# Daggy > Daggy is independent ad measurement and budget planning for higher-ed marketing teams: it shows which ads bring applications and where next year's budget should go. Daggy uses causal inference, forecasting and budget optimization. It doesn't buy media, so it has no stake in where a school's budget goes, and it works alongside the school's agency. Schools start with a six-week pilot: send Slate or CRM exports plus ad-platform spend (Google Ads, Meta, LinkedIn and the rest), or Daggy does the export with you. Daggy estimates what each channel did for each program, with the season and deadlines taken out, and proposes an allocation by channel, program and month, with ranges. The school decides what moves. Daggy works from totals by program, channel and week, never student records. UVA McIntire School of Commerce's marketing team is a customer. ## Pages - [Home](https://www.daggydata.com/home.md): what Daggy answers, a customer quote, the pilot, independence and the FAQ (data, security, who decides) - [Request a pilot](https://www.daggydata.com/contact) - [About](https://www.daggydata.com/about.md): the founder - [Blog](https://www.daggydata.com/blog/): posts are under /blog//index.mdx ## Optional - [MCP reference for agents](https://www.daggydata.com/mcp.md): Daggy's statistical engine is also an MCP server; tools, arguments, workflow and troubleshooting - [AI NFL predictions vs Kalshi](https://www.daggydata.com/nfl): eight AI agents, with and without Daggy's statistical engine, forecast every NFL game against Kalshi's market, scored all season: a public test of an unproven method, against a benchmark it isn't expected to beat. The agents never see market prices; the test is how close they get to a liquid market, and whether the engine moves them closer - [NFL week-by-week AI predictions](https://www.daggydata.com/nfl/2026/week-1): one page per week, every game's prediction and result, under /nfl//week- - [The Roundup](https://www.daggydata.com/nfl/roundup): weekly NFL recaps, under /nfl/roundup//index.mdx