Helping the network become profitable — what can operators contribute while staying viable? (calculator inside)

Product Discussions STORJLINGS
1

If you haven’t already, please read the open letter to the community (An open letter to the Storj token community: the restructuring, the network, and a proposed path to shared ownership) first. That thread is the place for questions about the restructuring, the token, and the shared-ownership proposal — this thread is narrower.

I work at Storj, and I’m asking a practical question on behalf of a decision we may have to make together:

How far could storage node payouts be reduced, before operators are forced to shut down?

I can’t answer that with just one datapoint from my nodes. I need to know your economics — and those vary enormously between a Pi at home on cheap power and a rack in a datacenter. So instead of guessing, I’d like to gather real datapoints.

A number you can share without exposing your setup

The trick is to normalize: your monthly running cost ÷ your current monthly payout = the share of today’s rates at which you break even. Expressed that way, a small and a large operator produce directly comparable numbers, and nothing about your node count, locations, or costs has to leave your hands.

To make that easy I built a small calculator. It runs entirely in your browser — nothing you type is sent anywhere — and it outputs a copy-ready result with only percentages and $/TB, no costs or counts:

→ Payout-rate calculator

Note on the math: I’m treating today’s Reed-Solomon numbers as fixed on purpose — they’re tuned for the balance between durability, storage overhead, and repair traffic that keeps download performance where customers need it, and that balance is worth protecting. Keeping RS constant means the ratio between data at rest and egress/repair traffic stays constant too — so any payout reduction has to apply proportionally to both storage and egress. Cutting egress alone would force the RS numbers to change, spike repair traffic, and degrade the download performance customers rely on. Baseline is today’s $1.50/TB storage, $2.00/TB egress.

The calculator groups the result into four tiers, from deepest cut to shallowest:

  • Survival — covers base running costs only; temporary, since nothing is set aside to replace a failed drive.
  • Barely sustainable — also covers replacing failed drives, so the fleet doesn’t shrink; but nothing left for growth or your time.
  • Slow growth — adds a small budget to expand capacity over time or just a small profit for your own pocket.
  • Moderate growth — adds a larger expansion budget.

What I’m asking of you

Run your own numbers and share just the result

How I am going to read your numbers is as follows:

  • The cut you can tolerate indefinitely would be somewhere between slow and moderate growth.
  • The deeper cut you could tolerate temporarily would be somewhere between survival and barely sustainable. Feel free to add for how many months you would be able to tolerate it.
  • Optional context you might want to add

Higher-cost setups matter as much as cheap ones — more, honestly. The goal is honest data to inform a hard decision, not a race to the bottom.

I’ll post my own numbers as one datapoint in a reply below — with the operator hat firmly on, not as any kind of official position.

2

Switching hats: everything below is me as an operator, not Storj speaking. It’s one setup, and a deliberately efficient one, so I fully expect my tolerances to sit at the forgiving end. Please don’t read my numbers as a target — the point of gathering many is that yours will look different.

My Storj node payout tiers — rate cut I could absorb (setup private):
• Survival: −67% → $0.50 / $0.66 per TB
• Barely sustainable: −46% → $0.81 / $1.07 per TB
• Slow growth: −26% → $1.12 / $1.49 per TB
• Moderate growth: −8% → $1.38 / $1.84 per TB

Two things about my setup that make these numbers what they are:

My running costs are low. Thanks to hashstore I was able to retire an old, power-hungry setup and replace it with something far more efficient — power is my biggest cost, so that pulled my whole floor down. An operator who hasn’t made that jump, or who pays more per kWh, will hit their limit at a much smaller cut. Weight my numbers accordingly.

My nodes are close to full. I count anything above 80% utilisation as full, and I’m at 75% now — close enough that scaling up is a real near-term decision, not a hypothetical. That’s lucky, because it means the growth tiers aren’t abstract for me: I know what full looks like, so I can estimate what adding to it would cost and earn.

Here’s how I think about each tier:

Survival (−67%) covers base running costs only — power, internet, connectivity. Nothing set aside for hardware. It works right up until a drive dies, and then that capacity is gone for good. This is a 3–6 month emergency bridge, not a way to live — I’d only accept it on the understanding that rates recover to at least the sustainable tier afterward.

Barely sustainable (−46%) adds a reserve to replace failed drives. The fleet stops shrinking and holds indefinitely — but there’s nothing left for growth or for my time.

Slow growth (−26%) adds a small budget for new drives on top of replacement. Concretely: I could still add about one drive a year. My idea of standing up a new node with four drives would stretch over a few years instead of months — but because replacement is already covered, it still happens. Slow, but real.

Moderate growth (−8%) is where expansion gets comfortable. Drive prices are a bit crazy right now, so I’m waiting for a good offer around €15/TB — but on a sensible cadence I’d buy 2 drives late this year and 2 early next year, which is a full new 4-drive node. Four drives in a single year felt like too much; the 2-and-2 split feels fair. That’s what “moderate” means for me. Anything above moderate feels like too much — sure, I’d take it, but there’s no need to overpay me. Moderate is already a solid incentive to keep running and scale up when the network needs it.

One honest caveat, operator to operator: Storj’s own advice has always been to run on hardware you already have, not to buy gear for the network — and my talk of buying drives sits in tension with that. So to be clear, the growth tiers aren’t a promise to expand. I might buy those drives; I might just as easily put the money toward a completely different hobby. I framed the margin as drives because “what could I add with this?” is far easier to answer concretely than the fuzzy “how much profit do I want in my pocket?” Treat the growth tiers as a yardstick for discretionary margin, not an expansion plan.

