Tending — From Almost No Online Sales to a Scalable Purchase Flow
Zero → One
Hands-on
Situation
Tending originally operated in Israel, where cemetery care services were sold mainly by phone. This worked because customers understood the process, local traditions were established, and people were comfortable discussing the service directly.
When the company entered the US market, the same model stopped working. Phone sales were expensive, difficult to scale, and poorly matched local customer behaviour. At the same time, there was almost no effective online purchase journey.
The first digital flow asked users to choose a monument type or size, then enter cemetery and grave details, select a service, register, and pay. In theory, this looked logical. In practice, it created friction because customers often did not know which monument type or size they had.
When I joined, online sales were effectively zero.
Task
My task was to help build a scalable online sales journey and make the service understandable enough for customers to purchase without needing specialist knowledge.
My formal title was Head of Design, but in practice I worked as a hands-on product design lead. I personally handled most of the research, analysis, UX, UI, and experiment design, working directly with the founder, developers, and marketing.
Action
We first tried to improve the existing flow in the usual ways. We tested different images, different step structures, and considered whether the process should be shown on one page or split into several steps.
However, I started to suspect that the real problem was not the interface. It was the underlying logic.
I collected completed cases and grouped them by monument type and size. A large number of examples were difficult to classify. I then asked the CEO specialist and internal managers to help validate the categories. They also struggled to identify which type or size many monuments belonged to.
That became the key insight: we were asking customers to make a decision that even people inside the business could not make consistently.
Instead of continuing to improve the same broken logic, we explored a different model. Together with the founder, we replaced technical monument selection with three service packages based on price and included services.
We later tested a simpler version with only two packages, but it did not produce a meaningful improvement, so we kept the stronger structure.
As the product developed, we also saw that many customers wanted care for several graves or monuments. We added quantity selection, but this made the calculator and the pre-payment form too complex. To reduce friction, we moved cemetery selection until after registration and payment.
We also redesigned the role of phone communication. Instead of using calls as an expensive acquisition channel, the call manager contacted customers after purchase to confirm details and identify additional needs.
Results
The first online flow with monument selection generated only around ten orders per month.
After replacing it with clearer service packages and improving the wider journey, online sales grew quickly. The business operated with this model for around six months and reached approximately 100–150 online orders per month.
Traffic and advertising improved during the same period, so I would not attribute the full growth to the design change alone. However, the package model removed a major decision barrier and created a much stronger foundation for online sales.
The post-purchase call also became a valuable upselling channel. In one case, a customer who initially ordered care for one monument added eight more during the follow-up conversation.
Learnings
The most important lesson was that the visible UX problem was not the real product problem.
We initially thought the flow needed better explanations and better interface design. Research showed that the customer was being asked to make a specialist classification that did not belong in the journey at all.
I was mistaken in thinking that business requirements were something fundamental and unchanging. When the owners saw that the sales model wasn't working, they easily changed it. And that showed me once again that I need to think beyond the scope of design.