You are currently viewing Hatching Chatbots for eCommerce in Bangladesh

Hatching Chatbots for eCommerce in Bangladesh

Online shoppers in Bangladesh are well-known to be among the first generation of f-Commerce buyers – the boom of social commerce that emerged with the popularity of Facebook and Instagram.

Today, there are countless online businesses across a wide variety of niches who secured funding from somewhere like this website, set up their own store, and tried to outsell their competitors. The primary source of many of these small businesses is Facebook and in certain markets, even Instagram, which serves the higher income classes.

However, social commerce also comes with the drawback of not being as scalable as a revenue channel as a website. You are only selling as long as you are promoting which means for low margin products, sustaining digital ads will not be feasible in the long run. As queries and comments pour in, manning the responses becomes troublesome especially when unanswered queries mean lost customers. For startups operating with lean teams, affording an eCommerce website or hiring community managers is not always feasible.

As COVID-19 ushered lockdowns in the beginning of the year, it also put an unprecedented demand on online stores as visiting physical ones posed health risks. While this led to a somewhat comical explosion of online grocery and essential item stores, the explosion of organic traffic did not solve the aforementioned problems of social commerce. As a consumer, it was difficult to track your order and see available products without having to navigate to and fro from a Facebook Page and Messenger.

All this after the inevitable 1-2 hour wait time for responses, back and forth for your confirmation often to find out that your desired item is not in stock. On the merchant side, an uptick in business means having to deal with inventory issues, suppliers providing incorrect or substandard products, logistics, queries coming in at all times of the day e.t.c.

To solve this issue, ChatLeads and DigiMarka partnered up to provide a systematic social commerce solution that would enable order tracking and online payments with the end objective of lowering risk of online merchants of returns and enhancing the customer experience of the end customer.

Our experiment took the shape of a Messenger chatbot – Hatch which gave its audience the unique value proposition of completing their weekly grocery in under a minute and checkout with online payment. We wanted to test the limits of the number of queries we could handle per day, and how a demographic, typically resistant to change, would adopt to a Messenger based eCommerce system.

Our objective was to measure acquisition costs, conversion rates and retention as a measure of understanding performance versus that of a website or app. Furthermore, we wanted to measure the cost of repeat orders to ultimately calculate the R.O.I on user acquisition.

 

Acquisition

As far as targeting was concerned we were targeting the female shoppers in select areas in Dhaka – which can be considered a premium online audience. The purpose of our ads were to engage and educate right within Messenger itself.

Clicking the ad would take users to Messenger and trigger an automated sequence.

Whenever users select an option, the chatbot asks for a user’s location and tags them accordingly. It enriches our audience database and subsequently helps in our retargeting campaigns.

Users could select the package of their choice, add it to a cart, enter their address and checkout with online payment as the only available option. Our acquisition cost for the lifetime of the 2 month campaign was BDT 16.24 where we are defining a user as someone who engaged with the chatbot up until the location stage. As a comparison, the cost of driving a message to a retargeted Mad Koffee audience ranges from BDT 60.00 – BDT 80.00.

In our experience, a website click for a similar audience would cost around BDT 10-12 for a well optimized campaign. However, with Facebook Messenger we were receiving the audience name and location as well.

Leave a Reply