I don't want to nit-pick, but your homepage says it is not unconstrained:
This plan covers all the needs of most of our customers, but if your site uses more than 100 GB a month in bandwidth or more than 5 GB of space for your content and images, we'll get in touch with you to charge separately (mostly at-cost) for the extra resources.
Came here to say this aswell, I visited the site wondering what the terms and conditions were. The page starts off by saying:
> For USD $25 per month, we install and maintain a full featured Ghost install for you with no restrictions (unlimited visitors, unlimited members, unlimited authors/staff)
Claiming absolute no restrictions but then goes on to goes addd a 100 GB bandwidth limitation, which while should be able to handle mild traffic is by no means unlimited. It would be better to call the plan a one with "generous" limits but certainly not unlimited
"$25 a month for managed, Ghost optimised hosting that's perfect for 99.9% of websites*
* Think you're in the .1%? Transfer more than 100gb a month (roughly equivalent to X million visits for the average site) and we'll charge you at cost for the overage. This works out to be around $x per x. You can read more about this [here]..."
Thanks, I've removed the "unlimited visitors" part, that does seem misleading. The competition restricts that and I don't want to — if the blog owners keep their JS/CSS/Images small they can go well past millions of visitors, but I do need a bandwidth cap.
Have updated the cap to 1TB as well, realized I could move people off CloudFront to BunnyCDN if they were very high usage.
Yup. I'm offering sysadmin / ops / IT management services. If a business already has ops/devops folks they can use the Bitnami image or the Digital Ocean official image for $5 a month.
Here I provide free bits of consulting, walk-throughs of all the setup you need to do yourself, handle everything else for you. Not really targeted at experts on HN, but if non-technical friends or family need a blog+site or substack alternative I can help.
This sort of reminds me of all the comments on the Dropbox Show HN. "I can backup all my stuff by doing XYZ, why would I need this?"
I'm not saying this product is as revolutionary, just that there's value in identifying a market and taking a punt. Assuming it's not much work for the OP, it could be a great side hustle, small business, or whatever else. Just because it can be done cheaper with a bit of effort doesn't mean someone won't pay for the convenience.
"1. For a Linux user, you can already build such a system yourself quite trivially by getting an FTP account, mounting it locally with curlftpfs, and then using SVN or CVS on the mounted filesystem. From Windows or Mac, this FTP account could be accessed through built-in software."
You’ve made valid points but have failed to see who the target market is for a service like this: people who are not technical enough to stand up a Ghost server by themselves AND are too price sensitive to pay for the lowest tier of $29/mo [1] from the makers of Ghost.
Essentially, the maker thinks their product will fill an unmet need for such customers, many of whom can be found in developing countries [2].
I have been reading the comments and and I agree with you. my thoughts are as below.
1) The poster has no market plan and is building one on the go.
2) Haven't done any kind of market research.
3) The poster does not seem to have an actual platform built. seems to only have the domain at this stage.
4) Very poor presentation of the product on the site.
5) Does not actually solve the problem the market has. Does not bring anything new to the table.
I have done research on ghost hosting and there is a demand for low end ghost instances but the problem with ghost is its not profitable due to its scalability and management problems at lower scale.
All of the points are true, but to paint them negatively contradicts YC Startup School advice. This is a minimum usable product that does things that don't scale. If you email me, you'll get a Ghost install you can log in to and an invoice on Stripe. I'll walk you through everything, Collison brothers style, and give you ideas on how to set it up for your team and do low-key editing for you, early-airbnb style.
If there are over a 100 people who want this, I'll build out a self service dashboard that automates everything, and get a designer. But I do think it would be idiotic to start this project by hiring a designer and writing thousands of lines of dashboard and automation / system state management code. Doesn't mean I can't do it (have done it professionally enough times).
