Stories

A Lesson Learnt 🤔

2 min read

If you're reading this, you're either an avid follower of Carryr or potentially crazy - either way, we're excited to have your attention for a couple of minutes.

You may be aware of how quiet we've been over the past few months since we launched our concierge chatbot via Facebook Messenger throughout the summer. Let me tell you why...

Our chatbot was an experiment run over a 12 week period to find logistical and technical problems and to validate a few of our original assumptions. Now you're going to ask what they were. So, here's a few assumptions we had before going into the experiment:

  1. We could gather enough order data to better understand which retailers would benefit from a same day delivery service - turns out that it wasn't that simple, we'll explain why in the list of problems below
  2. We needed to integrate into retailers' inventory and become a delivery option - we were right, and it saved us wasting time building something without knowing if it's needed or not
  3. We can deliver products from large retailers and independents on day one - technically we could but it was a nightmare

Here's the part you've been waiting for, the problems:

  1. Why couldn't we gather enough order data? It turns out that it's super hard to acquire consumers to a Facebook chatbot (we tried different paid advertising techniques too) and then be able to build up enough trust to get them to pay for the service either on their doorstep or within the chatbot.
  2. Availability. This was the worst part! Majority of the requests contained products that weren't in stock in the stores in Cambridge and it took us a while to find out - this was mostly for large retailers. On some accounts we recommended an alternative product from a different store but this took too long to manually search through the retailers products. I hated letting people down.
  3. Poor user experience. We designed the chatbot to be able to take order information as quickly as possible. In under a minute an order request can be submitted but, in reality, this was not the case. If a consumer has seen something they like online, they're more likely to buy it from the site that they're on rather than explore other apps to complete the order fulfilment process.

So, what have we learnt and what are we doing now?

We now know that the chances of getting a fulfillable order significantly increases by becoming an option on a retailers' website. Why? Because the customers are already there and they're ready to buy, plus we have access to eligible products i.e. we know the location of their inventory.

We're now working incredibly hard to build eCommerce integrations for the top platforms; Shopify, WooCommerce, Magento & more. Over the next couple of months we will be partnering with some exciting brands and retailers that operate in Cambridge and we'll be opening up conversations with larger retailers too.

Info on integration, details and partnerships to come soon, keep an eye out!

Oh and if we're lucky enough to have a retailer to read our blog, you can register your interest here - we'd love to have a chat with you 😆

Author image

About Chris Jordan

Founder & CEO at Carryr
  • London