Alpha version
Organization
B2C retail, brick-and-mortar store selling custom-made, high-end bicycles, 4 FTE (± 12 employees)
The challenge
People from all over the world managed to find the boutique, high-end bicycle store. Multiple days per week, employees found themselves prioritizing store visitors over customer questions from other channels, as there was not enough time to help everyone.
The result of this was that callers often needed several attempts to get through, many people that email needed to wait more than a day, questions on social channels were often an afterthought. On top of that, the store owner was spending time after closing hours answering emails on a daily basis.
How might the store be able to answer more questions in an efficient way, while saving time?
Roles
- Me: UX researcher
- Research project collaborators: store owner, web content editor
Constraints
- Budget: The store owner only allowed ad-hoc research with limited budget when the project leads to actionable and practical recommendations that increase the efficiency of the company.
- Resources: No ability to involve other personnel other than the web content editor. The store owner had limited time for the project.
My approach
Hypothesis
The customers of the store are people that like to self-organize their foreign bicycle holiday, using a high-end custom-made bicycle. Marketing theory tells us that risk reduction becomes more important when consumers are shopping for expensive products. Moreover, the complexity of the holiday, the complexity of bicycle customization, and the cost of the purchase could lead to many questions. Therefore, one of my hypotheses was that the store might be frequently answering many common questions.
From earlier research, I learned that the store had blind spots in their understanding of the buyers’ journey. My additional hypotheses were, that a selection of common questions might be the result of unclear steps in the customer journey. And another selection of common questions might be the result of not adequately answering these questions on the website.
To research these hypotheses, I took inventory of customer questions, and where in the customer journey these questions pop up.
Scoping the research
The store received questions from store visitors, and was contacted many times per day by telephone, email, and social media. Taking inventory of questions during store visitors would be very time consuming. Given the constraints, in-store techniques were no option.
Since the store employees already felt overwhelmed by the number of customer questions, it was not an option to assess the phone questions. Questions on social media were very infrequent, in low numbers, and often not related to the primary business. Therefore, I didn’t include questions from social media channels.
Email was the channel customers used most often for asking questions. A benefit of email is that it’s already documented, which makes analysis easier. Also, the fact that many customer questions often turn into email conversations made it easier to understand the background of people’s questions. So, I scoped my analysis to customer questions that came in by email.
I opted for a content analysis, combined with quantification to help with prioritization of questions.
Phase 3: Content analysis of e-mail conversations
To make analysis easier, I collected the contents of all relevant email conversations into a single Word document.
Selection criteria:
- Email conversations during the off-season and the high season
- Email conversations of 4 emails back-and-forth or more, because long conversations might indicate inefficient communication.
I made a tagging taxonomy based on:
- Tags known beforehand (phase of customer journey, product/service, customer’s region)
- Tags that depended on the email contents (topic, feelings)
Together with the web editor I tagged the emails. Next, I did a descriptive statistical analysis to understand which topics had the most questions.
Then I did a qualitative analysis of the conversations, to understand how some conversations ended up becoming longer than the average conversation length.
Result
Things I have learned from analyzing 379 emails:
- It became very apparent to me that all people were contacting the store because they felt unsure about something. The emails all showed people were looking for clues, feeling doubts, needed help deciding, didn’t know what to do next.
- During the orientation phase, a big part of the email conversations was spent on planning an appointment.
- Most questions came from people in a phase of the customer journey where they were waiting for their custom-made bicycle to be ready for pick-up. This phase could take weeks or even months. In this phase, customers were planning an appointment to receive their bike, asking when the bike was ready, and asking about making changes to the bike.
- Foreign customers had long email conversations, but their conversations were very efficient: They had clear questions and made clear decisions.
- The website appeared to not answer several common questions, questions a bicycle store might expect to be asked.
- The store had several existing initiatives that could have been leveraged to reduce the number of emails. For instance, the store was sending standardized informational emails after an online appointment has taken place, or after customers ordered a bicycle. These emails were not adequate in answering some of the common questions.
Impact
I identified 19 points for improvement.
- The store owner decided he would work on 12 of the 19 points, mostly related to changes of work procedures.
- I picked up 7 points related to the website.
Some of the decisions were:
- Communications policy – When email conversations lead to making an appointment, employees should consistently point people to the online appointment process.
- Created new webpages about topics that the website had not covered, yet.
Reflection
- If I was given more room to do research, I would also look into which questions people ask on the phone and via social media, since other channels might lead to additional insights.
- Initially, I tried to tag the emails with atomic research in mind. I learned from this attempt that I don’t yet understand how to apply atomic research. I will need to look into that some other time.