When Support is Underwater, Build a Submarine
When facing a ticket tsunami, hiring more people is like handing out more buckets on a sinking ship. It might keep you afloat for a while, but bailing only works when the boat is still above water. Trying to assemble a haphazard bucket brigade through too-rapid staffing in an effort to stay ahead of the deluge is futile when you're already sunk. Instead of trying to increase your capacity, focus on deeply understanding the drivers of your volume. In other words, build a submarine and map the contours of your ocean of tickets.
Categorize by the Atomic Unit
In my last article I talked about the first time Udemy hit our inflection point. The tsunami of tickets was so sudden, and we were so quickly underwater, that catastrophe was imminent. Support was drowning, pulling our customers’ trust down with us. We had 2 week response times, plummeting CSAT, and no relief on the horizon in our new normal. While hiring would of course be part of our near term strategy, it is not an immediate solution. Even an accelerated hiring and ramp up cycle, which sacrifices positive long term team growth for short term needs, would still leave multiple weeks of being underwater, sacrificing thousands of customers to bad a CX. The trick was to dive into the tickets themselves, and look for ways to release the pressure of the queue.
If you find yourself in this situation, the first step is to properly categorize your tickets by the specific issue or question. This is the atomic unit of the issue. Not “account issue” but “password reset” or “email update”. Categorizing used to be a grinding, boring task where through sheer force of effort you went through a large enough volume that you were able to identify categories that could then be manually tagged by the team on new tickets. These days, AI can make lighter work of the task.
Plug the Leak: Eliminate the Cause
With your submarine deployed and the ocean contours mapped in detail, you are now ready to ask the following question of each of your ticket categories from largest to smallest volume:
Is there any way, AT ALL, to eliminate this issue entirely?
The first time I was underwater and went through this crisis at Udemy, we already knew that refunds were our biggest ticket driver by far. When I dove more deeply, I saw this volume consisted of both full and partial refund requests. Breaking this 2nd group down to the atomic level I saw some people that bought a course before it went on sale, and many others who already had coupons, but when they’d checked out, hadn’t applied them.
At the time, the purchase flow consisted of a little blue “apply coupon” link that, when clicked, would open a field for the customer to put their code. I went to folks like Nik Laufer-Edel, Ameya Kulkarni, Mustafa Demir, Cuneyt ('Nate) Mertayak & Cansin Yildiz on the product & engineering teams with this discovered friction and in a few days they’d pushed an update with the coupon box already open and visible, rather than activated with a button. Instantly, those tickets all but vanished from the queue.
It’s important to note that just categorizing our tickets as “refunds” or even the subset of “partial refunds” would not have revealed this opportunity. We also didn’t know what we were going to find when we delved deeper into the queue. Only a dedicated exploration into the root cause of our ticket drivers, not just the general but specific issues, could reveal the defined hole in our boat that could be plugged.
Bend the Arc of Volume
These types of leaks are everywhere in startups. It’s an inevitable byproduct of the early stage product development adage: to build a car while it’s on the road, or, to reframe for my metaphor: build a boat while it’s on the water. Addressing and actually plugging these leaks is almost always via a partnership with Product, and getting Support needs prioritized by Product is the subject of half the talks and articles in the industry. Often however, and especially early in a product’s lifecycle, you can identify minor changes to the user experience that will have outsized impacts on volume, like my coupon example above. Consider hosting a 1hr brainstorming session over pizza with the Product and Eng team, where you present your big root causes and pose the open question: “what could eliminate this customer friction?” Such a simple exercise can yield huge impacts on your volume.
Of course the easy wins that eliminate significant ticket volume will run out, and I’ll discuss where to go next in a future article, but the “eliminate” step is exponentially more impactful to Support, your customer, and your company's success than any other effort you undertake. And this is not just a one-time exercise. As discussed in my last article, there is a fundamental correlation between user growth and ticket volume. Bending the arc of tickets away from the curve of growth can only be accomplished by finding ways to eliminate the root causes of customer friction. Start there.
