If you own or manage an e-commerce website, you will know that “traffic is growing but sales are declining” is a common scenario. When that happens, rows and rows of analytics data may not be enough to fix the situation and re-optimize for growth: you need to get closer to your customers and find out what is actually going wrong.
Jeff Bronson is a business analyst at Ecommerce Warriors who helps e-commerce shops grow metrics like conversion rates, average order value (AOV), revenue, and profits. Back in 2018, one of his clients (and former Shark Tank winner) started facing the “traffic is up, sales are down” challenge, and Jeff was asked to give a fresh perspective and some actionable advice on how to solve the problem.
Jeff started with Google Analytics, one of his favorite tools for uncovering what and where things happen on a website. But in-depth analysis requires more than just numbers: it’s about watching how people interact with a website and asking for their direct feedback. Jeff knew this, and wanted to use Hotjar… though he had a reservation:
My concern was around user privacy, what is/isn’t revealed, and how that kind of data is stored. I was happy to see that Hotjar had options for suppressing things like credit card numbers, so I reassured the client that in addition to suppressing keystroke data with a single click, we could apply this same treatment to specific elements of their webpage as needed.
As he reviewed the product pages, Jeff watched Hotjar Recordings to see what was happening when people reached them. He noticed something fairly soon: “on product pages with size options, it wasn’t obvious that a size needed to be selected in order to proceed with the purchase. So the shopper kept trying to click ‘add to bag’ with nothing happening. You’ll see by the red dots, in this recording it took someone over 3 click attempts before figuring out the size-selection process.”
A frame taken from a Hotjar recording, indicating a user’s mouse movement and clicks
Jeff could also see that out-of-stock notifications showed at the wrong time, which only made the issue worse and could cause frustrated users to abandon the page. At this point, based on a handful of recordings, he could already recommend a few actions for his client:
Explore new design and color options to communicate sizing and availability more clearly
Research (and fix) the back-end logic behind ‘out of stock’ message
A/B test different variations of the size selector
Potential customers were adding items to cart and then leaving the website. Jeff could see the drop-off rates in Google Analytics, so he used a Hotjar poll on the exit pages to ask a simple question: What information is missing or would make your decision to buy easier?
It wasn’t long before, in Jeff's words, “eye-opening responses started to roll in”:
Sorting through the answers and looking at the relative word cloud, Jeff immediately identified coupons as something that needed investigating asap, since customers were evidently not receiving, finding, and using them correctly—which, in turn, definitely had a negative impact on sales.
Using this data, Jeff could confidently suggest that his client run additional research into where these coupon challenges were coming from, and test different coupon code box placements to help people complete their purchases.