New on LowEndTalk? Please Register and read our Community Rules.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.
Comments
I've been using IDrive e2 for myself. Price is great. Many different edge locations all over the country so if you're close to one the performance is great.
Sorry, I need to throw bucket of ice cold water.
You cannot compete in this market.
I see that you even want to create own object storage from scratch, as MinIO is too expensive.
Lets talk about big players that are in market for decade or more.
Backblaze, Storj, iDrive and Cloudflare R2. Cheout their prices and limits. Are you sure you can offer same thing but cheaper?
What about Hetzner Storage Boxes that cost 2euro/TB and have unmetered traffic? Can you undercut them too?
You will invest a lot of money and nobody will choose your offer unless you undercut competition heavily. Can you provide reliable storage at $1.5/TB?
How big margins you will have? How much you need to overprovision to have always available space? 20%? 40%?
Selling object storage at cheap prices (or free tiers) will burn your money. Scaleway learned that (introduction of heavy retrieval fees from Glacier), Storj learned that (disabled all free tier services).
I would highly recommend to look at something other. Maybe you not only host files but also do something with them? Like video hosting, image transformations? Theres no plug&play service for Wordpress users that want to offload both photos and videos storage together with optimisation and delivery. Cloudinary or Bunny can do it, but you need to configure everything for them and its clumsy, Im talking about something that works directly from Wordpress panel.
I will use erasure coding to save on the number of copies needed. Durability will be at the rack level.
Specifically, data pieces will be distributed to drives located on different nodes, these nodes located in different racks. Along with auto-healing when an object lacks the necessary durability due to a drive containing its data fragments failing.
Unless there are multiple 9s of durability, it would be a hard sale for anyone with critical data to use a system where data is saved in same DC, different racks.
There are other providers out here which have proven and given way high level of durability, at a good price overall that wants to safeguard their important files.
Good luck!
Yes, this is a very straightforward and useful perspective for my project. Thank you for taking the time to share your views. I would like to analyze and respond to some main ideas as follows:
Regarding competitiveness in the market: I understand that this is a market with many large and long-standing competitors such as Backblaze, Storj, iDrive, Cloudflare R2. However, I believe that there is still an opportunity for a new service if it can bring different value and suit the needs of a certain customer segment. I will focus on carefully researching the market and customer needs to find my point of difference.
Regarding price and profitability: It is true that to compete, we need to have very good prices, even lower than our competitors. However, I will consider carefully between offering attractive prices and the ability to cover costs and make a profit. I will calculate and optimize factors such as infrastructure costs, bandwidth, backup storage... to have competitive prices but still ensure stable operation. The price of $1.5 / 1TB is feasible at the present time, but it will have to trade off in terms of service quality. Opening a price war is not advisable, because other competitors have nothing to do but wait for the dumping service to run out of money and collapse on its own.
Regarding service orientation: Your suggestion about not only storing but also processing data such as videos and images is very good. I have experience working in video hosting projects. Building the S3 project as a storage platform for other services such as video hosting is a suitable direction.
Lessons from other services: I will learn from the difficulties that services like Scaleway and Storj have encountered when providing too generous free packages. I will carefully consider free tier policies and prices to ensure sustainability.
Once again thank you for your very sincere and frank sharing. I will study more closely the factors you mentioned to adjust the direction accordingly. Hope to receive more valuable perspectives and suggestions from you and the community during the development of this project.
Honestly, even if you have 16 copies distributed around the world plus one on the moon, no one is going to trust you with important data because you're new and it's easy to make promises. Are you going to pay for regular audits? People will trust Google or Amazon with their data because they're credible.
You might as well run it on RAID0, call yourself YoloS3, and make it as cheap as possible. At least then you'd have a unique selling angle. Even then, people who want as cheap as possible are just running their own minio.
Really, this space is already dominated by huge players, and offering "cheaper" is rarely the path to success.
I must respectfully disagree, given that Wasabi no longer provides public CORS, or at-least on that pricing model. A more fitting comparison would be Cloudflare R2, which benefits from Cloudflare's renowned low latency and competitive pricing. Wasabi has also been known to kick people off for no reason unlike Cloudflare
.
Thank you for sharing your views frankly. These are very thoughtful comments for my project.
Regarding the issue of reputation and customer trust, I completely agree with you. As a new service, convincing customers to trust you with important data is a huge challenge. Big companies like Google and Amazon have outstanding advantages in reputation and trust. I will have to put in a lot of effort to prove my competence and reliability through service quality, security policy, SLA commitment and as you suggest, may need to undergo regulatory audits. term from a reputable organization.
Your observation about a market that is already dominated by the big guys and that going in the "cheaper" direction is unlikely to succeed is a very useful wake-up call. I will think more about how to create a truly different value beyond the price factor.
As you said, it is "rare" for a new service to be successful, not "impossible". That means I still have a chance even though the probability is not high.
I really appreciate your frank and profound comments. This is valuable information for me to review and consider adjusting the direction of the project. Once again sincerely thank you!
So... you have picked the brain of the community to polish your idea. Any thing more concrete you want to share ?
There are more and more competitors in the S3 market, previously it was the IDriveE2 with its low price policy, and this year, as far as I know, Hetzner will be taking a slice of the market - and given Hetzner's previous pricing strategy, there is reason to believe that this is excellent value for money.
I host a Minio myself, just to keep costs down. But I have to say that unless you can offer a competitive service at $2/1T or even slightly less than that, it's going to be hard to get a foothold (of course, I think you'll always have a certain number of customers, but only if you don't want to go big).
It's already a red ocean market. As a user who relies on Minio to store massive amounts of data, while I'm happy to see new players in the market, if I'm putting a lot of data into a business, I definitely want it to be reliable and sustainable. Data is expensive, data is priceless.
Compared to the players in this market, they all have deep backgrounds and credit backing. For you, how to convince new users to be able to trust you is a big hurdle, I believe, but of course if you can overcome it, I believe you might have a chance to succeed.