What small companies can afford that huge corporations can‘t: niche-ing

Photo by Climate KIC on Unsplash

Over the last couple of days, I was discussing a lot of startup ideas with my friends. There were crazy ones involving parsing twitter or Instagram in real time to present their data in a specific way. Those ideas sounded cool and doable — in fact, all the people I was talking to already had something working — but the projects were not successful, they were lacking audiences or had a tremendous bounce rate (like 90–95%).

The truth is that it doesn’t make sense to build on top of a huge corporation’s dataset — it’s hard to imagine they were never thinking about something global like that or even experimented on something already.

Big companies have millions of customers to test the value of anything quickly.

Big companies are running thousands of A/B tests and user researches to find the best ways of presenting their data and improve its monetization. They have hundreds of teams to try different directions of evolution and probably already tried all the things you could ever do with that dataset.

Of course, there’s a chance specific social network doesn’t have a specific feature you have idea about… and suddenly you decided to do a startup for it!

Is it a good idea? Probably not, even if you’ll prove it’s successful, the company could just go and implement it themselves.

That company can also:

  • just close the API for your project
  • sue your project
  • buy your company (which might not be a good thing, e. g. in a lot of cases you wouldn’t have a right to negotiate — in the end, they have much more marketing power and money to throw into a problem)

But let’s consider first, what’s the chance company didn’t think about it/it’s customers never suggested this idea? What’s the chance they never went there to understand what could be a possible market value? Maybe they just don’t have a feature because there’s no value at all?

Big companies have millions of customers to test the value of anything quickly. It doesn’t make sense to come up with ideas of showing the most popular Facebook users or parsing the whole GitHub to show more trend repositories they are allowing to see — if those features had value and impact— they would already know about it. And if that would work for you — what will prevent them from implementing that thing right inside their apps? Hope you’ve got my point.

There’s must be no sense to improve something for 10,000 users when you’re able to push a feature for 100,000,000 users and instantly get a feedback, an impact will be just incomparable.

Instead of trying to come up with another way of presenting their data and creating a new ‘holy grail’ tool try to niche, try to tailor the data to a specific audience and a specific problem. Get some other datasets to help you, mix them up.
Photo is taken from pinterest.com

An example of a problem with using the other’s dataset exclusively might be worldc.am a website. It was implementing a specific feature of looking at Instagram photos associated with a specific location. It worked more conveniently than Instagram, but in the end, Instagram just improved the interface and limited API, which made it hell harder to run a project like that.

At a time of writing, worldc.am doesn’t work at all.

Relying on the whole separate dataset also turned out to be unreliable. If it’s gone — you will have no way of getting the data, it would be hard or impossible to find the replacement.

Relying on thinking that other company would never implement a feature you’re proposing is unreliable either.

It’s hard to imagine Google or Facebook will ever focus on this topic, it would be maybe just 0.001% of their audience, really insignificant number, but as you can see it’s a really successful sustainable business by all means.

In contrast, there’s a project nomadlist.com. It’s built on top of many individual datasets and is targeting specific niche — digital nomads. It started just with tailored data and grew up becoming the biggest community of nomads and a successful business (I’d highly recommend to follow Pieter in Twitter https://twitter.com/levelsio and watch his YouTube videos).

It’s hard to imagine that Google or Facebook will ever focus on this topic, it would be maybe just 0.001% of their audience, really insignificant number, but as you can see it’s a really successful sustainable business by all means.

Mixing datasets to bring value to your customers and find new niches is another cool thing to do.


Last year, my wife and I were discussing that it would be interesting to see how Amsterdam looked like before city started to build bike lanes everywhere. I discovered Google GeoCoding API and an Amsterdam City Archive. Combining both together resulted in oldams.nl (before this project I’ve also been inspired by OldSF, which is targeting San Francisco historical photo lovers!).

Although the project is not commercial, it’s successful in terms of solving a problem nobody solved before.

Here I’d also like to introduce ‘hand-picked’ term in connection with bootstrapping projects — Google with all its power can solve data problems globally, but I can’t imagine Google picking up projects of providing historical photos of cities from all the world even having a huge collection of images in Google Images, world-class Machine Learning services and so on and be successful in providing the best possible experience to audience targeted above.

AirBnB has a great product called AirBnB Experiences where they decided not to rely on any data tour operators had and instead ‘hand-picked’ experiences from locals according to their target audience. That turned out to be successful as well.

Can you, I’ll ask again, imagine that any big company will go this way? No, they are aiming for the global fit-them-all audience. They are trying to find a solution which will fit billions of customers they already have not paid attention to niche markets, and it’s an opportunity for you to step in there.

There’re many niches in the world free to fit with a lot of room to bring the superior tailored user experience. Make your startup targeting those; you don’t have to target global audience to be successful.

P. S. Don’t get me wrong; there’s nothing wrong with big companies not filling in specific niches themselves. After all, targeting existing global market is the most efficient way of earning more money and growing the revenue for those. And that’s the thing which gives you a chance to step in.

Hi, I’m Valerii. I live and write in Amsterdam. I wrote this article, all views are my own. If you enjoyed reading it, make sure to follow me on twitter https://twitter.com/viatsko