The team at 1080bots has been thinking more about the Return on Investment model to help retailers justify their initial foray into the chatbot space. Having seen a range of chatbot applications, from Best Buy’s Order Status Bot to Ebay’s Shopbot, we wanted to analyze the benefits of augmenting a customer service call center with a chatbot.
For this research, we chatted to JD Sports in the UK via its Facebook Messenger service. I said “Hello” at 12.24pm and 30 minutes later I was greeted by Anisha, “Hi, How can we help?” That was a pleasant enough greeting on Nov 17th at 12.54pm so I responded immediately.
Then I waited patiently until the next day, sending a reminder at 4.54pm on November 18. “Anisha… what happened?”. At 6.20pm Sabrina replied. The JD Sports Messenger did say “Typically replies in a day” so the timing is consistent with that.
I am looking for a Christmas Gift for my 8 and 12 year old nephews and I had found something specific on Amazon’s US site. I sent a link and a new representative, Kirsty, let me know that this is not an item that JD Sports stocks.
When I asked for a similar product, Lexie replied asking for more information. I’d already provided a link so I referred back to my earlier comment. Then Kirsty replied again – was this Kirsty I’d spoken to a few days ealier? Not sure, but it didn’t matter because even she could not see the whole thread. The conversation continued until finally Anisha came back to let me know that JD Sports did not actually stock football nets. Possibly the same Anisha but she didn’t remember me or the original conversation.
When I compared the conversation to one with a chatbot, the experience was lacking in three areas:
Let’s now go back to the matter of return on investment on the Chatbot. What if JD Sports put a chatbot in to handle the initial Facebook Messenger interactions? The bot would be able to handle certain queries itself and hand others off to Sabrina, Kirsty, Anisha or Lexie.
Our hypothetical call center has 50 seats and answers calls between 8am and 8pm. The average call is 5 minutes long, and each agent takes about 10 calls per hour. That’s 2,400 calls per day. Hypothesize that 50% of those calls are for order status, product enquiries or FAQ type questions. If a chatbot is used to completely handle 50% of the order status, product and FAQ calls – that’s 600 calls.
Let’s say that average cost of an inbound call to a customer contact center would be around $5. For 600 calls, that’s $3,000 per day. That is a total of $780,000 per year, far outweighing the cost of the chatbot development and maintenance.
We’re not advocating at all for chatbots to take over all human customer service agents tasks, but used as the frontline mechanism for Facebook Messenger or any other of the messaging platforms, we believe that a chatbot can very quickly provide a positive return on investment.
If you want to build the business case for an ecommerce chatbot or to prototype a customer service chatbot that works in unison with your call center agents, then drop us a line at firstname.lastname@example.org