Let's talk about the elephant in the room: The Storj economic model (node operator payout model)

IsThisOn · February 28, 2023, 9:32am

It is pretty hard to discuss stuff when we live in totally different realities.
Let me describe a hypothetical situation as an analogy.

There are 3 coffee bean companies. They are called, PremiumBean, GenericBrand and BlackRifle.

	PremiumBean	GenericBrand	BlackRifle
Quality:	They offer award-winning high-quality premium beans.	These are the generic medium-quality beans you can find in any Walmart/ASDA/Aldi.	These are the generic medium-quality beans you can find in any Walmart/ASDA/Aldi. The quality of the packaging is not very good.
Reputation:	They are an old trusted brand and thus known by customers.	They are a cheap alternative brand and thus known by customers.	There are pretty new and small, customer knowledge is pretty limited.
price per 1kg:	100$	21$	20$
volume:	100t	100t	1t

PremiumBean and GenericBrand have in common, that they only process their own coffee beans. They own the farmland, employ their own farmers, and roast the beans themselves.
BlackRifle does it differently. There are independent farmers that farm and roast their beans and sell them to BlackRifle. BlackRifle packages the roasted Beans and sells the finished product.

Currently, BlackRifle suffers multiple problems.

Problem 1: They only sell 1t a month. That is nothing compared to the other companies

Problem 2: Because of that very low volume, BlackRifle only buys very small amounts from farmers. New farmers can only sell a fraction of the coffee they produce.

Problem 3: Farmers don’t have a contract with BlackRifle. They have no idea how much they will get paid in the future per kg, nor do they know how many BlackRifle will buy from them. That makes planning for farmers very difficult and investing a gamble.

Problem 4: The packaging is of poor quality. Lots of buyers hesitate to buy BlackRifle because of that, despite the slightly lower pricing than GenericBrand.

Problem 5: BlackRifle loses money with every pack they sell. They currently pay the farmers 20$ per kg and need another 10$ for the packaging. They only can do that because of VC money.

Discussing these problems with other farmers in the BlackRifle forum is difficult, because we can’t agree on where we stand at the moment. IF you believe in my understanding of the current situation, some arguments from the other farmers almost seem delusional.

Farmer 1: It is unsustainable for us farmers! We need higher prices than 20$ per kg from BlackRifle.
Me: That does not make any sense. They are already selling it at a loss and now you want even more?

Farmer 2: It is getting unsustainable for me. I need wood for my roasting machine and wood prices have risen.
Me: You shouldn’t have bought wood in the first place but only use unused wood you get from your land. If you invest money by buying wood, you have to compete with the big guys. They use big solar installations, there is no way you can get costs down to their level on a bigger scale.

Farmer 3: BlackRifle undersells themselves. They should spike prices and compete with PremiumBean.
Me: We aren’t even really competing with GenericBrands on a quality level. You always claim that BlackRifle is on the same quality level as PremiumBeans but that is just wishful thinking and not backed by any award.

Farmer 4: Damn, that coffee-selling business is hard. The margins are way too low! We should sell Nespresso like coffee machines that only work with BlackRifle. Then we can sell pads at way higher prices.
Me: Before the BlackRifle factory workers start building coffee machines, can they please get the packaging done right first?

BlackRifle PR 1: Blackrifle is 80% cheaper than PremiumBean!
Me: Sure, but that is only true if you buy a single pack, which nobody does. If you buy it in a multipack, prices go way down. Also, you can’t really compare the two, PremiumBean plays in a completely different league. My suggestion would be, ditch the military look of the packaging. You scare away more people than you attract by that.
Farmer 5: But you do know, that the US military is our largest customer?
BlackRifle PR 2: Actually that is not true. US military only makes up only for a fraction of our sales.
Me: Then I even understand less why you would not ditch the military look!

