Stop Giving Money Away! Using AI Negotiation Chatbots To Reduce Margin Sacrifice
Home » Podcast library » Ecommerce capability » Stop Giving Money Away! Using AI Negotiation Chatbots To Reduce Margin Sacrifice

Stop Giving Money Away! Using AI Negotiation Chatbots To Reduce Margin Sacrifice

An interview with Nibble founder and CEO Rosie Bailey on using AI negotiation chatbots for negotiating deals with online shoppers to protect margin and reward loyal customers.

Listen on your favourite player

Watch the video


  • What is AI negotiation and how does it benefit an ecommerce team?
  • How do you avoid giving away money you didn’t need to?
  • Business controls to determine when to enable negotiation
  • Examples of retail ecommerce stores getting results from this approach

Nibble is an award-winning chatbot for ecommerce that lets customers make an offer on the product page for instant results. Nibble’s aim is to nudge undecided shoppers into action, to drive higher conversions and bigger baskets, with less margin sacrifice than traditional discounting.

To avoid giving away unnecessary discounts, you can program Nibble to appears only when you want it to and to specific customers. For example, you can make Nibble a subscriber-only benefit, product-specific, or trigger only on exit intent or dwell time. There are many use cases, including clearing excess stock through targeted offers and taking back control of discounting to protect margins and strike deals based on value, not just price.

Tune in to listen to founder and CEO Rosie Bailey explain how the chatbot works and how other retail ecommerce businesses are using it.

Discussion notes

  1. Let’s start with the basics for anyone who hasn’t come across this term and use case for AI before: what is AI negotiation and why is it relevant in ecommerce?
  2. Where did the idea come from, what led you to creating Nibble and how long have you been live for?
  3. And it’s worth getting the elephant out the room upfront as ‘negotiation’ will bring up visions of slashing prices and killing margin: what is the goal of Nibble & how does it avoid a race to the bottom of aggressive discounting to secure a sale?
  4. I always advise people to run hold out tests wherever possible, to compare a cell using the new tool vs. a cell not exposed to it, can you share some examples where retailers have done this and what they’ve seen in the data?
  5. You’ve got a lot of knowledge and experience around the psychology of negotiation, can you summarise some of the scientific learning that is being used to hone how the AI works and delivers results?
  6. Let’s talk about setting up the chatbot: how does a business ensure the language, tone, style etc. are aligned with the brand (I like the example of seeding brand and product USPs into replies e.g why the product is high quality)? How much brand control is there vs. applying a few settings and letting the AI learn from there?
  7. How can ecommerce teams control who sees the chat prompt, can you share some examples where retailers are restricting who sees it, displaying if certain conditions are met etc.?
  8. And how extensive are the business rules customers can apply, can you talk us through some of the most commonly used rules?
  9. I’m going to ask another probing question on impact: have any of your customers modelled the impact of customers interacting and purchasing with Nibble on LTV, so repeat purchase rates and spend vs. customers who don’t have a Nibble interaction? If not they should!
  10. I love the tool, it’s a great idea well executed, the UI is slick and non intrusive so well done. What are the key updates & new features customers can expect soon?
  11. How does the integration work – is this about adding Javascript, or do you have pre-integrated apps on any of the leading platforms?
  12. And finally, can you explain the pricing model, as the tool is giving away a discount so how do you ensure it’s affordable?

Want to suggest a topic or guest for a future episode? Contact us via the website or on LinkedIn.


Listen / Subscribe