Retale launches a Facebook messenger bot

Retale launches a Facebook messenger bot

Retale launches a Facebook messenger bot

Sean Keith

Can you tell us a little bit about yourself?

I’m a part time worker at Bonial, enjoying the freedom of having my own Special Operations department within the Online Marketing team that focusses on rapid realisation of growth and experimental tech projects.

I’m in my fourth year at the company, started off with a focus on conversion rate optimisation on mobile and web and quickly became involved in a variety of cross-departmental projects, from building a CRM tool for automated push notification campaign management to prototyping an immersive virtual reality shopping experience.

I have been given the opportunity to develop a very diverse and flexible skill set alongside gaining invaluable practical knowledge as I take on projects, that are hard to fit into usual, everyday company operations.

Retale has just released a Facebook Messenger bot, tell us more about this feature.

With the hype around chatbots it came as no surprise that we got our thoughts around how we could leverage this technology to amplify reach and grant users access to our content within their beloved chat environment.

RetaleBot now delivers your weekly ads and coupons to you via Messenger. Besides getting the popular deals in your area, you can ask RetaleBot for the closest Walmart or Target store or let it manage your shopping list by adding items and then have it present local deals for each item on your list. You can also subscribe to receive updates when new weekly ads for your favourite stores become available.

To get started with it you simply search for Retale in the Messenger app and start a conversation just as you would with your Facebook friends.

Who worked on the bot and how long did the development take?

The bot is still in an early stage of development. We iteratively design and develop the bot as we learn how our users interact with it and what their expectations are.

I am the only developer directly working on the bot, but it is tightly coupled to the deals and coupons data infrastructure that is built, maintained and fuelled by the IT and Content departments.

The design of user flow, definition of scope, phrasing and other optimizations happen in close collaboration, most notably Karsten Lehmann, with the Inbound Marketing team and we also had at least a handful of testers, who’s feedback was invaluable in making the last steps towards publishing.

To give a rough timeframe, reaching the point of publishing took around four weeks, but this is an early milestone,enabling us to continue developing the bot in the right direction through evaluating real user feedback.

What challenges arose while developing Bonials first chatbot?

 A major challenge was the openness of such a conversational interface. At the beginning you could set up some basic functionality, hand the bot over to a colleague and you could be sure he would fail to formulate the phrases in a way you had anticipated. So when considering and laying out possible input phrases, you really need to become creative and keep them as open as possible.

It quickly became clear, we had to guide the user by carefully formulating responses, providing the right interactive elements and keeping the scope clear at all times.

Defining the scope in the first place has also been a challenge. You need to figure out which features really make sense in a conversational environment and then how one can project them into a text-dominated experience.

In a classical app or web interface, for example, you can provide huge amounts of visualised data, that a user can quickly scan and filter to find the relevant information. In a text based environment, every irrelevant sentence means extra work of reading for the user and can be very distractive and misleading.

Do you have plans to expand the bot with new features?

With the publication of the bot, we’ve just reached a point where we can gather insights on how exactly users interact with the bot, and from here we will closely monitor, test and evaluate to discover features that support our mission and help users find the right deals at the right time.

Will the Messenger Bot become accessible for other Bonial Brands?

Yes. We are currently teaching the bot German for kaufDA and French for

What’s the main learning?

We’ve started to learn how to deal with conversational UI. This will help us come up with solutions not only for text-based, but also for voice-controlled environments, such as Amazon’s Alexa or Siri.

RetaleBot gives us insights on how people interact in such conversational environments, preparing us to go with technologies that may potentially dominate the future beyond mobile apps.