How does Carrd make profit?
So this might be a weird question but after using Carrd for a bit, I've tried to do some number crunching because it seems like the company would not be profitable.
I have two thoughts. First is that hosting 100 websites for $19/year would barely cover the underlying hosting costs of the ec2 instance/droplet/etc. Thats assuming every person gets their own instance (which I could be wrong about). One year of the smallest AWS instance is ~$8 per year which makes a margin of $11/customer. My followup thought is that it seems like there are new templates that come out every so often and the labor required to make those would cut into the slim margins that already exist. Finally, the cost to acquire a customer would be at least non-zero and thus would diminish the margin further.
Anyone else curious about how this is all working?
His costs to acquire customers is 0.
The websites are hosted statically I believe.
Most websites get close to 0 traffic and cost nearly nothing to host.
It is a one man shop so costs are very low.
I believe BLM recently used carrd a lot and might have drove up server costs a lot.
I agree, they probably have some jam type of set up where you don't really need a lot of compute.
Plus, most sites I bet don't get any traffic, so the costs could be very low.
Many venture funded businesses in emerging industries focus on market penetration before profitability.TThis could very much be the case for them.
There are also strategic paths they can take once they have enough mass: domain registrations are one of the oldest businesses online, plus integrations can become paid (like you have in Spotify). In pretty sure they already make money on affiliate commissions to stripe and other services.
Point is: their cost could be much lower than you anticipated and their revenue streams may not be that clear to understand only by looking at their pricing page
cc @ajlkn
Oof I did not know he was on here. Thanks for tagging him @csallen!
one of the best podcasts ever on IH
There is a IH pod with him as well :) https://www.indiehackers.com/podcast/087-aj-of-carrd
follow him on twitter too, his updates are quite interesting
As these are all static websites and mostly drive little to no traffic you can easily host a few hundred or even thousand on a single droplet.
I used the 5$ DO droplet and managed to get about 100 req/sec with a PHP site, I'm pretty sure you could serve at least 1000 req/sec on these small droplets.
I think AJ has a few bigger instances (AFAIK with IBM and AWS) where he can easily host all the websites.
Ahh makes sense!
Operating costs for low traffic sites are cheap and do not require instancing. All of the blogs on https://bearblog.dev (which I created) operate on a single Heroku dyno hobby tier ($7 per month) with a Hobby dev postgres DB ($9 per month).
I currently have over 1000 blogs running on Bear Blog, and my server is currently running at about 25% capacity. I assume Carrd works in a similar way. I do not believe that each website is a separate instance, but is just doing fancy stuff with domains and subdomains.
I hope this helps :)
I asked AJ the exact same thing on twitter with my guesses:
and he replied with a
"pretty close"Wow thats awesome you got him to respond!
Just wanna let you know that many of their premade templates are also available in https://pixelarity.com/ .Not sure if they are correlated.
Yeah it's not that much profitable and hectic. I used to run a similar service. You need lots of customer support because majority of your customers are users who never opened a single html in editor and neither they have idea how to setup a simple page.
I had 2000 plus customer but profit was not good and it was taking too much of time. So I sold it.
Since he's not marketing large scale to non-tech audiences, his support might be significantly lower.
My first job was at a webhosting company (started in tech support) - you're absolutely right that non-tech customers need a lot of support!
I think he nailed his niche, and that helps a lot.
Hey, Carrd is popular now. This is what I have learned after running similar service:
Name or Branding matters a lot in link service. Mine was too (I can't reveal name because the guy who bought my site is now pretending to be founder of that site).
Short customer lifetime. Most of time user need link for just 1 day or for a week. Mostly they need during an event. So basically they just get their work done in trail period.
Need higher number of customer support as compared to other SaaS service.
Getting banned in some conservative countries for using your link service for redirecting fans to x-rated service or using for social events which triggers governments.
The service is profitable if you have nice, short and addictive domain name just like Carrd or LinkTree which is very hard and costly to get these days.
Thats what my gut feeling was but maybe hes hit an inflection point where its worth it?
I have replied in details above.
Large, old hosting companies are hosting often even 15,000 websites on one dedicated server.
That's with Cpanel, a lot of unoptimized stuff, and demand for some processing power.
Afaik Carrd does not need even a fraction of what Cpanel and PHP website needs.
The cost of bulk hosting small websites with the same tech goes into pennies/year.
I see so at scale the economics works out
Exactly. And access to the cloud allows you to start small. With AWS free $100K credits, it's an amazing help.
Every person most likely doesn't get their own instance (this isn't enterprise software). Rather, it's probably load balanced and centralized across multiple large instances. That way, costs are fixed and margins remain high.
These are likely fully static sites, when hosted in s3 they will cost next to nothing to host & serve. There is lots of profit to be made at $19 a year.
The important thing that you missed is - "Most sites using carrd have a almost less than 10 users per day". (There may be execpetions)
With that low traffic, it is profitable.
As @rosiesherry said, most people wouldn't care to cancel Carrd and so the churn is will be low.
I see the volume needs makes sense.
It's also a no brainer to just renew every year at that price, I bet his churn is super low.
Whereas when someone is paying $30+ per month for something they will always be thinking about whether they need to renew it.
I didn't even think about that! Now that you mention it before switching I often asked myself whether by webflow was worth paying so much just to host landing pages.
When you subscribe to Carrd, you're paying for an indie hacker to work on a project he's passionate about, full-time, not the hosting.
Speaking from experience, these kind of businesses work great when they're run by a single owner-operator, who can build and support the entire product from top to bottom.
Static sites can be incredibly cheap to host, usually an order of magnitude cheaper than the cheapest DO droplet and a wordpress site.
If you need to do it yourself, try something like BunnyCdn or Backblaze b2 as your low-cost storage bucket, and put that behind the Cloudflare free plan, and crank up your cache-control headers for images.
Especially if you have something like a Carrd site, that only gets a tiny amount of traffic. Note that for a static website, 50k visits per month is a tiny website.
Your total costs will range from a couple of cents per month to 50¢ or so. I imagine the unit economics of this business works out to a nickel a site per month, or less.