Review engine
The scenario:
Before a prospect books a call, they look you up. Website first, then the reviews. Yours say 2024.
You have delighted clients. You fixed a disaster last month and the client called your engineer a hero. None of it is written anywhere a prospect looks. The gap between the service you deliver and the proof on display is costing you deals you never knew you were in.
The prompts:
Your AI tool of choice.
STEP 1 - Find the moments
Step 1 – Find the moments
Run it in your AI tool of choice.
You are building a review-generation system for an MSP.
Context:
Where reviews matter for us: [Google, industry directories, case study quotes]
Recent wins: [list 5 recent moments clients thanked you, praised an engineer, or renewed]
Current review count and age: [numbers]
From the recent wins, identify which clients to ask, in what order, and what specific moment each ask should reference. A reference to a real moment doubles response rates. Flag any client where an ask would be badly timed.
-
STEP 2 - Write the asks
For each client from step 1, write the ask message:
Under 80 words, personal, referencing the specific moment
One link, one action, no pressure
A variant for email and a variant the account manager says on a call
A thank-you reply for when the review lands
-
STEP 3 - Make it a system
Build the standing process: the trigger moments that prompt an ask (ticket praise, renewal, project completion), who owns sending it, the monthly cadence cap so no client is over-asked, and a simple tracker with columns for client, moment, date asked, and result. Keep the whole system inside 30 minutes a month.