Yell for Business app

Yell is a leader in the business directory with over 2.9 million listed businesses across the whole of the UK. It works closely with a number of leading global technology companies in order to deliver the best possible products and advertising results for its customers. For over 50 years they have supported the marketing needs of local business offering a number of solutions and tools to support every business and enabling them to quickly create an online presence. Using their relationships with the leading players in the industry - including Apple, Facebook, Google, Microsoft and Amazon, Yell ensures their solutions are cutting-edge, using automation and machine learning to deliver the best returns. Yell’s main purpose is to help the business to be found, trusted and chosen over the competition.

Download:
https://apps.apple.com/gb/app/yell-for-business/id1078222002#?platform=iphone
https://play.google.com/store/apps/details?id=com.yell.business

The challenge

The Problem:
Yells service for business customers is to set up a business listing online and provide as much content as possible. Generally speaking, the more content the customer provides about their business the more useful that listing becomes to users and the more likely they are to contact that business.

One way of adding content to a business profile page is via the Yell for Business app. Merchants are able to update their information in the app, view and respond to customer reviews, view analytics, upload photos and message customers, all while they're on the move. To find a way to create an easier marketplace between the merchants and their customers, we saw an opportunity to embrace conversational commerce. Using Apple Maps and Spotlight search, via Apple Chat Suggest which provides messaging functionality for the business anywhere that their telephone numbers is listed online, which could include Facebook, Google my Business and their own webpage. The Chatbot using ‘Apple Business Chat’ in collaboration with LivePerson would need to be able to handle thousands of conversations, guiding customers to the information they needed, and then finally handover to the merchant. The merchant would be able to respond via the Yell for Business app in their agreed respond time.

My role:
As a senior UX & UI Designer, I’m regularly communicating with all key stakeholders while discussing the projects with the product lead and the dev team. Together with my lead UX manager we are using various usability testing methods to produce quality research by scheduling unmoderated usability tests, surveys, card sorting and planning user interviews. Once we have the insights we build wireframes and prototypes using the Yell Design System to improve the performance and the processes. Outside of this I’m also advising on other UX and UI issues for other Yell products, working closely with SEO team, marketing team, outsource dev team and partners to put together design ideas for features ideas.

Research

Understanding the problem

Based on industry observations, users prefer to message and chat with businesses rather than speaking with them over the phone. Using chat as a term implies an instant response where users expect a clear indication of live status and expected response time, whereas messaging typically refers to conversations conducted on social media, such as via WhatsApp and FB messenger. Chatbots have the potential to help businesses significantly cut labour costs:

• 29% of customer service positions in the US could be automated through chatbots according to McKinsey. BI Intelligence estimates that this equates to $23 billion in savings from annual salaries.

• 36% of sales representative positions in the US could be automated resulting in total annual estimated savings of at least $15 billion from salaries.

But who is using messaging? The usage of messaging apps is increasing year on year. Our research implies that over 53% of costumers said they were more likely to shop with businesses they can chat with.

Messaging metrics of success

Merchant defined success

Metrics

• Increased customer ratings

• Reduction in duration of chats

Costumer defined success

Metrics

• Chat time vs Customer satisfaction

• Wait time vs Customer satisfaction

Yell defined success

Metrics

• Costumer satisfaction rate +85%

User research and personas

With access to millions of businesses that are already signed up with Yell, I was able to gain insights to create persona and empathy maps to better understand our merchants. Before I created the personas I was able to learn more through our web analytics such as Mouseflow, Adobe analytics, surveys and video calls. We were able to classify our users and run 12 separate workshops around the UK with segmented demographics using stimulus material to understand where the pain points occurred. Due to sensitivity of the data I’m unable to show our findings and our validated user personas, however, I am able to share below a persona I created before the validation phase.

