the older uploads and one to the recent ones (up to 20 videos per question), then synthesizes the diff. Every observation cites its source chunk.
What signals should I ask about beyond wording?
Demo personas, the industries in customer stories, which integrations get airtime, what the pricing conversation avoids. Positioning leaks through examples even when the tagline stays frozen.
Related use cases
Import every competitor's channel and interrogate all of it in one conversation — "what did they emphasize this quarter that they didn't last?"
Import each launch, summarize claims, and query the differences — battle cards grounded in what they actually said.
Mine how real users describe the problem in reviews and vlogs — copy that converts is usually a transcript quote.
Run this in your own assistant
Connect the veridive MCP server to Claude, ChatGPT or Cursor and paste the prompts above — setup takes about two minutes.