Exactly as predicted, it took four months to collect another thousand reviews and hit that beautiful milestone: 5,000 reviews.
We did it. And I’m incredibly proud of this result and of the team behind it. In this post, I want to unpack what happened over these four months and what I’ve learned from this next chapter in our “reviews story.”
Let’s start with a bit of theory and math.
Between mid-August and mid-December 2025, we averaged about 250 reviews per month—so 1,000 reviews in four months. That number still blows my mind. What’s even more interesting: the exact same pattern showed up between 3,000 and 4,000 reviews. Four months again. That kind of consistency is very satisfying.
Roughly speaking, this means Fenster can generate around 3,000 reviews a year. If I’d had that insight back in May 2017 when we first opened, it would’ve changed a lot of my planning.
On the technical side, the new thousand reviews haven’t moved our rating yet. We’re still sitting at a 4.6 average. Subjectively, it feels like this thousand had noticeably fewer one-star ratings than the previous one—but even that wasn’t enough to shift the overall number.
My guess is that we’re seeing a kind of geometric effect over time: the larger the total number of reviews, the more you need to move the average by even 0.1. My rough forecast is that if we keep going at this pace and with this level of quality, we’ll hit 4.7 somewhere around 6,000 reviews—so approximately mid-May 2026.
A small spoiler: I’m currently working on a project to collect detailed, reliable statistics on ratings. Once that’s up and running, I’ll be able to give much more precise summaries and forecasts. I’ve decided to take this area very seriously.
Now, let’s talk about how Fenster itself has been doing over these four months.
We’ve changed more than half the team. We’re now working with a “new generation” of colleagues. Training is fully based on updated guidelines, and we already have very clear data on what works and what doesn’t. That’s a huge advantage. The fact that our review growth stayed stable during such a big transition is, to me, a strong sign that the training is on the right track. That’s encouraging.
At the same time, I’ve noticed a slight drop in the number of long, detailed, highly complimentary reviews. That tells me we need to push harder on developing consistently strong, warm, and confident service skills among the newer team members. A lot of the people who left were extremely good at delivering the kind of service that feels “very Fenster.” I really hope the new team will catch up to that level—and maybe even surpass it. I’m confident they can. It just always takes a bit of time.
Another interesting observation:
Just by looking at surnames and profiles of the people leaving reviews, you can start to see patterns in “typical” tourist behavior from different countries. You notice certain styles of ratings and text reviews repeating themselves. It’s fascinating.
In theory, you could apply different approaches to guests from different regions, based on that data. I won’t give concrete examples out of respect and ethics—but I can definitely tell you which countries tend to give more one-star ratings (and why), and which ones almost always leave fives. The same goes for the themes in the text: there’s a clear trend where, for some tourists, the main complaint is “high price,” and for others, the focus is on the atmosphere in a very positive way.
All of this makes me think that reviews and ratings are really their own separate branch of marketing—both in terms of theory and in terms of practice. It’s still a niche area, but there are already a few serious players worth watching.
Personally, I’m building everything on Fenster’s data for now. I’m not planning to “package and sell” this experience—but if the opportunity shows up one day… why not?
So, stay tuned for the next milestone: 6,000 reviews.
It might take a bit longer than four months this time—winter activity is always a bit lower—but I hope there won’t be any unpleasant surprises along the way.
And before that, I’ll do a separate post on the theories, assumptions, and Google Maps rating/review algorithms I’ve been working with so far.
Leave a Reply