Regarding it not solving a problem the market has, that may be the case, in which case I'm down a landing page, will think up a new idea next month. But the world is a big place, and the market is bigger than we all think. Would be nice if we could be nicer to people - at the end of the day we're all still normal humans trying to figure out what to do after a pandemic has turned careers upside down. A simple downvote would have been enough.
HN is a tough crowd, especially for technical products that are familiar to many.
Even though it’s pretty clear who your target market is, I think technical products cause commenters to roast your product from a technical end user frame, rather than from a non-technical user frame.
Anyway, wishing you the best on your entrepreneurial journey!
I think parent meant it's too cheap? I have a license for Excel, so I can confidently say that the margins work for me, for my expenses and lifestyle. Assuming I hit at least 100 customers, of course. At 500 or 1000 customers I'd retire and write a book.
It's more a WPEngine inspired service. I can maintain margins of 80% counting only CoGS, not my time. So with about 350 customers I can live like a king (third world country), send kid to the best private school and do holidays in London / New York. 500 customers would be enough to exceed all expectations, and I manage that easily with very basic automation.
And re the contact email domain, thanks for the feedback, will fix that soon. I'm planning to do some stuff with DNS to support silent CDN switching between providers, so I don't want to put my MX records on it yet. Might have to slide it around on Cloudflare or Route53 a few times till I figure out all the details.
"Ghosting" sounds as if you will take the money and disappear...
> Ghosting is a colloquial term used to describe the practice of ceasing all communication and contact with a partner, friend, or similar individual without any apparent warning or justification and subsequently ignoring any attempts to reach out or communication made by said partner, friend, or individual.
Sorry, I just have a really bad sense of humour. When I decided to do Ghost Hosting, Ghosting was the best name I could think of that had a decent domain available.
Think ShellRack was the other option, think .com was available. Ghost In The Shell, so I give you a Shell and maintain a rack of Shells... get it? Seemed a bit of stretch to me.
Happy to take name suggestions, especially if I can find the .com for them.
Thanks for all the feedback so far. For context, I'm looking at this plan ($25/month) as a business plan, so non-technical businesses can get a blog / site / newsletter out at an affordable price.
Once I build the dashboard and fully automate blog creation, and maybe get a discount on CDN, I want to do personal blog hosting at $5 a month. Will maybe limit to 3 writers and ask you to manage your own backups. If anyone wants this plan let me know, I can set you up right now along with the businesses.
I think you should consider a three-tier pricing plan. $25 is only slightly undercutting ghost.org in dollar terms and also targets the cheapest customers (who are the biggest pain in the ass and tie up the most of your employee resources).
Chasing the bottom-feeder market is at best mildly profitable (unless you scale big), but always highly stressful and requires a very disciplined and thick-skinned team.
My perspective on this is a little different — $25 is low-end for the people who currently pay for Ghost, but my environment (India) lower rung than that. So I'm not actually targeting the bottom of the developed world, I'm targeting the top of the developing world, if that makes sense. This price point gives me a healthy margin, and it opens up publication opportunities to the worldwide Fortune-5,000,000. It allows the 20-person agencies in developing countries to set up a Ghost blog without thinking twice, both for themselves and their customers. That isn't happening with the current official Ghost pricing.
This is a brilliant proposition. Contacted you guys separately.
Most people forget the cost of human involvement, my time, and moreover, Ghost.org charge $29 /mo for a limited visit. The only time I'd consider doing something to save $25 /mo would be that I'm publishing all the static HTMLs to, say, an S3 and the worst cast of a downtime is S3 goes down (rare).
Personally, I'd still want to take over if needed but would love your $25 /mo DevOps subscription to the installation. I hope you don't start limiting to constraints such as - 1 domain per subscription! :-)
> Personally, I'd still want to take over if needed
There's obviously no way you're getting access to the servers, but Ghost has a fully import / export option already built in. I can also zip up your images / themes etc. for you, so there's no lock-in. Even the domain will continue to be your own nameservers, you'll just point your custom domain to abcd1234.ghosting.dev.