Apple Business Chat and Chatbot analysis

To gain better understanding of how the Apple Business Chat worked we had a series of questions that dig deeper to provide a fuller picture of how conversations performed, what information was passed and similar. Below are presented some of the questions we were trying to get answers for:

1. How the conversations performed?

2. How the informative and instructive are they?

3. How they set up expectations with the user?

4. How quick each question was answered?

5. How and when the bot handed over to the merchant?

6. What tone of voice did they use?

7. How easy or hard was the conversation? And why was it easy or hard work?

8. How each questions lead to the next?

9. What information the bot handed over and in what format?

10. What did the bot look like?

To find answers to those questions we looked into the businesses that already were using Apple Business Chat like, Apple, Hilton, Burberry, Four Seasons, BuddyBank and T-Mobile, where we proceeded through a set of questions to evaluate how well it worked for the user.

Before we write down the findings, we looked into other conversational commerce that didn’t use Apple Business Chat such as, Facebook, Dominos, DPD, Starbucks, Santander and Volkswagen.

Overall, our findings indicated some of the businesses to come across very robotic, whereas some of them had a touch of personality by using emojis and ending the conversation with a friendly tone. A few of them ended with surveys asking the user questions on their experience.

To gain extra knowledge of chatbots we researched a couple of articles on how to approach designing and structuring a chatbots conversation. One of the articles we looked at was from Nielsen Norman Group the world leaders in research-based user experience, The User Experience of Chatbots. Based on the articles we came with the following findings:

• The most important advantage was speed.

• If you have a bot, you should have a really good one.

• Disclose upfront to their customers that they are interacting with a bot.

• When users realised they were talking to a bot, they tended to be more direct, use keywords-based language, and avoid politeness markers.

• Make it easy for user, add date pickers or reduce typing where possible.

• Carousels, the UI element that bots use for showing sets of results, are simply not the best choice for displaying long lists.

• User were generally annoyed when the bot repeated the same answers over and over again.

• Owning a failure and offering an escape hatch (phone number of a live agent) is generally perceived favourably.

• If the bot is too rudimentary, people will lose trust in the company and will feel ignored and unappreciated.

• Do chatbots have any advantages? In their current embodiment, they just have one: less information overload.

• Be upfront about using a bot and not a human.

• Clearly tell people what tasks the bot can do. Make sure you don’t create false expectations.

• Don’t be overly ambitious: create bots for simple tasks. Complexity is not well handled in the limited bot interface.

• Tolerate typos and ambiguity.

• Allow people to interact with the bot both through free-text input and a selection of links.

• Allow sorting and filtering to let people narrow down through results.

• Save information from one task to the next.

• Program some flexibility into the bot: infer context and allow people to jump forward and backward in a linear flow.

• Be honest about not understanding.

User journeys and low-fi wireframes

We had gathered plenty of information and knew essentially want we wanted to achieve. We had put together a hypothesis for how we imagined it should work and how we wanted the bot to steer the conversation, so we could meet both merchants and customers needs and goals.

For us the main objectives were:

1. Set the expectation up from the first message: be honest and tell the customers it’s a chatbot and can help with basic information.

2. Offer an ‘Advanced Bot questions’ option for merchants to set up their own questions to be asked by the bot.

3. Provide a summary at the end before handover to ensure customer information is correct for the merchant to read.

4. From the merchants end it was required that they select a response time, which could be provided to the customer at the point of handover to set an expectation of when they would be contacted.

5. Chatbot to send prompts to the merchant as the time window for their response started to close, an alert so they didn’t miss a potential lead and the app didn’t annoy the customer.

Once we had the objective set up, we made a low-fi wireframe to ensure that the product lead and the stakeholders had a clear understanding of the functionalities and designs.

Advanced Bot questions research

User stories for Advanced Bot questions

To find out more about offering the merchants ‘Advanced Bot questions’ we used the user stories method to encourage productive and user-centred discussion, so we could see the bigger picture. Some of the user stories we created are presented below:

1. As a customer who had Advanced Bot, I want to have default questions available for the Bot to ask my costumers, so that I am able to get more information from my customers without having to ask them myself.

2. As a customer who had Advanced Bot, I want to be able to write/overwrite my own questions via the Yell for Business app, so that I am in control of the questions asked by the bot and I know that they are the right questions for my business.

