Sharing is Caring

Put Cookie-Monster on a Diet.

Vincent Warmerdam koaning.io
04-25-2020

There’s an imbalance on the internet.

imgs <- c("img1.png")
knitr::include_graphics(imgs)
What it should feel like to see a website.

Figure 1: What it should feel like to see a website.

Ever since GDPR there’s been an increase in websites asking me permission to use my behavioral data. I’m open to that idea and I think websites can become better by using data appropriately. But I often do not know how my data is being used. I’m usually supplied with a notification stating that my privacy is super important after which I am told that my data will be shared with 3rd parties.

imgs <- c("img2.png")
knitr::include_graphics(imgs)
What it starts to feel like to see a website.

Figure 2: What it starts to feel like to see a website.

Sure, they’re being transparant in the sense that they are telling me about their plans. But let’s be honest here. I can’t pretend to know what is going to happen with that data. How is anyone? There’s a suggestion of control by asking me for permission but I’m still struggeling to understand what is happening when I click accept. Does the third party collect data independantly per website or is it merging everything from other platforms? What am I approving when I click “accept”?

imgs <- c("img4.png")
knitr::include_graphics(imgs)
What it also starts to feel like to see a website.

Figure 3: What it also starts to feel like to see a website.

And that bothers me. The meaning of the word “approval” no longer implies trust on the internet. It implies “we’ve covered our legal asses”.

This is tricky territory.

I’d really like to trust the website but it’s hard for me to trust some third party unless I can varify what they do with it. Then again, I also get that you’re doing yourself short if you’re not running analytics on your website. It can reduce churn if you track what clients are doing on your platform and installing google analytics is way easier than inventing in your own stack.

I’m a techie and even I wouldn’t want to run analytics myself for any of my own projects. It’s just too much gruntwork to maintain my own analytics pipeline even if it is a simple one. Then again, I really like to know how many people read this blog. I’d also like to know how my side projects are doing.

This is where the slippery slope starts. A few years back I used google analytics for this blog just for the view counts. But google gave me way more. It’s an overload of information. There’s no way to tell google to track less. I enjoy seeing where the traffic to my site comes from but I really just don’t care about demographics. It’s not just the overload of “insights” that makes google’s analytics product a pain to use. It’s also that I’m using software that tries to estimate the age of the people who visit my website.

I can’t imagine anybody giving permission for that.

Constrained Solution

So here’s my take on how to fix this mess.

I’m not going to ask for permission on behalf of third parties that you’re going to have to magically trust. Instead, I’m going to just open up my analytics dashboard. The analytics from this blog and all of my other projects are now viewable by anyone who is interested. This way you get to see what I track. The idea is that I care, and therefore I should share.

imgs <- c("img3.png")
knitr::include_graphics(imgs)
That's better.

Figure 4: That’s better.

There’s a few nice effects when you open up this way. Not only can you demonstrate that you genuinely care about privacy but by keeping the dashboard open you’re incentivised to only track data that you’re comfortable sharing to the open world. That’s a sane constraint to build into your analytics.

Plausible

It turns out that there’s actually a service for this: plausible.io. They’re a privacy aware analytics company and you can view their data policy here. All the data that is collected is aggregated and anonymised. This service is not free and I like it that way. By getting paid by people like me they have the incentive to stick to their promises. The source code of the service is even available on github so people can inspect it.

One refreshing selling point of their service is that they really only give you a few simple metrics. It actually does less than their competition. This is nice because there’s just a basic summary. I read it and after two minutes I am up to date. I am not distracted or overloaded. It is a calm experience so that I may carry on with my life.

Concluding

I’m running a couple of blogs and websites. For each of them you can now view the open dashboards;

I’m looking forward to a new norm; sharing is caring.

Appendix

And yes, plausible has an open dashboard too.

Citation

For attribution, please cite this work as

Warmerdam (2020, April 25). koaning.io: Sharing is Caring. Retrieved from https://koaning.io/posts/caring-means-sharing/

BibTeX citation

@misc{warmerdam2020sharing,
  author = {Warmerdam, Vincent},
  title = {koaning.io: Sharing is Caring},
  url = {https://koaning.io/posts/caring-means-sharing/},
  year = {2020}
}