How to Shoot Yourself in the Foot?

Year: 5, Month: 2, Week: 2, Day: 5

We almost had a dentistry case that turned into a live lesson on how to waste time and money.

The client came to us right before New Year 2024. We even moved the payment date to meet them halfway.

They arrived with two landing pages. Small spoiler: both didn’t convert. I also made a strategic mistake as a lead. I suggested pushing traffic from two channels at once: Yandex Direct (Yandex’s PPC platform) and VK Ads (ads platform of VK, a Russian social network). They agreed.

We launched traffic and quickly saw the same picture from two sides. We felt it. The client felt it. In Yandex Metrica (Yandex’s analytics tool) we could see a clear bottleneck on the conversion path.

The landing pages were not long. Bounce rate sometimes dropped to 30%, but more often stayed around 40–48%. That still looked like relevant traffic: some bounces (visits under ~15 seconds) were expected because the page was short. Average time on page across four content blocks was around 2–3 minutes — CARL (internet meme).

And yet… no leads.

The first tests happened in the last week of December. After the holidays, still within the first month of work, on 16 January we concluded the client needed reviews on an independent platform like ProDoctorov (Russian doctor-review site). On 23 January we asked them to find loyal patients for interviews — at least three. We reached the right line of work in time.

Interviewing loyal customers is hard. Still, with good question design you get real signal for cold-traffic messaging. These people already trust the clinic. Before they became patients, their decision path looked different.

We wrote those points in the reports on 16 and 23 January. We didn’t explicitly flag them as “important”. At that moment we naïvely assumed: if the client pays, they treat our insights as actionable and respond to requests, because those requests exist to move their result.

Only on Friday, 26 January, we got a reply in chat: “If needed, we’ll find patients for interviews.” On Saturday, 27 January, the client messaged me privately: “Where are the leads? Give me an answer.” I repeated the conclusions from the report, in detail. I explained that leads can appear tomorrow or in a month, and these interviews are exactly the way to define the shortest, most important steps on the road to the outcome everyone wants.

After that they found three patients for custdev (customer interviews). Here’s what we got from those three interviews and follow-up analysis:

1. Fifteen points to consider when launching a new dental clinic into a highly competitive market — and how to frame advantages for cold traffic.
   Example: chain clinics scare some people because they feel like a big machine that needs volume, not care. Small private clinics win here. Some people trust them structurally: smaller practice often means lower rent, which lowers the pressure to upsell an implant instead of restoring a tooth.
2. The clinic must appear on all three map platforms: 2GIS (popular maps app in RU), Yandex Maps, and Google Maps.

Yandex was quick. With Google Maps we spent three weeks trying to get the client to record a verification video. The plan was straightforward: once the clinic is present everywhere, message existing clients and ask them to leave reviews. Honest reviews. If there were rough edges, say so. Give a 5 only if you genuinely want to. People often trust 4-star reviews more — perfection can look sterile.

Then, after we had already closed the information gap with specialist agencies in medical/dentistry and spent the remaining time pushing the client to do what we had been insisting on — like Gandalf in front of the Balrog — the client chose to leave for competitors exactly on the payment date for the next month.

I believe those niche competitors will deliver. It still looks like shooting yourself in the foot. You hired a team that doesn’t rely on opinion only. It checks itself against data. The team learned the product, focused, built the concept, and you replaced them. Adaptation costs time. Time costs money. The client decided to pay that cost twice.

They’re doing fine now, and I hope it stays that way. Still, I sometimes imagine the same move in medicine: after a clinic has built a treatment plan and taken X-rays to remove an impacted wisdom tooth, the patient says on surgery day they’ll do it somewhere else — because the tooth isn’t removed yet.

Don’t do that.

Result you can apply: If the landing doesn’t convert, fix proof and the conversion path first, then scale traffic. Switching teams mid-run makes you pay for context twice.