3. As a customer who had Advanced Bot, I want to be able to amend these questions at any time, so that they are always up to date for my customers and my business needs.

4. As a customer who had Advanced Bot, I want to be able to see the answers to these questions when the conversation is handed over to me, so that I know what the costumer wants and whether my business can help.

5. As a customer who had Advanced Bot, I want to be able add or reduce the number of questions being asked at any time, so that they are always up to date for my customer and business needs.

Messaging Customer Panel

To validate our assumptions and user stories we run a customer panel focused on gathering customer insight on Yell Messaging, about the following:

• How customers currently communicate with their customers.

• How the bot could work for their businesses.

• Value of digital assistant capable of answering basic and specific advanced questions.

A mix of industries joined the Messaging panel which we split in 2 sessions of 6 people to provide their views. The additional comments were provided during a live video call. Due to sensitivity of data I’m unable to provide the full insights, however I can present some of our findings:

• Majority of the customers find the bot as a positive time-saving tool to automate responses.

• Some of the customers were concerned if the bot could qualify leads by answering more complex questions .

• Majority of the customers thought the bot could add value by providing their catalogue, price list and bookings.

• A few of the customers who haven’t used messaging before, struggled to see the benefit of it.

Overall the customer panel went really well with majority of the customers finding the messaging platform and the advanced questions asked by the bot valuable for their business, especially if they are able to customise those questions and to be able to select when the bot responds to them (e.g., an out of hours service).

High-fi Designs

Based on the information we had collected so far we decided to design two different journeys. One journey that would take the customer down a successful path and the other down paths which weren’t as straight forward. This would enable us to understand what we were trying to achieve and what is the better user experience for our customers.

The ‘first’ journey would show the customers entering the Yell app via Apple maps and then enter into a conversation with the chatbot, which after that would be handed over to the merchant. We illustrated the journey from both merchant side of the Yell for Business app and also the customers. This journey would also help us to understand how different the conversation tone would be from chatbot to a real human being.

The ‘second’ journey would show where we believed our biggest blockers would be and how we can overcame them such as, how the bot can handle a message that it doesn’t understand and what to do when customers refuse to give a name as this was a requirement for Apple Business Chat.

The same example of design presented above was also used as a starting point for our lab testing with suitable candidates.

Designing the Hartley bot

Designing the chatbot and what kind of language tone the bot would use was a highly important task.

• Would it be male, female, non gender?

• What should the tone of voice be like? Happy? Informative? Funny?

• What should it look like? A robot? A person?

• Should it have a name? Should the name be related to Yell?

• What colour should it be?

Many questions needed to be asked with different stakeholders to ascertain its identity, but once we got the information I was able to sketch the first drawing of Hartley and after that the final design.

Usability Test

A moderated usability test was performed divided into 4 sessions of 2 hours face to face discussion groups. Two customer sessions with under 35’s and over 35’s, so we could see if age played a role in what they thought about the Messaging and Hartley bot features. The other two sessions were with small and medium size business owners for the merchant section of the app, across all ages.

To avoid bias we had hired a facilitator to work directly with the participants, guiding them through the session and asking open-ended questions. We, the UX team and the product manager were observers and did not directly interact with the participants, however we were able to observe their body language and pick up on subtle behaviours and responses.

After we collected all the data, it was time to analyse the results and make conclusions. As we went through it we pulled out the most frequent problems that user encountered. Overall the test was received very well, highlighting some areas we needed to amend or look more in-depth.

With enough information, the app features were first trialed in Manchester to a reduced number of people for testing purposes. During this period of social distancing and lockdown, businesses were looking for new ways to connect with their customers. This provided us with an opportunity to show that our solution could support them perfectly. We received great feedback from our listed businesses and customers and the service is now rolled out across the whole of the UK.

However, the work is not done yet, we are aiming to improve the Hartley bot monetisation, trying to research new features and improve the overall user experience for our merchants and customers.