Farmer 6: Sure times are currently rough, but just wait until the new US military budget comes out!
Me: What makes you think, that the budget will be higher and not lower?
What if it is already at its peak?
And even if it gets bigger, what makes you think, that the military needs coffee?
And even if they need coffee, why should they buy it from us and not from PremiumBrand or GenericBrand?

BlackRifle employee 2: Dear farmers, please don’t invest in any new tools! We will shortly have some adjustments and lower prices!
Farmer 7: I think that it is wrong from BlackRifle to discourage farmers from investing! If we don’t invest, we will never be able to sell 100t a month
Me: We currently can’t even sell more than 1t a month! Let’s try to sell our overproduction first before we ramp up production even more.

In case you need help for the analogy:
PremiumBean = AWS
GenericBrand = Wasabi, Backblaze, Ionos
BlackRifle = STORJ
farmer = node operator
packaging = S3 compatibility
Nespresso = Photo or Backup solutions
US military = Web3 adoption

BrightSilence · February 28, 2023, 11:51am

Analogies are great at a high level, but tend to break down at the details. You represent BlackRifle as being universally worse than anything else out there. While this is probably true for some use cases for Storj, there are many use cases in which Storj outperforms the big guys even without considering the price difference. And this isn’t from Storj PR, but from some of their actual customers. For Storj, that’s mostly large file distribution use cases where latency isn’t important. (Eg. software distribution, large scientific dataset distribution, blockchain snapshot distribution, video distribution not streaming) But also (less profitable) backup use cases. Coffee beans simply can’t represent a complex product like data storage in a reasonable way. Storj has the option to lean in to their niche while they improve the product for other use cases to expand into. I’m not sure what would be analogous to that for your coffee bean analogy. But pricing your product to fit the worst use cases is not the way to build a brand. Finding your niche and pricing it accordingly is. Then you expand from there.

Coffee bean farmers have other options, if BlackRifle isn’t profitable for them, they will disappear. And there is no brand to scale if you run out of suppliers. It’s a balancing act and yes, node operators probably have it too good right now, but you can’t just cut one side when you try to optimize. You’d need a holistic approach to optimize processes and make it fit the market on both ends.

IsThisOn · February 28, 2023, 1:17pm

Analogies personally help me put things into perspective. Especially when my belief systems, religion, prejudge, or something else secretly clouds my judgment.

No, I represent BlackRifle as being worse when it comes to quality, packaging, and marketing.

Tell me about these use cases and how it outperforms the big guys.

So we both agree, that BlackRifle is not competing on a quality level with PremiumBean? Don’t say that too loud on the BlackRifle forum

I would put it this way: Coffee bean farmers survived until now because of subsidies. When these subsidies go away, the kg price goes down. Some farmers will disappear because of that. BlackRifle CEO already made an announcement, that in fact will lower the kg price! Now, if BlackRifle cuts down the price tomorrow, or step by step over the next years doesn’t matter that much. What matters is if this is a viable business model or not. BlackRifle is not a new startup, they had 8 years to prove that point.

BrightSilence · February 28, 2023, 1:31pm

Which are literally the only properties you mentioned in your overview…

Keep reading…

How is high speeds across the globe for large file delivery not quality?

I don’t know about your fictional BlackRifle, but Storj encourages these conversations. I have no fear mentioning Storj’s downsides here. Time to first byte being probably the biggest one.

I don’t disagree with anything in that summary. I mostly disagree with how you skip over the upsides of Storj and argue it should be bargain bin priced. I don’t think that fits the product on offer. And even you still mention that farmers will leave. That’s exactly the balancing act I was pointing to. You either drop the subsidies and see whether node operators can survive it or you find your niche, price accordingly and take less risk on losing node operators. Of course in the case of Storj there are other things to tweak too, like tuning RS settings to require lower expansion ratio’s, making repair cheaper, charging for costly services (like the gateway MT) separately. Consider for example $5 for egress when using native implementation or gateway ST and charging an additional $3 for ingress and egress for usage of the gateway MT, as this incurs egress costs for the Storj hosted gateway MT in both directions. This would cover the costs and incentivize customers to use the cheaper native implementation or a self hosted gateway ST.

