15× more repeat orders

A national dry-cleaning chain had a large database of past customers who had not placed an order for some time. We built an AI assistant that used their order history to start personal conversations. This generated 15 times more repeat orders than the usual bulk messages to the same customer base.

Give people a reason to reply

The chain sent typical bulk messages: “We haven’t seen you for a while. Come back for a 10% discount.” Most people ignored them because it was obvious that everyone had received the same offer.

Yet the CRM already held enough information to make each conversation personal: what the customer had brought in, when, and what they had discussed with the staff. We connected an AI assistant, Mary, to this history. Before writing, she checks the previous order and can ask, for example, whether the customer was happy with how their coat had been cleaned last season. A question about your own coat gives you more reason to reply than another discount offer.

The conversation follows the customer’s response. Mary can ask about the previous order, find out whether anything else needs cleaning, and offer the same discount a few messages later. The offer has not changed, but now it comes after a conversation about that person’s experience. More than a third of customers replied to these messages on WhatsApp and Telegram.

When someone wants to place an order, Mary asks for a description and photos of the items, then hands the conversation to a member of staff. This is the client’s preferred process: AI brings customers back and starts the sale; people complete the order.

We had to slow it down

Conversion to repeat orders increased fifteenfold, and the chain soon struggled to handle the volume. We added automatic control over the pace of outreach, taking the business’s workload into account. The company could now bring customers back at a rate it was ready to serve.

The same offer, to the same customer base, brought in 15 times more orders.

An unexpected benefit: finding unhappy customers

Some customers had stopped returning because of a previous problem: poor cleaning, a missed deadline or a damaged item. They ignored promotional messages, but when Mary asked about their experience, they explained what had happened.

We added a mechanism that alerted the business and summarised each complaint, so a manager could join the conversation and resolve it. Sometimes a discussion was enough; in other cases, the company offered a bonus or compensation. It finally had a way to repair relationships with customers whose dissatisfaction had gone unnoticed.

The company restored relationships with most of these customers, and many placed orders again. Alongside repeat sales, it gained a way to discover why people had left and speak to them while there was still a chance to put things right.