Automated support for greater efficiency and lower costs

Want to see how agent-based AI is used in practice at NeoTaste? Click here for the webinar.

This case study shows how NeoTaste, working with moinAI, has built a customer service system that keeps pace with growth: one that resolves cases, is integrated into the company’s own systems, and provides clear guidelines for every action. The path to achieving this involved a careful analysis of the company’s own support tickets and the courage to go live with a lean version at an early stage.

From a voucher book to an app with over 2 million users

Voucher books featuring deals for restaurants, saunas or special offers: NeoTaste reimagined this concept in 2020 and brought it back to life as an app. Today, the foodtech start-up operates in 52 cities across four countries, from Düsseldorf to Vienna and London.

As the business grew, so did the demand for customer service. Within a year, the number of users rose from 500,000 to over 2 million, and with it the number of support tickets. If a faulty newsletter is sent out to 200,000 people, 1,500 additional tickets end up in the inbox within 24 hours. Nevertheless, outsourcing was out of the question for NeoTaste, as quality and proximity to customers were the top priorities. They were looking for a solution that would scale with the business without costs rising at the same rate. The objectives were clear:

1. Make support scalable without taking on extra staff
2. Resolve recurring issues automatically
3. Reduce the cost per ticket whilst maintaining the same level of satisfaction

Why moinAI?

As a start-up, NeoTaste is judged on its cost-effectiveness. Before a decision was made, four questions were therefore on the table: What can the solutions on the market do, how much do they cost, how is the pricing model structured, and how quickly can they be implemented? In the end, moinAI gave clear answers to these:

• Transparent pricing model: Fixed prices that do not increase with every additional enquiry. Increased traffic therefore reduces the cost per ticket, rather than increasing it.
• Integrated rather than isolated: moinAI is connected to HubSpot and the payment provider Stripe via webhooks. Enquiries are resolved directly, not merely answered.
• Full control over the conversations: NeoTaste determines for itself how structured or free-form the conversations are and when staff take over.

Less free-form text, more solutions

NeoTaste relies on clearly structured dialogues: the most common enquiries are listed directly as options in the chat. These include refunds, questions about subscriptions and changing your phone number. A practical example: when processing a refund, the AI agent checks whether a payment has been made in the last 14 days: “I’ve found the following payment. Would you like to request a refund for this?”

Just one click is all it takes to trigger the refund via Stripe. If the payment was made outside the cancellation period, the AI agent explains this and immediately offers to hand the matter over to the support team. More complex enquiries from restaurant partners are forwarded straight to account management with all the necessary information, rather than getting lost amongst hundreds of tickets.

Clear results

The cost per ticket has fallen from €4.44 to under €3, and customers are just as satisfied as before. At the same time, the support team – comprising three full-time staff members – can handle 25 per cent more volume without having to take on new staff.

With moinAI, NeoTaste has established a customer service system that keeps pace with the company’s growth:

• Case-by-case,
• integrated into our own systems and
• with clear guidelines for every action.

The path to achieving this involved a careful analysis of our own tickets and the courage to go live with a lean version at an early stage.

Interested to hear more?

In this case study, you’ll learn, amongst other things:

• How NeoTaste filtered out the enquiries from 10,000 tickets that could be automated
• Why clearly structured dialogues work better at NeoTaste than an open text field

‍

• How NeoTaste increased its full automation rate almost fivefold within three months
• Why it took just a few hours per person to implement the system
• How NeoTaste aims to resolve 50–60 per cent of enquiries via chat in the next phase of expansion
• What companies in the growth phase can learn from this project