All I’m saying is I really hope we don’t see just a “we’re gonna pay SNO’s less” announcement without other efforts to make the whole system more profitable.

IsThisOn · February 28, 2023, 1:55pm

No, there is also a very small price advantage

Large files and high latency is not PremiumBeans quality. That is GenericBrand quality.

I honestly don’t know how BlackRifle should beat the competition. I just think that BlackRifle does not exist in a vacuum and there are other brands. So it has to compete with them.

I just see that their current quality is at best on par with GenericBrand, the pricing is slightly cheaper, and the marketing poor.

Please, just give me a use case! Come on, give me a real example. And please don’t say

This is the point where we go round in circles over and over again. I simply believe this is not true. I believe that there are only two types of coffee customers. Type A just wants the highest quality. They don’t care about prices. Type B is not concerned about quality. This type is more interested in price, that the packaging does not fall apart, and that his visitors don’t make fun of him because he has a BlackRifle coffee bag in his kitchen.

You might disagree with me on that. But the fact that GenericBrand sells 100x the amount of BlackRifle is a strong indicator that coffee customers agree with me.

IsThisOn · February 28, 2023, 2:27pm

Another interesting side note:

During Q4 2022 BlackRifle coffee used
1.2M to pay farmers for beans
2.7M to pay coffee service providers
3.5M to pay BlackRifle employees
22.5M for the repurchase of company shares and other payments

So I am with you on that one. I really hope for the farmers that BlackRifle won’t cut the kg price too much. Because as you can see, beans from farmers only account for a fraction of the total costs.

It really makes me wonder if BlackRifles company structure just got way to inefficient or if some people are making a lot money on the back of VC and farmers. It seems that with such a high spending, BlackRifle would have a lower cost of doing business by just going to Walmart, buy GenericBrand coffee and resell it as BlackRifle coffee

BrightSilence · February 28, 2023, 2:47pm

This is why your analogy doesn’t work. Storj is capable of delivering higher and more consistent speeds globally for large files than even the giants like AWS, GCP and Azure are capable of. Furthermore it’s multi-region by default and at no extra cost. It’s nothing like the price competitors in this space. You keep consistently ignoring Storj’s strong points.

I think you can use the search function yourself as well. There are multiple lauding customer testimonials in software distribution, large scientific dataset distribution and blockchain snapshots. You can look at some case studies here: Storj Case Studies search the forum or this new thing called the world wide web for further testimonials. I don’t have them at hand at the moment either.

Imagine how little beliefs matter… You expect me to back up my stance, but rely on your belief. You can really stop with the coffee analogy, because it breaks down. Quality isn’t a single measure in the storage space. There are use cases that only care about low latency regional access and use cases that care about high speed, latency independent global distribution. Customers have specific needs and Storj can meet the needs for many use cases incredibly well, while not best suited for others. It just isn’t a simple quality vs price comparison.

No it isn’t, it could be just a good indicator of how well known the brand is.

They won’t be after scaling up. BlackRifle is growing and if they fix unit economics, that means their profit will outgrow the currently higher static costs that won’t grow linearly with scale.

node_operator0815 · February 28, 2023, 3:25pm

I think it is about time that I also write down my thoughts. I’ve been a node operator since May 2021 and have really thought a lot of the Storj project so far.
Recently I became a Storj DCS customer and I am not really satisfied with the performance. My upload is maxed out, but my download is almost not at all and I think this is due to, please forgive me for this expression, third world countries that provide maybe 56 KBit/s for upload.
I’m happy if I can use my 250 MBit/s downstream at peak for one second, but on average we’re talking about maybe 50-70 MBit/s here. That’s clearly too slow for my 2.2 TiB backup.

My suggestion: Run speed tests in both directions on your test satellites and then categorize the nodes into tiers. If you want premium speed in a region of the world where it is safe - I’m talking here for myself and Germany, for example - you should dig deeper into your pockets.