And no, I don't think I'll need to limit the number of domains that point to your installation, but of course you'll want to do that yourself for SEO reasons.
The official Ghost.org do seem high, but I empathize because it does pay the salaries of an international team. But they're so high as to be inaccessible to normal people and companies. An individual writing a blog (especially outside a developed country) or a small company wanting to start a team blog isn't going to pay that much.
Would want to give a portion of revenue to the Ghost non-profit and support development as soon as I get ramen profitable, though, which will probably be happen at 100 to 150 customers.
> But they're so high as to be inaccessible to normal people and companies. An individual writing a blog (especially outside a developed country) or a small company wanting to start a team blog isn't going to pay that much.
They'd much rather use your reasonably well priced service. \s
OK. Will you also have an option to auto-push the generated HTML and other static files to our S3? That can solve all the bandwidth issues while using your EC2/Server to just do the CMS part.
Ghost is a NodeJS app, and doesn't automatically do static site generating, but there's first class support for using other static site generators via the CMS API. That would be entirely up to the users, once you take over the installation you can either use it as an API for site builds or send people directly to it. Makes no difference to me either way.
Yeah, there's it's not world changing or anything. The problem I want to solve is that the Ghost(Pro) offering gets expensive really quickly - I want to give small teams and small businesses all over the world a way to start a small site or team blog for $25 without worrying about head count or visitor count suddenly costing them a bomb.
Then I want to give individuals a way to start writing on beautiful blogging software for $5 a month. This bit isn't possible, but will be if I can get a few hundred business customers.
IOW, you are consciously competing with Ghost(Pro)’s lowest plan $29/mo, which I have no problem with.
How do you intend to manage the actual Ghost instances: single-tenant or multi-tenant?
I’m assuming customer instances will be run single-tenant—isolated from others, which means with time, different customers will be running slightly different versions of Ghost? I’m I wrong?
That's right. The server will be multi-tenant by default, unless you're a celebrity of some sort. Ghost is super fast, and under load you're mostly stressing nginx. I start with the bigger instances on AWS, and I've load tested them pretty heavily for multi-tenant.
Even with multi-tenant, each Ghost installation is its own version, own database and own settings, theme. So they're all independently configurable. I would like to move everyone through versions in lock-step, but I have no objection to letting people trigger their own upgrades (after I have the necessary automation in place).
I’m a bit confused by the opening sentence of your second paragraph. Perhaps we are using multi-tenant to describe slightly different things.
A multi-tenant setup avoids the pain of maintaining multiple version in production. All multi-tenant customers are on the same server & db versions by design, so any future upgrades can be done for all customers at once. Also, upgrades are expected to be painless (at least in theory) since there is never more than one version in production.
Customers with special needs are hosted as single tenant and of course you get to charge them extra $$$ for this special treatment.
Replying the nested comment, my current approach is to put multiple Ghost installations on one big AWS server. Each Ghost installation has its own codebase and database, so they're very independent of each other, with upgrades and migrations all happening independently.
Dunno if you'd call that single or multi tenant, but that's the gist.
I can automate the upgrades and migrations and running them for all the installations at the same time, or leave your installation alone until you ask for an upgrade / migration.
Change the name. It's an unnecessary question mark on your brand that gets in the way of potential customers asking the questions you want them to ask. Instead of, "how do I pay?" customers ask, "ghosting...that's bad right?"
I like the Ghosting = Ghost Hosting play, and this isn't customer facing either way. You'll still put yourdomain.com over it. Doesn't seem oppressive or insulting to anyone, and I'm happy that you're more likely to remember the name now.
I don't suppose you've committed to donating at least a small portion of your profits to the Ghost open source project?
Ghost is released under the MIT License so, obviously, you don't have to, but I (personally) wouldn't "feel right" pointing any potential customers your way without something along those lines. I'd just send them straight to https://ghost.org (who are hiring folks to work on Ghost).
