Background
The florist is a family-owned shop in a large desert city in the Southwest. It delivers the same day across the city and its suburbs and makes wedding, event and sympathy flowers. Its customers ask ChatGPT, Google AI or another AI engine which florist can deliver today, who does funeral flowers or who makes the best wedding bouquets, and they call one of the shops in the answer.
In the first two weeks of Reachd analysis, AI ranked the florist first among 93 and named it in 28% of answers. Tatum Flowers and three other local shops were named almost as often, in 22% to 23% of answers, close enough to take the top spot on any given day, and in mid-September PJs Flowers & Events did. A lead that thin meant the florist was one of several equal names in an answer, and the customer could just as well call another.
Goals
- 01 Turn a thin lead into a clear first place in AI answers
- 02 Win the questions where competitors were still preferred
The owner also wanted the lead to hold as AI answers change from day to day. And AI had to be able to answer the practical questions customers ask before they order, such as by what time an order has to be in, where the florist delivers and whether the flowers survive the summer heat.
Results
- #1 of 74 florists The florist became the one AI recommends most in the city. It was named in 98 of 190 answers, Tatum Flowers in 60, PJs Flowers & Events in 52 and Anne's Flowers in 49.
- One customer in two Its share of AI answers doubled from 26% in mid-September to 52% at the start of October, and its AI visibility score went from 22 to 45.
- 8 in 10 ChatGPT answers ChatGPT named the florist in about eight answers out of ten, up from about three in ten in the first two weeks.
How the florist used an AI agent to take first place in AI recommendations
The owner needed no developer for any of it. Her Claude agent was connected to Reachd through MCP and to the Shopify store, so the same agent that read a recommendation could publish it. It pulled the day’s analysis, with the questions customers asked AI, the florists named on each and the recommended changes with ready text, and then published the change in the store. The next morning’s score showed whether the answers had moved, and most changes went live within a day of the recommendation.
- Mid-September
Delivery terms and a wedding page
The delivery page was rewritten in plain sentences: same-day delivery Monday to Saturday for orders placed by 3 PM, Sunday delivery for orders placed by 3 PM on Saturday and a fee from $15 by zone. The studio page gave the address and the same terms, and a new page described the wedding work.
Back to first place - Around September 21
How the flowers survive the heat
A new page explained how the florist keeps flowers fresh in a desert summer: flowers bought almost every day, a walk-in cooler, a refrigerated compartment built into the delivery van, and no arrangement left on a doorstep from May through September.
From 33% of answers to 52%
When AI recommended the florist, it most often cited the studio page, the arrangements collection and the new wedding florist page. ChatGPT repeated the delivery terms from the studio page almost word for word: the 3 PM cutoff, the Monday to Saturday schedule and the $15 starting fee. In another answer it called the florist the best choice for a truly custom bouquet.
AI passes delivery terms on word for word, so a shop that writes down its cutoff time can be recommended for a same-day order.
The cutoff, the delivery zones and the refrigerated van were part of the business long before they were on a page. Once the agent wrote them down where AI reads, they started reaching the people asking. A business can check what AI says about it and see which of its pages AI already cites.