Those who want less and, above all, inconsistent speed will pay less accordingly. And since electricity prices have already been discussed in other threads. New customers in Germany pay at least 0.40-0.65 EUR per kW/h, which is no longer unrealistic, so that you would scare away your most valuable SNOs.

You could adjust your payment for storage node operators accordingly and not generally punish those who consistently and securely provide storage for the network.

BrightSilence · February 28, 2023, 3:33pm

I recommend taking a look at this topic: Hotrodding Decentralized Storage

You don’t have to worry about slow nodes as they usually get eliminated by long tail cancellation. Storj is also not as fast for smaller files. But you can overcome a lot of that by downloading more of them in parallel using uplink. The topic should provide you with plenty of info on how to optimize.

Knowledge · February 28, 2023, 3:43pm

As BrightSilence points out, your settings may not be optimal. Slow transfers are trimmed off and there is less incentive for those nodes to even exist as they will lose the race to much faster nodes who will earn egress. The system is fetching a percentage of the fastest available files, discarding the rest, and reconstructing the ones it doesn’t need. By leveraging increased parallelism you can achieve very high transfer numbers.

MattJE96011 · February 28, 2023, 4:30pm

Storj claims to be competing with other S3 providers that clearly charge far more at around $23/TB and varying rates for bandwidth which are substantially higher. If this is truly the use case they want to focus on and that’s what Storj is best for then fine, but why is Storj so cheap compared to the competition? It seems Storj is priced to compete with a specific part of the market that they don’t actually cater to. This also hurts them in terms of the market they DO cater to because although it’s far cheaper it also clearly indicates the company isn’t making a profit and therefore isn’t sustainable long term. There’s to much uncertainty around it for these larger customers to make the switch despite the cost savings.

From running nodes myself for years, they tend to average around $4/TB. More TB’s cost money for hardware but more bandwidth doesn’t (with obvious limitations of course). I would still be happy with $4/TB even if my bandwidth is maxed out, especially on the outbound side since generally that’s not used for a whole lot anyway in residential situations. Since Storj is a decentralized network that takes advantage of many internet connections with otherwise idle bandwidth, why not monopolize on that. Really use that to your advantage.

Since node operators (with some acceptions I’m sure) generally don’t have any additional costs for outgoing bandwidth, why not offer that up as your selling point? Use the real advantages of the network to sell it. Let snos contribute more by taking advantage of something we already have available to give away as opposed to cutting payouts.

So what about this…
Price Storj at say $15/TB
Offer a large FREE egress allowance. Enough to WOW me but not so much that it can be overly abused.
Keep $7/TB egress for anything over free allowance. (Free allowance can vary by customer.)
Pay snos flat rate of $4/TB + egress bandwidth (if any is paid).
3 copies of data = $12 with $3/TB left for Storj (I believe it’s 3, is that correct?)

This alone, with the current data stored (assuming you could just hike up existing customers pricing) would generate $648,000/yr for Storj… and the networks still small yet.

This would generate plenty of income for Storj especially at scale.
No more subsidizing.
This would give me as a potential customer far more confidence in the sustainability of the company.
Customers in the targeted market still save ~$7/TB right off the top plus all that bandwidth cost.
Node operators have steady and predictable earnings and no longer have to worry about low egress.
Theres room to slowly reduce costs over time:
- Storj can decrease their cut as they scale as well as afford “bonuses” to encourage quicker growth.
- Sno payouts can be cut over time relative to decreasing data costs and increased storage density etc. (But don’t forget about increased energy costs.)