As soon as I get ramen profitable, yes, I'd like to. Will make a public commitment once I have enough customers to cover rent.
If I cross $250K a month I think I can commit to something like 15% voluntary royalties. And I'm happy to hire engineers and contribute to Ghost when we're not in the middle of support tickets.
And this is just common sense, I guess? If I wind up hurting the project that would be killing the goose that's laying the golden egg. I'm trying to make good software accessible to people who couldn't use it otherwise. The Fortune 500 will still use the official Ghost Pro offering, I don't have any doubts about that.
I think this is a great idea (it's one I've had myself and I might execute on someday), a few questions that I'd also think about?
- who are the biggest non-Ghost ghost hosting companies?
- how would you scale bandwidth costs?
- how will you reach this customer in particular?
- how would you scale support load? the users with the least time and money often have the most requests and need the most help
You might benefit from looking into/reading up on the microconf community, they're the kind of community of bootstrapping types that might help you get more insight on when/how/what to think about.
Thanks for the thinking points. The ones I have actually thought about so far are
* bandwidth (BunnyCDN is the best I've found, and they and AWS do private pricing after hitting an aggregate of a couple petabytes a month). Offering yearly plans would help bring in enough cashflow to buy bandwidth options. BunnyCDN allows me good margins at the retail price.
* support load: will likely put in an hourly rate for needy customers, or do the CD Baby / Derek Sivers "buy us a pizza" thing. I'm also okay with helping customers off the system or firing them outright if they're toxic. But ideally will automate whatever gets asked for the most, and offer services from a team for creative work, at hourly rates.
* customer acquisition: put my own blogs on it first (I have a personal blog, and a couple of OSS projects), bring on the companies I used to work for that avoided Ghost because we had 20 people teams and it was too expensive, then go one by one from there. Will get my mom and her embroidery boutique on, then friends with their cooking blogs. If everyone is happy, can evaluate next steps.
I'm bootstrapping as well, but the good news is there's no cost to this (my own blogs will already warrant the baseline infrastructure). So I'd like to think I can take my time with this.
I think it might make sense to at least create two tiers, one with "unlimited" support (where you charge the hourly rate and let them go crazy) and one without. Access to a knowledgebase that you update frequently (as a result of working with customers who had issues) might be available at lower tiers, but at higher tiers it comes with support?
Yeah. I’ve figured out a way to offer $5 a month, so will put that on community support, and change the business plan with human support to more than $25 if there’s a crowd.
55 comments
[ 4.6 ms ] story [ 119 ms ] threadThis plan covers all the needs of most of our customers, but if your site uses more than 100 GB a month in bandwidth or more than 5 GB of space for your content and images, we'll get in touch with you to charge separately (mostly at-cost) for the extra resources.
> For USD $25 per month, we install and maintain a full featured Ghost install for you with no restrictions (unlimited visitors, unlimited members, unlimited authors/staff)
Claiming absolute no restrictions but then goes on to goes addd a 100 GB bandwidth limitation, which while should be able to handle mild traffic is by no means unlimited. It would be better to call the plan a one with "generous" limits but certainly not unlimited
What would be a good word to convey that there are no artificial constraints, and that the only (soft) limit is the bandwidth you actually use?
I'm not even planning on enforcing the bandwidth, I think it'll even out over everybody for the most part, unless someone is being abusive.
"$25 a month for managed, Ghost optimised hosting that's perfect for 99.9% of websites*
* Think you're in the .1%? Transfer more than 100gb a month (roughly equivalent to X million visits for the average site) and we'll charge you at cost for the overage. This works out to be around $x per x. You can read more about this [here]..."
Or whatever.
Have updated the cap to 1TB as well, realized I could move people off CloudFront to BunnyCDN if they were very high usage.
