All products

Advertising proposal research with AI

Build a traceable evidence table before turning market findings into a campaign argument

Advertising proposal research with AI starts with the decision your pitch needs to support. The ad-research-collector skill turns that question into search terms, prioritises original reports and records evidence in a table rather than handing you an unexplained pile of links.

Each finding carries its year, source quality, limitations and suggested slide use. The supplied source guide centres on Chinese market reports and platforms, so your agent must adapt its research to the market you specify. It structures research; it does not grant rights to reproduce third-party reports or create a finished presentation.

What it does

  • Investigate market size, buying occasions or channel changes
  • Separate primary reports from media summaries
  • Record links, publication years and confidence levels
  • Map a finding to the proposal claim it could support
  • Expose outdated evidence and gaps requiring further research
  • Keep unsupported conclusions labelled as hypotheses

Install and use

  1. Copy the skill folder into Codex’s skills directory and start a new session with web-search access.
  2. Provide the campaign question, target market, relevant dates and decisions needing evidence.
  3. Ask the agent to search official sources first and identify secondary reporting explicitly.
  4. Review source links, limitations and reuse conditions before quoting findings in a proposal.

FAQ

Source

Made by Alex Dongqing / ilexlyu. Original skill: ad-research-collector, published on GitHub under the MIT licence.

View on GitHub