I moved from GA to Plausible a while back. Very nice product and easy to set up. Its a paid product but I would rather keep my user's data private.
This product seems interesting but its free so I fail to see how you will keep maintaining it in the future if you don't get paid.
I understand that running it will cost you a few dollars a month if that at the beginning but eventually, between the time spent on maintenance and the ongoing costs, you are going to have to start charging your customers.
I am doing this together with my regular job. At the very moment it does cost me about the same as the Netflix account I am paying for my parents. The plan to be profitable is by mass. There was a lot of time and thought spend on making this service really really cheap to serve. Even now with this HN my redis database is two megabytes in size and before this HN post with happy and active users the CPU did not went over 1% usage - on a cheap VPS. So the bottom line is if we can get in only a small fraction of what other services have as a fee I am happy. I think we are talking about less than one cent per user per month. In any case I feel comfortable to get that with Pay as you want and one advertising banner. Don't get me wrong. It is a great product and we do expect users to actually pay for it if they are using it on a commercial project. We just have the luxury to ask for a payment instead of charging it. I also think plausible is a great product. If you are happy with it even better :-)
Worth mentioning that the "pay what you want" business model can be a lot more profitable than people give it credit for.
My hobby photography business, which my wife and I run on a "pay what you want" model, brings more than enough to pay our monthly bills and have some leftovers for savings. While the last time I checked about 97% of photo shoots paid below average, the 3% that paid above average more than made up for it.
We also donate part of the profit of every photo shoot to a non-profit that helps women fight breast cancer, which aside from the good feeling that it give us, also helps with marketing on its own.
It is an interesting business model which not many people take advantage of, and may be perfect for Counter.dev for sure.
That's interesting. Here in Berlin there is/was also a restaurant where you could pay what you want when you leave and it seemed to also work there. I can imagine important is to explicitly state that you do have costs and are expecting payment - it's just not such a strict thing and if you are a student of course you pay less. Let's see how it goes.
Agreed. I got the idea from the Humble Bundle[1] and wondered if it could work with physical products as well, instead of digital. Turns out there are more businesses than I thought doing it, even in the city I live, although it mostly seems limited to be coffee shops and restaurants for some reason.
If nothing else, this experience showed me that people can be pretty generous if you give them the chance, which changed my outlook on life for the better.
I may try offering a plan with this model for a limited time on my hosting business, to see how things go, as an experiment. It may be a little difficult since dedicated servers are not cheap, but depending on how much each people pay, it may offset the costs and turn into a profit. Tom Morkes[2] has a book on this model that might be worth the read as well[3].
Personally, I go with no analytics whatsoever for my business, for personal reasons, but we do offer a GoatCounter to customers to entice them to move away from Google Analytics. I will definitely keep an eye on Counter.dev as an alternative to that, good luck!
I get this, and that's alright. The only reason I have tracking set up is to get an idea of how the site is doing. I should probably just record the events I need myself instead.
I find it funny how the stats get skewed by developers or people coming from HN (83% from HN at the time of viewing).
Windows, the most used OS in the world is listed under "Other" when looking at the Platforms pie chart. Pretty funny that in the HN community Linux is more commonly used than Windows, at least from the people that clicked on the link.
I do have a question... If Counter is free, what is the revenue model? How will you sustain the platform? Although it's open source, it seems to be a SaaS platform that runs on your servers.
Yes, so although it is optimized to be very cheap to operate many users there still are costs. We hope to get that via donations or as it's communicated in the website "Pay what you want". One single advertising banner is also something we are considering. So the amount we have to get from active users is really low. Currently there are 80 integrations with multiple sites on the cheapest linode VPS. The redis database is 2 megabytes on disk and the CPU does not go over 1% usage (before this HN post). If you just take our current integrations and divide them by the costs the VPS makes you get to pennies per user per month. I did not put in my calculation that not all integrations are active but you get the idea :-)
I was not expecting it to work flawlessly with my open source LMS CourseLit which I am currently building as CourseLit still has rough edges. But it did!
I was able to collect the analytics from my website and the setup was easier than that of Google Analytics. Great work.
Have you considered applying to YC? Email hn@ycombinator.com if you'd like some tips or feedback with that. (Same offer goes for anybody! Just please realize that I can't always reply quickly and sometimes take embarrassingly forever.)
Well, people that know me know that I am not arrogant but actually I'd like to counter (pun intended) that offer. My Co-Founder and me want to work one year full time on the project. Here is our mail: hey@counter.dev
Can't speak for the OC, but I'd say decision paralysis for these sorts of utilities is more about "will I regret having used this after a year vs something else" and less about "can I dump this after trying it for a week".
Competition is great, but people are people and want to know they're making a good longer term decision, which isn't always clear in the short term.
This is just something you can't help with outside of provide information, especially since use cases and desires of users of these services change over time.