Someone please tell me why this wouldn’t still be a highly competitive as well as profitable strategy for the particular market Storj is trying to break into…

BrightSilence · February 28, 2023, 5:08pm

I like the concept, but I think those prices are too high in the current market. Storj is still trying to break through and low prices are a very good way to do that. Backblaze offers $5/TB with $10/TB egress costs. Storj could match the $5 for storage and keep SNOs at the current $1.50/TB. The expansion factor is around 2.5x on average I believe. So $3.75 would go to SNOs, Storj keeps $1.25. Though I believe they have a revenue share of 20% with channel partners… or something in that range. So that would leave only $0.25 for Storj if a channel partner brought them on board. Wasabi offers (slow) egress for free up to the amount of data stored per month. Storj could possibly replicate that and only pay SNOs for egress that is also paid for by customers. Or calculate a reasonable flat egress rate for SNOs.

The problem remains that they have to cover egress costs for their gateway MT. Which isn’t cheap. I think an added charge there is quite reasonable, since customers have the option to avoid that added charge by using the native uplink or self hosted gateway ST.

But this only works if there are enough large egress cases so there is some money to be made on that.

Realistically speaking though… Node operators do have it quite good right now. At current pricing you could make $750 profit in 5 years on an 8TB HDD of $180 after subtracting initial purchase cost and ridiculous energy costs of $0.66/kWh. I am shooting myself in the foot by saying this, but that’s kind of a ridiculous ROI. So they can afford to drop payouts a bit.

8TB HDD
Purchase cost: $180
Energy costs for 5 years: $250
Estimated revenue with current payouts over 5 years: $1180
Profit: $750
Cost as part of revenue: 36%

20TB HDD
Purchase cost: $350
Energy costs for 5 years: $250
Estimated revenue with current payouts over 5 years: $2400
Profit: $1800
Cost as part of revenue: 25%

But if they cut payouts in half, that would change ROI significantly as the costs becomes a MUCH larger part of revenue. 72% in the first example, 50% in the second. That’s already a very limited profit potential. And sure, most HDD’s will probably last beyond 5 years, but I picked that as that is generally the timeframe they can reliably be used in data center applications.

So yeah, there is some margin to play with… but it’s not as big as it might initially seem.

Larger HDDs will help with that though.

MattJE96011 · February 28, 2023, 5:57pm

I agree that payouts have a little room to drop if they had to, but that still doesn’t leave anything for Storj. And comparing Storj to Backblaze and other cheap options isn’t really fair as they are not the same products. But to be honest, node operators should have it good right now. This is how you encourage network growth. The network is still small. 50 PB is peanuts. And I understand super low introductory pricing, but it appears this 4/7 pricing is THE price. Storj prices itself right down there with the absolute cheapest options.

When I buy any product personally, as a general rule of thumb I buy around the mid price point for most things. I find that’s a good balance between buying something that will underperform, and overpaying simply for the brand name. When I buy a product for a company though, this tends to shift to somewhere between mid to higher end depending on what you actually gain from the price increase. In terms of an auto repair shop I used to own I buy MAC and SnapOn tools as opposed to craftsman. They are stupid overpriced so why? Mainly the lifetime warranty, AND they come right to the shop every week. If I need things replaced or something new it’s right there.

Now… I don’t know how many of you are familiar with Harbor Freight tools but those who are will know right away what I’m getting at here. For those who don’t, Harbor Freight is the super chepo discount tool brand. Ok for some things and hobby DIY’ers but everything is make out of chinesium. One time use type crap that’s practically guaranteed to break. I don’t even buy this stuff for my garage at home. So… if I happen to be walking down the street and see some new tool place, stick my head in and see prices on par with Harbor Freight, I’m not going to give the place a second look. The salesguy could sit there telling me their tools are comparable to SnapOn and MAC until he’s blue in the face and I’m just going to laugh at him… maybe he’s telling the truth, but I have no way of knowing that. Maybe they offer lifetime warranties… but I don’t know if they’ll be around next month to honor them. I’m sure you can see where I’m going with this. I’m sure everyones heard the term “you get what you pay for”. In the case of Storj, it may not be true, but that concept will still run through peoples minds.