For $25/m you can get a VPS from Hetzner that has 8vCPU, 16GB mem, 240GB disk, and 20TB traffic.
For $3/m on Heztner you can get a VPS with 20GB disk and 20TB traffic, which is still a 4x multiple for storage and 100x multiple for traffic.
Here I provide free bits of consulting, walk-throughs of all the setup you need to do yourself, handle everything else for you. Not really targeted at experts on HN, but if non-technical friends or family need a blog+site or substack alternative I can help.
I'm not saying this product is as revolutionary, just that there's value in identifying a market and taking a punt. Assuming it's not much work for the OP, it could be a great side hustle, small business, or whatever else. Just because it can be done cheaper with a bit of effort doesn't mean someone won't pay for the convenience.
Just my 2c!
https://news.ycombinator.com/item?id=9224
https://ghost.org/pricing/
Essentially, the maker thinks their product will fill an unmet need for such customers, many of whom can be found in developing countries [2].
[1] https://ghost.org/pricing/
[2] https://news.ycombinator.com/item?id=25415538
1) The poster has no market plan and is building one on the go.
2) Haven't done any kind of market research.
3) The poster does not seem to have an actual platform built. seems to only have the domain at this stage.
4) Very poor presentation of the product on the site.
5) Does not actually solve the problem the market has. Does not bring anything new to the table.
I have done research on ghost hosting and there is a demand for low end ghost instances but the problem with ghost is its not profitable due to its scalability and management problems at lower scale.
If there are over a 100 people who want this, I'll build out a self service dashboard that automates everything, and get a designer. But I do think it would be idiotic to start this project by hiring a designer and writing thousands of lines of dashboard and automation / system state management code. Doesn't mean I can't do it (have done it professionally enough times).
Regarding it not solving a problem the market has, that may be the case, in which case I'm down a landing page, will think up a new idea next month. But the world is a big place, and the market is bigger than we all think. Would be nice if we could be nicer to people - at the end of the day we're all still normal humans trying to figure out what to do after a pandemic has turned careers upside down. A simple downvote would have been enough.
Even though it’s pretty clear who your target market is, I think technical products cause commenters to roast your product from a technical end user frame, rather than from a non-technical user frame.
Anyway, wishing you the best on your entrepreneurial journey!
Also the domain and contact email domain being completely different is just plain unprofessional looking
> Ghosting is a colloquial term used to describe the practice of ceasing all communication and contact with a partner, friend, or similar individual without any apparent warning or justification and subsequently ignoring any attempts to reach out or communication made by said partner, friend, or individual.
https://en.wikipedia.org/wiki/Ghosting_(relationships)
Think ShellRack was the other option, think .com was available. Ghost In The Shell, so I give you a Shell and maintain a rack of Shells... get it? Seemed a bit of stretch to me.
Happy to take name suggestions, especially if I can find the .com for them.
I'm offering to remove the limit on writers and members, cost is lower, but still secondary.
Once I build the dashboard and fully automate blog creation, and maybe get a discount on CDN, I want to do personal blog hosting at $5 a month. Will maybe limit to 3 writers and ask you to manage your own backups. If anyone wants this plan let me know, I can set you up right now along with the businesses.
Chasing the bottom-feeder market is at best mildly profitable (unless you scale big), but always highly stressful and requires a very disciplined and thick-skinned team.
Most people forget the cost of human involvement, my time, and moreover, Ghost.org charge $29 /mo for a limited visit. The only time I'd consider doing something to save $25 /mo would be that I'm publishing all the static HTMLs to, say, an S3 and the worst cast of a downtime is S3 goes down (rare).
Personally, I'd still want to take over if needed but would love your $25 /mo DevOps subscription to the installation. I hope you don't start limiting to constraints such as - 1 domain per subscription! :-)
And best of luck and wishes to your initiative.
