The 100% Optimisation Score that Cost 66% of Business Growth
Last week, I sat down with the owner of a medical walk-in practice in London. He was initially quite proud of his Google Ads dashboard. He had achieved what many advertisers consider the holy grail: a 100% optimisation score. He had clicked every blue recommendation button, followed every automated suggestion, and cleared every notification bell.
But there was a problem. Despite the perfect score, the practice was quiet.
When we looked at the year-on-year data, the reality was stark. Despite a healthy increase in clicks and a perfect score from Google’s bot, the clinic had lost two-thirds of its qualified leads compared to the same period the previous year. He had traded patient growth for a digital pat on the back.
This is the hidden cost of the vanity metric trap. In high-stakes lead generation, following recommendations without understanding their economic impact is a recipe for disaster.
The broad match trap
The primary culprit in this case was the broad match trap. Google’s algorithm constantly nudges advertisers to switch from specific keywords to broad match, promising more reach and efficiency.
However, for a specialised medical practice in a competitive market like London, broad match often acts as a budget-hungry vacuum. It ignores the subtle nuances of intent that a private clinic relies on.
Patients search behaviour leading to wrong conversion data
The reality of search behaviour is that users often have a remarkably low attention span. In this instance, the broad match expansion caused the ads to appear for patients specifically searching for other clinic names or local GP surgeries. These users, often in a hurry and looking for a quick solution, simply clicked the first phone number they saw without realising they were calling the wrong practice. The client was paying premium London CPCs for calls that ended the moment the receptionist said the name of the clinic.
"Positive" signals from the wrong audience
The secondary, more dangerous effect of this shift is the negative feedback loop. When a campaign starts capturing low-quality traffic, those users often click buttons or fill in simple forms incorrectly. The algorithm sees these as successful conversions. Because the system is getting positive signals from the wrong audience, it doubles down, prioritising more of the budget toward those same low-quality searchers. Essentially, the algorithm was being trained to send more crap leads because it thought it was succeeding.
The machine was learning, but it was learning the wrong lessons. It was efficiently finding the easiest, cheapest, and most irrelevant clicks available because they satisfied the mathematical goal of the campaign, even while failing the business goal entirely.
For niche medical services, precision is a requirement, not an option. You cannot expect a general algorithm to understand your specific patient demographics or the quality of a lead just by the volume of clicks.
Lesson learned
The lesson here is simple: The more specialised your audience is, the more careful you need to be with broad match keywords in lead gen. Automation is a powerful co-pilot, but it should never be the one setting the destination. We had to go back to basics, disabling the auto-apply settings and returning to a laser-focused exact match strategy to protect the remaining budget.
If your optimisation score is perfect but your treatment rooms are empty, you are likely pleasing the bot rather than your bottom line.
Is your Google Ads account chasing vanity metrics or genuine business growth?
Book a consulting session on Fiverr or send me a message on LinkedIn.