Idle On-Site Analytics

Hey everybody,

This is a research about decentralized analytics tool for site tracking. That’s something that’s of interest to me as a marketing guy. As the times are changing, and privacy and data is becoming a commodity people are more and more aware they have and actually want to own.

I would specifically like to point out the need to implement some site tracking tools (Analytics) for Idle website. That means it’s different than the already proposed grant ([GRANT] IDLE Analytics Site ) as that is for the analytics on the blockchain. And what we’re talking about here is on-site analytics, which would help getting to know the users by having various reports that include: demographics, behavior patterns, behavior flow, goals tracking (that need to be set up first) and many more.

All this can be used in conjunction with the blockchain analytics (Dune Analytics currently being implemented as of the time of writing) - to track which type of content is successful, which sucks, what specific article is performing really well, how can other articles be improved, where the people are coming to the site from, how to pinpoint the right user which we want the Idle to grow with. And much much much more.

Why having site-analytics is important, expanded: Ex. Writing the content or any other marketing activity is arguably pointless as you don’t really know what is the ROI (return on investment) of that activity. You invest X amount of money/time/energy, and what do you get as a result? How do you measure it? How many people are reading that piece of content, are they really raading it or just skimming it, are they bouncing right off and leaving, are they sharing the article (and on which channel), is that article helping with the main goal which is expanding the number of people using Idle, what kind of content is good (“how to article”, a “listicle”, a “comparison article”, or maybe something more in depth about Idle…) and much more.

Before getting started, I want to point out that all the solutions are basically centralized. I did a research and tried to find a decentrailzed one but to no success (as it also has a possibility of a B2B cooperation and acquiring ther tokens). As the only one I found is this one https://w3a.phala.network/ (and here in the interview format: Polkadot Eco AMA 丨Phala.Network : Web3 Analytics, a New Privacy Protected Analytics Tool | by Polkabase | Medium ) but unfortunately it works on Polkadot, not Ethereum. If someone finds something plausible, great, please write below! And be careful, as usually the results when searching for the web3 analytics are blockchain analytics.

Okay, here’s the analysis of the options on the market. Mind you, this is all pretty new, so we’re basically pioneers in all of this.

The things that I considered important when conducting the research:

  • Cost

  • Privacy (the relation to the GDPR, usually directly tied to having cookies)

  • Usability of the reports and funcionality of the interface

  • The depth of data that’s collected from the users (the variety of metrics, dimensions, drilldowns, cohorts…)

  • How “fat” the code snippet is (heavy code will slow down the speed of the site → which negatively affect UX and ranking as it’s a ranking factor)

  1. The “golden standard” is Google Analytics.

Pros: Uses cookies (opportunities for remarketing, better UX etc), extensive reports, native integration with other Google tools (Google Ads, etc)

Cons: Uses cookies, Google owns the data, invasive for privacy

Digging into the code, I found out that for example https://blog.coinbase.com/ are using it.

  1. Yandex analytics

Pros: Uses cookies, extensive reports, has built-in heatmaps and user tracking.

Cons: Uses cookies, Yandex owns the data, data shared with 3rd parties, too invasive for privacy (even more than Google)


Below here we have the solutions that are sexier. Even though they are centralized, they don’t use cookies (GDPR headache is no more, but we lose the valuable more detailed data and the opportunity of remarketing to the qualified and ripe potential customers). They also enable you to own the data (as opposed to Google or Yandex). They are not as developed as Google Analytics or Yandex in regards to functionality and therefore will be less useful, but are very plausible alternatives to the big corporations. So, none use 3rd party cookies from what I’ve seen, less Javascript bloat, no data is shared with third parties, and generally a break from large corporations.


  1. Plausible Analytics https://plausible.io/

I like their Q&A session they did: Why you should remove Google Analytics from your site - DEV Community (on the same link you can read the comments to learn more about those kinds of analytics tools in general.

I personally like this option as they look professional and there’s a whole team behind it from what I can see. Code is open-source. You can also self-host the data. 30 days free, after that reasonable monthly prices: Plausible Analytics | Simple, privacy-friendly alternative to Google Analytics

  1. Goatcounter https://www.goatcounter.com/

Open-source (code availible here GitHub - zgoat/goatcounter: Easy web analytics. No tracking of personal data. )

Not sure how reliable it is, as it seems run by one person.

  1. Fathom Analytics https://usefathom.com/

Pretty similar to Plausible Analytics, but more pricey. Open source and all that.

  1. Simple Analytics

Pretty simple as the name suggest, not a lot of metrics tracked. Pros: privacy-preserving and cookie-free session tracking method. But might be a bit too simple.

  1. Matomo https://matomo.org/

Honestly, this might be the best one, as it obviously has the most money injected into it. If one were to stay for a long time (and potentially replace Google at some point in time), this might be it. Open source, data prrivacy, data ownership etc. And it’s actually free if you host it on your own server.

  1. User Track https://www.usertrack.net/

Similar to Matomo, but with the addition to Heatmap tracking which is very good speaking from experience. Also, you basically buy it one-time and have a lifetime licence. Though might collect a bit more data than Matomo, so could be problematic privacy-wise? Not sure.

.

My personal favourites are Matomo, UserTrack and Plausible Analytics.


Note:

There are also some other options out there, such as using Google Tag Manager. That’s basically a tag management system, which allows you to deply multiple tag all under one umbrella tag. Such as Facebook Pixels, mouse tracking, heat map analysis, A/B testing, various remarketing tags etc. But also enables you to track some more advanced stuff via Analytics such as engagement with the blog posts or click some specific buttons (ex. engagement can be “triggered” if a person simoultaneously scrolls more than 50% of the blog post and spends at least 30 sec on the blog post). That’s usually one I use, but in the spirit of the blockchain and making data more private, probably not the way to go for now (all depends on the communities opinion).

This is also a good article on the analytics providers I used as an inspiration and you can find some additional info and opinions on the first 4 alternatives: Privacy Preserving Google Analytics Alternatives - Yasoob Khalid

I’d love to hear what you guys think.

7 Likes

sorry for the late reply… lots to catch up here lol
ok… i am really not gonna waste my time with stuff that does not make me money, so I will leave issues related to open source analytics to others who care and know that stuff.

Now, I DO CARE about your phala network “find”. A Lot… because i see money lego potential

After reviewing the links you posted, i like this part in particular:

Looking forward to see this in production also in the ethereum blockchain soon after polkadot.

I did inquire about the possible bridge for the ETH blockchain in the Phala network community, and they said it’s planned and to wait for the announcement. So keeping an eye out!

1 Like