"We only track the first page the user views, this is again more privacy friendly and additionally also results in substantial less HTTP requests the server has to handle. As a result of this strategy, Counter is able to show top landing pages but not top pages"
May not be big for some people, but that seems to be a very big consideration folks should be aware of.
Fun fact. Once I made a Twitter account for my open source project. Twitter says you can use the "date of birth" field for your actual date of birth or the launch date of your product. Once I entered 2 May 2018, I was suspended because now I'm too young :|
I didn't want to go through the hassle of getting my account reinstated, because Twitter isn't that important to me.
Thanks for writing. There was absolutely no activity done on that twitter account. In any case we swapped it to one that is actually active :-) https://twitter.com/NaiveTeamHQ
I don’t know. I have several Twitter accounts that are suspended for rules violations and they’ve tweeted and followed people zero times, so I don’t think you can necessarily read too much into that.
There's no real way to know for sure (unless you work for Google), but if it did affect your ranking Google would be breaking antitrust law. They've gotten in trouble in courts for doing that with other products though so it's still potentially the case, legality aside.
Oh by the way, we are looking for a Co-Founder that we are convinced is able to grow counter. We really don't care how, chould be content marketing. We are here for the long term. Write a couple of sentences to hey@counter.dev
Can someone versed in open source licenses chime in in what it means to use this service since it is licensed under the AGPL? I'm not an expert but as I understand it AGPL can require your own code to come under AGPL as well?
No, it should prevent someone else taking the code and starting a competing service with said code without opening its source.
If it was GPL, you'd be forced to open your source if you distributed your software to others - it does not apply to a web service where the software is always ran in the owner's computer. So I could take your GPL code, build a service running on my server (without ever distributing the actual software) and not open my source. AGPL tries to fix that loophole.
So if you are a user of this service, license has nothing to do with your own project, you are just a user of the service. But if you grab the source and create your own analytics (or something else) service from it then you have to open that source even if it runs on your own server(s).
> Web Analytics made simple and therefore privacy friendly
One small remark: not storing personal data doesn't mean it's privacy friendly. Yes, it can be private for the user, but not for the business owner. If we still centralize data, then potentially a single entity gets information about traffic and browsing habits of user cohorts, which can then indeed be used for "evil", such as creating a competitor of that business by using the collected data or to guide those users towards a different website/direction.
Was looking for something like this the other day. Looks perfect! trying it out right now.
Little confused how it works out the page though. Looking at the given snippet it doesn't seem to include anything and the pathname I am hitting is showing up on the dashboard as "/". Maybe missing sending "window.location.pathname"? And a super tiny thing but "navigator.sendBeacon" maybe better over "fetch". Other that super work!
The script from Counter is pretty special compare to others. It's a single fetch() call.
And for some reasons, I can't see the request being logged to the Network tab in Chrome. Maybe I'm doing something wrong? My site: https://devutils.app/
Ahh the request is only done for unique users - so if you already was there once he won't send another request again until you clear your browsing data (or just the sessionStorage)
Well, I do have Dos protection so you'r traffic would get rejected at some point in time but it would make me unhappy because I'll have to fix up the data for demo or restore the backup for it. You'll have to do a little better than just that but generally speaking I can't 100% secure myself about the demo data being skewed. Other analytics may have a better position here because they do track and mess with the requester's IP address but I don't. [Writing in my lunch time]
Nevertheless, some bad actor may take advantage of the simplicity of the implementation (which is, at the same time, also a strong feature of your product). Of course this would affect every user, not just the demo account.
I think this is one common issue with analytics products (Google Analytics itself has ridiculous amounts of spam referrers!). I guess may be a good idea to be proactive and plan some countermeasure at some point.
For every user you'd have to know all usernames which could be brute forced guessed or gathered with some sort of crawler.
It's definitely a great point you bring up I think with web analytics an attacker has a natural advantage since captchas would obviously not be feasable. But it is worth to but effort in not making it too easy.
At the first sight they really look quite similar. Counter does not use a relational database and is highly optimized to serve users for free. If you want more details and really dig into the data plausible could be a better fit. If you value a free product and don't need too much details or actually even value less data collection then I would say go with counter.
If you're collecting data, you generally want a) as much as you can get and b) be able to make use of it any way you want.
Yes users of a privacy centric analytics tool are opting into less information, though the inability to do more may be a large downside to counter (if it is the case that it has less data collection capabilities vs plausible at this time). You never know what data you will need later, better to collect however much you're able to in a fashion you can do anything you need to with it.
This product seems interesting but its free so I fail to see how you will keep maintaining it in the future if you don't get paid.
I understand that running it will cost you a few dollars a month if that at the beginning but eventually, between the time spent on maintenance and the ongoing costs, you are going to have to start charging your customers.