Why the best researchers have empty desksBy Gareth Dunlop
I have long admired the old–fashioned manager who enjoys wearing the shoe leather. There is something very comforting about being at a hotel or restaurant, and seeing the owner or manager meeting customers, chatting with the staff, checking the cleanliness of the cutlery and crockery as well as the quality of the food and making sure that everything is just ‘so’ for customers.
Unfortunately, the sight of such managers is increasingly rare in today’s commodity–driven retail environment, but historically it was common for the manager of a clothes shop or shoe shop to tread the boards to ensure that customers were getting the attention they deserved. Occasionally they would be gainfully employed to close a deal or agree a final negotiation, and the best of them would use their customer interactions to make sure that customers were being well looked after and happy.
When one considers the not inconsiderable pressure on managers’ time, it is impressive that they forgo the comfort of their office for life on the front line, to see what’s really going on. Managers who still walk the shop–floor are wired a certain way. They embrace an unspoken belief that the health of their business can’t merely be described by balance sheet and P+L reports, but must be felt on the shop–floor, by observing and understanding customer interactions, on the coal–face, where the action is.
I can’t help but wonder if such managers have much to teach those of us who make their living by better understanding digital customers. And I wonder if we can be too tempted to use the myriad tools available to us (such as online surveys, card sorting and heat–mapping), all of which provide us with valuable insight, but don’t put us physically in contact with the customer in their natural environment.
Ultimately, they can leave us one step removed from where the real action is.
We recently did some work with the largest e–commerce company in Benelux. They are one of the few companies in the world who have been able to stop Amazon becoming dominant in their territory. They effectively are Amazon in the region, offering a virtually identical service, selling directly, and allowing vendors to sell through their platform for a commission on sales.
Many of these vendors sell thousands or tens–of–thousands of product lines, and for such vendors, neatly ordered spreadsheets and rigorously named matching image files are essential for efficiently managing their stores. Images related to their products can be uploaded individually or in batch, depending on the needs and size of the vendors.
The e–commerce company keeps a close eye on their support logs from both phone and email and noticed that there were regular queries from vendors about individual image uploads, but not from multiple image uploads. It was clear that individual image upload process needed further work to make it easier to use, however that was balanced with a sense of relief that the multiple–image uploads appeared to be working well.
The e–commerce research team decided to travel to meet a range of their vendors (large, mid–sized and small) to see how they could improve the single–image upload experience. They didn’t express it, but in their minds they wondered if they could use some elements of the multi–image uploader to improve the single–image uploader.
At their first vendor, they spent some time getting to know the business, looked around the warehouse and had a good conversation about their marketplace, challenges and opportunities. As they got down to business, it was clear where some of the single–image upload issues arose, and the researchers took ample notes around pain points and some ideas about how they could be eased. As the conversation drew to a conclusion they couldn’t help but ask “are there any elements of the multi–image uploader which we should incorporate into the single–image uploader?”
They couldn’t have been more surprised at the answer.
“Oh, we’ve never looked at the multi–image uploader. We’ve clicked on it a few times, but it was so complex looking that we didn’t go any further.”
Could it be that the reason the multi–image uploader didn’t generate any support issues is because no–one was using it?
So, the researchers asked the next vendor. And the next one. And the one after that. Consistently, they learned that no–one dared go near the multi–image uploader because it was so difficult to use.
This led them to the inevitable (and accurate) conclusion. The lack of support calls to do with the multi–image uploader wasn’t to do with its ease of use: rather it was because no one was using it. How they perceived user behaviour, and actual user behaviour, were polar opposites.
They learned a valuable lesson through the exercise. The hallowed environs of the design studio are artificially sanitised. Reports, analytics, and research tools – valuable as they are – can only tell us so much.
There is always value in embracing an old–school approach and researching where the action is.