There's obviously no way you're getting access to the servers, but Ghost has a fully import / export option already built in. I can also zip up your images / themes etc. for you, so there's no lock-in. Even the domain will continue to be your own nameservers, you'll just point your custom domain to abcd1234.ghosting.dev.
And no, I don't think I'll need to limit the number of domains that point to your installation, but of course you'll want to do that yourself for SEO reasons.
The official Ghost.org do seem high, but I empathize because it does pay the salaries of an international team. But they're so high as to be inaccessible to normal people and companies. An individual writing a blog (especially outside a developed country) or a small company wanting to start a team blog isn't going to pay that much.
Would want to give a portion of revenue to the Ghost non-profit and support development as soon as I get ramen profitable, though, which will probably be happen at 100 to 150 customers.
They'd much rather use your reasonably well priced service. \s
Then I want to give individuals a way to start writing on beautiful blogging software for $5 a month. This bit isn't possible, but will be if I can get a few hundred business customers.
How do you intend to manage the actual Ghost instances: single-tenant or multi-tenant?
I’m assuming customer instances will be run single-tenant—isolated from others, which means with time, different customers will be running slightly different versions of Ghost? I’m I wrong?
Even with multi-tenant, each Ghost installation is its own version, own database and own settings, theme. So they're all independently configurable. I would like to move everyone through versions in lock-step, but I have no objection to letting people trigger their own upgrades (after I have the necessary automation in place).
A multi-tenant setup avoids the pain of maintaining multiple version in production. All multi-tenant customers are on the same server & db versions by design, so any future upgrades can be done for all customers at once. Also, upgrades are expected to be painless (at least in theory) since there is never more than one version in production.
Customers with special needs are hosted as single tenant and of course you get to charge them extra $$$ for this special treatment.
Dunno if you'd call that single or multi tenant, but that's the gist.
I can automate the upgrades and migrations and running them for all the installations at the same time, or leave your installation alone until you ask for an upgrade / migration.
Ghost is released under the MIT License so, obviously, you don't have to, but I (personally) wouldn't "feel right" pointing any potential customers your way without something along those lines. I'd just send them straight to https://ghost.org (who are hiring folks to work on Ghost).
If I cross $250K a month I think I can commit to something like 15% voluntary royalties. And I'm happy to hire engineers and contribute to Ghost when we're not in the middle of support tickets.
EDIT: I mean $250K a year.
- who are the biggest non-Ghost ghost hosting companies?
- how would you scale bandwidth costs?
- how will you reach this customer in particular?
- how would you scale support load? the users with the least time and money often have the most requests and need the most help
You might benefit from looking into/reading up on the microconf community, they're the kind of community of bootstrapping types that might help you get more insight on when/how/what to think about.
* bandwidth (BunnyCDN is the best I've found, and they and AWS do private pricing after hitting an aggregate of a couple petabytes a month). Offering yearly plans would help bring in enough cashflow to buy bandwidth options. BunnyCDN allows me good margins at the retail price.
* support load: will likely put in an hourly rate for needy customers, or do the CD Baby / Derek Sivers "buy us a pizza" thing. I'm also okay with helping customers off the system or firing them outright if they're toxic. But ideally will automate whatever gets asked for the most, and offer services from a team for creative work, at hourly rates.
* customer acquisition: put my own blogs on it first (I have a personal blog, and a couple of OSS projects), bring on the companies I used to work for that avoided Ghost because we had 20 people teams and it was too expensive, then go one by one from there. Will get my mom and her embroidery boutique on, then friends with their cooking blogs. If everyone is happy, can evaluate next steps.
I'm bootstrapping as well, but the good news is there's no cost to this (my own blogs will already warrant the baseline infrastructure). So I'd like to think I can take my time with this.
I think it might make sense to at least create two tiers, one with "unlimited" support (where you charge the hourly rate and let them go crazy) and one without. Access to a knowledgebase that you update frequently (as a result of working with customers who had issues) might be available at lower tiers, but at higher tiers it comes with support?