Thanks for posting, Rosie! I was literally thinking of doing this with our analytics.
Tracking revenue, platform sign ups, newsletter sign ups, and SEO performance definitely seem like decent metrics to follow. Also, using Similar Web for rough traffic trends would give a reasonable overview every month or two.
I like his take on why he's removed GA. I think if you want vanity metrics such as page views, then you're probably better off with server side analytics. Digital Ocean has this built in, and you can add CloudFlare analytics as another option.
GA is way more powerful than just page views though so you're probably not using it to it's full extent anyway. My personal take for removing GA was more for my reader's privacy. Instead, I chose JetPack analytics for my WP blog, and Matomo for custom built websites which is hosted on my own DO droplet. This way, I have more control, and I'm not recording any data that isn't needed.
I'm definitely on the record here as pro-GA, but this move can make sense if you are particularly sensitive to website performance and at small scale.
For certain sizes of 'small,' it's less accurate to say "you don't need a powerful analytics tool" than it is to say "you don't need a powerful analytics tool yet." If you are still validating your idea, and still have few enough customers that you have the bandwidth to talk to each of them, then focus on that qualitative data over quantitative info from a small number of users that skew easily.
Add analytics back in when you have sufficient customers that you can't manage to talk to them all, so that way you have to see aggregate usage information to understand whether your users value your latest improvements.
this is interesting, indeed? how would one measure performance of paid advertising, conversion rates, retargeting and finding issues in the customer journey?
But I think that's when it matters the most. I don't do a ton of analytics but I do look at # of visitors, channels, what pages they're viewing, and ranking keywords.
Yes that’s the consensus, I just am not sure I agree with it so I’m experimenting without. I may be incorrect, but my theory is that sign ups, newsletter subs, google search console data for rankings, etc is enough data and Google Analytics data is an unnecessary distraction
I don't know that there's a right or wrong answer honestly. It's what you value and like. If all I were interested in is growing our mobile app, then I would do the same. I'm also wanting to test our Popular Bluetooth Devices Marketplace concept and so I need to grow our web presence as well. So, for sure, I need to know if my efforts are paying off with increased visits.
I would love to migrate away from GA. But I’d hate to lose 13 years of metrics. Even just page visits and MAU. Are there any tools that can import something from GA? Or a recommended manual approach?
On my site GetTheAudience I use Matomo analytics (open source, self-hosted version). It allows you to connect it to your GA account and imports (slowly slowly) your existing GA data into its database. It’s slow because it has to respect GA’s cap limits for API calls per day. But this doesn’t really matter for me, once the import is over, it’s done!
That's not going to be the main problem. You can export the data or use core reporting API to migrate the data to your new setup.
I'm experimenting with manual approach with internal analytics, retrieving unique page visits isn't a problem as request headers have what's necessary but designing, implementing and maintaining a proper database would be the key problem to solve.
To start, just to store the Country names, we need a table with 195 columns now extrapolate it with states, cities etc. Projects like https://github.com/stefangabos/world_countries can help to some extent but still replace GA with internal analytics is a huge endeavor.
Interesting! I do wonder if they would consider other web analytics platforms in the future. It's nice to see what other providers have to offer (especially the privacy minded ones)
Analytics combined with https://www.fullstory.com/ is a great combination for early startups or services. I monitor problems and instantly respond to my customer needs with an email.
I visit analytics multiple times a day so that I can be available when my customers might need me.
This guy is famous ?
Because I don't see anything interesting in this article ...he removed google analytics because he doesn't need it, nothing more, nothing less
Thanks for posting, Rosie! I was literally thinking of doing this with our analytics.
Tracking revenue, platform sign ups, newsletter sign ups, and SEO performance definitely seem like decent metrics to follow. Also, using Similar Web for rough traffic trends would give a reasonable overview every month or two.
Reading this has made up my mind!
I like his take on why he's removed GA. I think if you want vanity metrics such as page views, then you're probably better off with server side analytics. Digital Ocean has this built in, and you can add CloudFlare analytics as another option.
GA is way more powerful than just page views though so you're probably not using it to it's full extent anyway. My personal take for removing GA was more for my reader's privacy. Instead, I chose JetPack analytics for my WP blog, and Matomo for custom built websites which is hosted on my own DO droplet. This way, I have more control, and I'm not recording any data that isn't needed.
This is an interesting thought but I'm not quite convinced.
I'd suggest these simple steps to make GA productive and not addictive:
i don't use them at all on anything.
I'm definitely on the record here as pro-GA, but this move can make sense if you are particularly sensitive to website performance and at small scale.
For certain sizes of 'small,' it's less accurate to say "you don't need a powerful analytics tool" than it is to say "you don't need a powerful analytics tool yet." If you are still validating your idea, and still have few enough customers that you have the bandwidth to talk to each of them, then focus on that qualitative data over quantitative info from a small number of users that skew easily.
Add analytics back in when you have sufficient customers that you can't manage to talk to them all, so that way you have to see aggregate usage information to understand whether your users value your latest improvements.
Nice move! 👏
this is interesting, indeed? how would one measure performance of paid advertising, conversion rates, retargeting and finding issues in the customer journey?
I'm guessing the point is that you don't as not everything needs to be measured, maybe @Ramy can confirm?
Pretty much, especially in the early days when the numbers are so tiny anyway
But I think that's when it matters the most. I don't do a ton of analytics but I do look at # of visitors, channels, what pages they're viewing, and ranking keywords.
Yes that’s the consensus, I just am not sure I agree with it so I’m experimenting without. I may be incorrect, but my theory is that sign ups, newsletter subs, google search console data for rankings, etc is enough data and Google Analytics data is an unnecessary distraction
I don't know that there's a right or wrong answer honestly. It's what you value and like. If all I were interested in is growing our mobile app, then I would do the same. I'm also wanting to test our Popular Bluetooth Devices Marketplace concept and so I need to grow our web presence as well. So, for sure, I need to know if my efforts are paying off with increased visits.
I would love to migrate away from GA. But I’d hate to lose 13 years of metrics. Even just page visits and MAU. Are there any tools that can import something from GA? Or a recommended manual approach?
On my site GetTheAudience I use Matomo analytics (open source, self-hosted version). It allows you to connect it to your GA account and imports (slowly slowly) your existing GA data into its database. It’s slow because it has to respect GA’s cap limits for API calls per day. But this doesn’t really matter for me, once the import is over, it’s done!
That's not going to be the main problem. You can export the data or use core reporting API to migrate the data to your new setup.
I'm experimenting with manual approach with internal analytics, retrieving unique page visits isn't a problem as request headers have what's necessary but designing, implementing and maintaining a proper database would be the key problem to solve.
To start, just to store the Country names, we need a table with 195 columns now extrapolate it with states, cities etc. Projects like https://github.com/stefangabos/world_countries can help to some extent but still replace GA with internal analytics is a huge endeavor.
Interesting! I do wonder if they would consider other web analytics platforms in the future. It's nice to see what other providers have to offer (especially the privacy minded ones)
cc @Ramy 😇
Thanks for posting this 😊
Don't follow this unless you are really sure.
Analytics combined with https://www.fullstory.com/ is a great combination for early startups or services. I monitor problems and instantly respond to my customer needs with an email.
I visit analytics multiple times a day so that I can be available when my customers might need me.
It does waste time but, totally worth it!
This guy is famous ?
Because I don't see anything interesting in this article ...he removed google analytics because he doesn't need it, nothing more, nothing less