You base recommendations on the customer's purchase, not dead stock
Deciding to offer one more product used to start with whatever was piling up in the shop or cost more, without checking if it actually matched what the customer had already chosen. With the question about how the add-on improves the purchase coming before the offer, every recommendation now starts from the main product chosen. That means more relevant offers and less time spent suggesting something the customer had no reason to buy.
You send one offer at a time, not stacked
It used to be common to suggest several products at once, hoping one of them would land, which usually sounds like pressure instead of a recommendation. With the habit of offering one add-on at a time, each message becomes easier to consider and looks less like a sales list. The shop ends up with clearer offers that the customer can weigh without feeling pushed in several directions.
You shift your pitch from insistence to benefit
Recommendation lines used to sound like insistence, repeating the offer or highlighting only the higher price of the suggested product. With the habit of explaining how the two products complete each other before recommending, the pitch starts showing the benefit of the combination, not the shop's wish to sell more. This lowers customer resistance and makes the recommendation easier to accept.
You keep a written record of each offer cycle
Every offer made after a sale used to end right there, with nothing kept about what worked. With the task of noting the product chosen, the add-on offered, how it improved the purchase, and the words used, the shop starts building a real record of what earns an accepted recommendation. This makes it easier to spot which combinations actually work for the shop's catalog.