IsThisOn · February 28, 2023, 7:00pm

BlackRifle coffee is just better than PremiumBeans! Source: trust me bro. /s

Even if I believe you for a second, you don’t have to convince me. You have to convince the customers. There are currently not convinced.
Source: Low sales numbers of BlackRifle

Haha no. I went down that rabbit hole once just to find out about a niche crypto project that shares the ledger rollup with a 150GB STORJ link. There are testimonials and there are testimonials. John Miller buying 1kg every month from BlackRifle is not the same testimonial as ASDA announcing buying 1t every month. I am too lazy to research that further. And I don’t have to. The source is again: Low sales numbers

No matter what the reason is, this is one of the most important metrics. And it is pretty bad.

Sure, but that scaling up has to happen before they run out of ICO money.

BrightSilence · February 28, 2023, 7:20pm

Yeah, that is a risk. And it seems to have @IsThisOn convinced too🤣

I really don’t care if you do. It seems you are already convinced and I’ve learned to stop arguing when people show that that is the case.

As for adoption… Just take a look at the graph I posted here: SNO's please don't buy anything - #34 by BrightSilence

mars_9t · February 28, 2023, 10:54pm

Managed to read the whole thread in less than 24h. I’ll give myself a clap for that - mostly because I endured all the filler posts, like in the anime.

Since Backblaze is coming up over and over again, could someone please tell me why* their products are not your competition?

If they are really not, but the Amazons’ are, why do you set price at what Backblaze has in their offer? Like if a BlackRifle decided to compete with PremiumBeans but at the prices lower than what Herbee has for their teas. Parallel/interweaved market, sure, but that’s not your competition, so why set the prices based on that?

*ain’t nothin’ but a heartache

BrightSilence · February 28, 2023, 11:17pm

You have my respect. I’m not sure I could have done that if I wasn’t around since the start. Also, welcome to the community!

I would say it’s not a matter of better or worse in general. Traditional storage providers tend to operate in a fairly similar way, so when comparing those you can sometime rank them on a kind of general quality basis. Storj is just a bit of a different beast, which is in some aspects better than the big guys and in some aspects worse than the price competitors. I would personally argue that that should put eventual prices somewhere inbetween the two. But because it’s kind of a new quantity, I can also see why you may want to market it as a lowest price offer at first. Just to take pricing away as an argument not to try it.

In short. Storj is great at high throughput for large files globally. But it isn’t good at time to first byte and many small files. So that means it’s great for high speed global large file distribution like software, data sets, blockchain snapshots or backups including database backups. It’s not so great (yet?) at streaming video, photo services or any other thing where latency before the first byte comes in is important. And in those extremes it outperforms the best competitors on one end and underperforms the price competitors on the other end. Does that make sense?

mars_9t · February 28, 2023, 11:44pm

Is Backblaze good for that as well?

Is Backblaze?

Isn’t this cold storage or at least close to one? And reps already stated multiple times here that Storj is not meant for that. So this won’t be a “selling point”, so it won’t contribute to scaling factor.

I know I am going Backblaze, Backblaze, Blah-blahze, but the price is directly competing with them, so I want to know exactly why that’s the case, despite stating multiple times that they offer different product(s). I know that Storj has built-in geo-redundancy and overall resiliency is quite high, but that’s not what it can do from the use-case/API perspective.

Hello

MattJE96011 · February 28, 2023, 11:45pm

They are still competition, but they are not the same. One major drawback to Backblaze is your data only exists in one location… in one datacenter. If you want to replicate it to another, guess what… your paying for it. Storj replicates the data globally. Large geographic areas of the network could go out and files would still be accessable. This might not matter to some, but does to many. I can’t speak to data speeds etc as I’ve never used them. I know other cheap options can tend to be slower though.

mars_9t · February 28, 2023, 11:51pm

So functionally Storj is the same/similar, but its ace is built-in geo-redundancy?

So Storj should have price a bit higher, somewhere betwen 1-1.5x of Backblaze, clearly stating that if you want geo-redundancy there, you need to basically pay 2x. If there are other benefits, the price should go even higher then.