Comment by nickdandakis
7 years ago
To everyone complaining about the price point for this service.
You are part of The Problem.
This is a solo dev's venture, that has a relatively pure and straightforward goal. If you can't afford it, don't use it and pick one of the others.
Do NOT compare this with a B2C offering that has nothing to do with analytics.
Do NOT compare this with a B2B offering that's free and feeds your user's data into the parent corporation's advertising revenue stream.
Do NOT compare this with a B2B offering that is open-source, with a team of a dozen core contributors that has had a decade of development under its belt.
Heh...I had the opposite reaction to the price. As someone building something in the analytics space, $12/mo seems so low that it won't get traction beyond the hobbyist demographic. If you want to sell to business, the price point needs to be at least $200/mo.
Plus, I have zero confidence that someone using a naive postgres implementation can scale an analytics backend with customers paying only $12/mo unless all those customers get barely any traffic. Perhaps if he was using Timescale on top of postgres, but even then, $12/mo seems awfully low.
But as it is, the price point signals that he doesn't think it's a particularly valuable service.
How do you know the postgres implementation is naive? I've worked on several analytics platforms...including offshoots of google analytics within Google itself, and this problem domain is ridiculously easy to shard on natural partitions. And after sharding, you can start to do roll-ups, which Google Analytics does internally.
By 2014 when I left, we had a few petabytes of analytics data for a very small but high traffic set of customers. Could we query all of that at once within a reasonable online SLA? No. We partitioned and sharded the data easily and only queried the partitions we needed.
If I were to do this now and didn't need near real-time (what is real-time?) I'd use sqlite. Otherwise I'ld use trickle-n-flip on postgres or mysql. There are literally 10+ year-old books[1] on this wrt RDBMS.
And yes, even with 2000 clients reaching billions of requests per day, only the top few stressed the system. The rest is long tail.
1. https://www.amazon.com/Data-Warehousing-Handbook-Rob-Mattiso...
There's a comment elsewhere in this thread where he talks about his backend. He didn't explicitly say it was naive, but he definitely gave off that vibe. Is it possible to use postgres in a sophisticated way to work as an analytics store? Sure...Timescale does it and gives you the majority of what you'd need. But it's hard to get right and the creator hasn't given the impression that he's well-versed in this space.
It's an MVP right? Nothing stopping him from changing the model in the future
Incrementing counters for pageviews, visited page, referer and page width, and putting that into chartjs is something I can put together myself in two hours. It wouldn't be nearly as polished, but it would be 90% there and good enough. Plus I would have a much better idea how well it scales, and generally have less unknowns and risks.
The goal is great, the design is sleek, but at the current price point (which is already lowered to $9) and feature set, it's just not worth it to me. For that price the tool has to provide more actionable data or other value.
There's probably a market out there, but most of that market is probably not the type of person you will find on hacker news.
There's always one.
https://news.ycombinator.com/item?id=9224
Dropbox is a glorified FTP client, Slack is IRC with a nice skin and inline pictures, Spotify is bittorrent without the hassle of downloads (and more legal, but that hasn't stopped people). Convinience matters, and convinience sells. But it is only one of many factors that play into the decision to purchase. Dropbox for example wouldn't sell all that great without the free tier.
2 replies →
I made an argument about this offering vs other offerings, not this offering vs rolling your own, which is a whole different issue.
It's incredible how many developers undervalue their own time, effort, and liability. I believe you're mistaken if you think this'll only take two hours of your time. Even if that's true, I believe you're mistaken that your two hours of time is cheaper than $9/mo. I'm also certain that being responsible for analytics sets you up for liabilities and maintenance that distracts from your main value proposition.
There's definitely a market for this, and that market is absolutely here, but apparently a lot of developers don't know how to pick and choose their battles.
> I made an argument about this offering vs other offerings
No you didn't. You argued what it shouldn't be compared to, so I compared it to something else.
> I'm also certain that being responsible for analytics sets you up for liabilities
Outsourcing analytics opens me to the same or worse legal liabilities.
> Even if that's true, I believe you're mistaken that your two hours of time is cheaper than $9/mo.
I don't live in Silicon Valley, so $9/mo pays for two hours within a few months. Sure, there's hosting and an uncertain maintenance burden, but on the other hand buying a service has its own uncertain overheads.
What happens if the service is down, what if it fails to scale, what if it gets hacked? What if it just disappears because there weren't enough customers? Most of these are much easier to answer and take up less time with a self-built service.
2 replies →