This guide covers the fastest, lowest-effort way to use RUSHA: as a live co-pilot while you're on the phone or in a Quarterly Business Review (QBR) with a client. No setup, no waiting on a background job — just ask.
When to Use This
- A client asks a question on a call you don't have the exact number for.
- You're prepping talking points 10 minutes before a QBR.
- A client emails "why did our spend jump?" and you need a real answer, not a guess.
How It Works
- Open the client's dashboard → RUSHA AI chat.
- Type the question in plain English — RUSHA has live tool-calling access to every connected platform for that client, so it pulls real numbers, not cached guesses.
- You get an answer in under a minute, with the exact figures it used to get there.
Because every number RUSHA states is mathematically checked against the raw integration data before it's shown to you (see Mastering the RUSHA AI Insights Engine), you can read the answer to a client verbatim without double-checking it yourself first.
Example Prompts That Work Well
Simple lookups (instant):
- "What was our Google Ads spend last week?"
- "How many leads did we get from Meta this month?"
- "What's our current cost per conversion on the Brand campaign?"
Comparisons & root cause (a bit more reasoning, still one insight):
- "Why did our CPA increase this week?"
- "Compare this month's Meta performance to last month."
- "What caused the traffic spike around August 8th?" — RUSHA will cross-reference every connected platform for that client to find the real driver, not just the platform you asked about.
Things it will correctly refuse:
- Anything unrelated to the client's connected data (e.g. general marketing advice, unrelated small talk) gets a graceful "that's outside what I can verify" response instead of a made-up answer — this is intentional, not a bug.
Tips for Live Use
- Ask before the call, not during it if the question is complex — give RUSHA a few seconds to reason through a multi-platform question rather than typing it live while the client is waiting.
- Trust the number, not your memory. If RUSHA's answer differs from what you remembered, RUSHA is checking the actual connected data right now — your memory isn't.
- Follow-up questions keep context. You can ask "and how does that compare to last month?" right after, without repeating the client or date range — RUSHA remembers the thread.
- Every chat turn (question → answer) uses 1 insight credit from your monthly allowance — see Pricing, Quotas, & Billing. A failed or "not enough data" response is never charged.
Related Guides
- RUSHA AI for Internal Audits & Proactive Monitoring — catching problems before the client asks about them.
- RUSHA AI for Client Reports & Proof of Value — turning verified insights into things clients see automatically.
- Mastering the RUSHA AI Insights Engine — the technical deep dive on guardrails and how answers are verified.
