Is it me you're looking for?
In my current role as visual and user experience designer to the M&S international team I am responsible for managing the digital content of 7 websites that give users a seamless and rich commerce experience. These localised sites have helped grow the international business which delivered a combined revenue of £1.1bn in 2016.
Brief from M&S stakeholders:
In July 2016 the Marks & Spencer contact centre got in touch with the international product team and requested development based on observations they had made of customer behaviour: “I am a customer in the south pacific and I want to get in touch with you outside of UK working hours, but I can’t.”
Stakeholders in the business interpreted this and presented me with a solution: “Design a pop up contact form for customers to fill in their phone number and name so that they can submit a request for a call back from the contact centre. We’ll put a link to the popup in the footer.”
Identifying the problem:
I received this brief and questioned why it had come to me as a 'finished' solution. I went back to the start, researching the original story and then set to work re-framing the problem. I was interested in why customers in the south pacific wanted to get in touch with the contact centre. Only by understanding the customers’ first problem and the reason for their contact would I be able to design the best solution to fit their needs.
The research process:
In my deep dive on customer queries to the contact centre, I found that overwhelmingly customers were frustrated and making contact because they wanted information. 72% of the contacts made were because the customer wanted to know where their order was.
I then heuristically evaluated the current customer journey as it is when making a purchase. I focussed on heuristics that I knew would be important as measures of success for a customer when placing and receiving an order :
- What help and feedback was sign-posted to the customer
- What ongoing interaction with the customer there was during the order process
- What support and teaching was provided to the customer throughout the process
I found a list of areas that could be improved upon, and made the following recommendations:
- Better communication of the ordering process and expected delivery dates
- Better follow up emails after an order is placed which should be more informative
- Bring forward frequently asked questions so that they are easier to find and access
- Increase support for customers so that they can make contact if needs be
- Signpost delivery times more clearly throughout the website
As well as conducting a heuristic evaluation I wanted to focus on the customer and so made effort to understand how the customer would be feeling and what thought processes there would be whilst placing and waiting for an order. I created a journey map to deeply examine the experience of the customer and therefore help me identify where the process was falling down.
As a result of creating the journey map I discovered that when placing an order as a customer I found it frustrating to not get immediate answers to common questions such as, 'how long will it take to receive my order?' Also, having to wait until unsociable night-time hours to speak to an advisor when I should be sleeping made me feel alienated and angry!
The solution:
From my research I set about generating ideas as to how to solve some of the frustrations and problems I had found. As well as some quick wins which were straight forward to implement, like signposting delivery information and FAQs in a clearer way, I wanted to create a live chat feature for customers to use that would enable customers to make instant contact and remove the frustrations they had as a result of not getting their answers immediately.
As designers, we are aware that any product we design for a customer should spark joy and solve a need. It should definitely not make life more difficult or more frustrating and so the chatbot would need to be seamless and well planned. The setup of a good chatbot is complex, at its core it would require increased resource and training in the contact centre to support such a product which at present M&S international does not have the budget to support.
So, with that problem in mind and working as an avid agile practitioner, I set about creating a chatbot minimum viable product (MVP). My MVP would be the most basic version of a chatbot, a bot that isn’t able to ‘chat’ at all. Instead, my bot would feel to a user like it’s an automated bot but in fact behaves as a useful tool. The design would help by guiding a user towards frequently asked help topics with an aim to helping them self-serve and resolve their own problems with support and signposting. In this way the product would be seen by the customer as a 'helpbot,' but not require additional resource or investment from M&S.
From my research I knew what customers were getting in contact about and therefore I would be able to pre-empt their questions and serve them the answer through the helpbot instantly. It’s a win-win solution as for the customer, they get the answer they want instantly and for M&S, as the number of contacts would decrease, thereby saving cost.
Sketches / Prototypes:
I worked quickly to create paper prototypes of various helpbot solutions so that I could test and learn from the models. I needed to be sure that customers would understand how to use the helpbot, where to find it and that they would like the interactions, so I tested my paper prototypes on my stakeholders and colleagues. I then refined the language used with the company contact centre as my guides – they are the people who speak to these customers regularly and so it was important to me to follow their learnings and tone so that the helpbot felt authentic and part of M&S as familiarity is key to a well-rounded user experience.
The ‘finished’ product:
I use the word ‘finished’ in inverted commas as for me a product is never finished and this was merely the MVP. Using the results of my testing I moved to the computer and to Adobe Photoshop to create the finished visual design, I delivered all UI elements and wireframes for the responsive design of the helpbot to the product team in order to enter the team sprint. I worked closely with the developers and testers to ensure that the finished helpbot was in line with my designs. We then released the product to a portion of customers through use of an A/B test and we would measure success by monitoring the ratio of contacts by order into the contact centre. We would expect to see that the helpbot would reduce the need for contact.
View the helpbot here: www.marksandspencerlondon.com/au/contactus
Results:
The results were phenomenal. We found that very quickly customers were interacting with the helpbot and clearly they were able to get the support they needed because the ratio of contacts by order reduced from 40% to 13% in one month, this result would go on to equate to a business saving of £80,000 a year based on each contact costing M&S £5.
Next steps:
I’m a designer who believes in continuous improvements and so the journey doesn’t end there. I am closely analysing data from the contact centre to see what works well and where there are still frustrated customers as well as tracking the success of the product to see where I can refine the design further.
In addition to this, the success has given me a strong business case to look further into development of a true chatbot or use the helpbot across other areas of the journey such as size guides and styling.
© Sara Wright 2016 & Marks & Spencer LTD