So let me turn that into a concrete offer, rather than leave it as a table of tiers:

Opening offer: −15% to -20%, permanent. That still leaves enough margin for my hobby, and I’d hope it’s meaningful cost saving on Storj’s side. If a modest trim is all that’s needed, this is where I’d happily land.

If more is needed: −50% for up to 6 months, then settling at −10% to %15 long term. I’ll take a deep temporary hit to help through a genuine crunch — but the trade is that the permanent rate then lands gentler than my opening offer, not deeper. I’d give up more now precisely so the long-run cut can stay small. That still sounds fair to me.

That’s my honest read of my own setup. What I’d really like now is yours — especially from operators whose costs look nothing like mine. Run the calculator and drop your tiers below.

3

I will have to add - asking your operators to “break even” is asking them to work for free.

I will just leave this here:

4

Yea not great wording on that part now that you are pointing it out. I must have missed that. The issue with break even is that we all have a different understanding which of the 4 tiers we call break even. For me it would be the slow growth tier that I would call break even.

5

Anything less than we got now, will make this not worth it for me. And I honestly think people fool themselves if they accept less than we get now.

I think the math needs to include:
Taxes,
Time spent
Hardware running cost
Power
Hardware replacement cost
Internet costs.
And yes - make the dumb comment that I don’t understand the project @ rabbit.

I don’t care. Any reduction and I have to bail.
The margins ARE so thin - and in this market, they might now even exist.
I say that as a 6 year operator, with custom scripts, earning predictions and mapped out earnings VS expenses. I have done the math serveral times over. My 450TB JUST break even. I can’t do less.

6

My Storj node payout tiers — rate cut I could absorb (setup private):
• Survival: −3% → $1.46 / $1.95 per TB
• Barely sustainable: +11% → $1.66 / $2.22 per TB
• Slow growth: +38% → $2.07 / $2.76 per TB
• Moderate growth: +65% → $2.47 / $3.30 per TB
(− = reduction from today’s rates · storage $/TB · egress $/TB · scaled together per RS balance)

That’s very optimistic. A few degrees warmer outside, and the numbers will change.

7

It does. For my own calculation I have included electricity, internet connection (difference to the cheapest contract I would have without my storj nodes) and some other recuring costs.

Hardware replacement I put down for my calculations as well. I am lucky to have no HDD failure for a long time now but one day they will break. In my case I would expect about 1 disk per year to break. Plus a bit extra for other hardware that can die along the way.

Time spent wouldn’t work for me. I am not tracking how much time I am spending. Instead I just add another drive per year as a slow growth target and if I don’t expand that will be 300€ for my pocket as a compensation for my time. Feels easier to calculate for me.

Taxes I don’t pay on expanses. total payout - total costs and that gets taxed.

8

My Storj node payout tiers — rate cut I could absorb (setup private):
• Survival: −33% → $1.00 / $1.33 per TB
• Barely sustainable: −10% → $1.35 / $1.80 per TB
(− = reduction from today’s rates · storage $/TB · egress $/TB · scaled together per RS balance)

I want $1.5/TB storage and $0/TB for egress or repair
I just need my disk to be filled.

Real question is HOW MUCH THE CUSTOMER PRICES CAN GO DOWN:
What we need is customer price from
$7/TB and $7/TB egress to some maybe $3/TB and $0/TB

So they actually use the network.

9

$0 per egress would mean storj can reduce costs by lowering the RS numbers to a point where you would see nonstop repair traffic. I would estimate total egress traffic would at least 3x. With no egress payout storj might even try to max out that cost saving.

Keeping the ratio remove that issue and we can maintain current RS numbers. We do have a littlebit room to reduce the egress rate more than used space that would not be worth an RS change. I will have to think about that.

That is missing the expansion factor.

Edit: And ofc I should have started with thank you for sharing your numbers.

10

My Storj node payout tiers — rate cut I could absorb (setup private):
• Survival: −10% → $1.35 / $1.80 per TB
(− = reduction from today’s rates · storage $/TB · egress $/TB · scaled together per RS balance)

I think this is affordable only, if the following problems are solved:

  • the inconsistency between the real on-disk occupied data vs what I’m paid for, numbers don’t match. I host many but get paid like 15-20% less, always. If we can solve this, then -10% is acceptable even for long term.

  • new data, more consist data, like backup data, of course.

11

My appoligize for that. We have migrated most of the infrastruture to self hosted and kind of disabled every service that makes that migration more difficult. You should have received new bloom filter yesterday. I believe we are still missing US1. In any case it will take about 7 more days until that space will get cleaned up. Not entirely because some overhead is normal. There are ways to reduce that but they will increase compact runtime.

12

You are NOT an operator. Have at least 100Tb then call yourself SNO. This calculator is another joke.

13

Where did that number come from? Being a SNO isn’t just running at least one node?

Can I change the number to 450TB and you just go away for a couple months? :wink:

14

A calculator that requires me to calculate all the expenses myself in advance is genius. ))) What’s the point of it anyway? Do you want to know at what level the SNO won’t PAY you EXTRA for being a SNO? ))) When you go to the store for sausage, do you ask the owner for a financial report so you can demand that they sell you sausage exactly at cost? They tell you outright that you CANNOT reduce payments; no one will work for you for free. Not by -10% or -1%. All you’ll get is a boost in recovery traffic, which you’ll pay for by spending the “savings.”

15

I do. There is an old thread here in the forum with my latest setup. You can look that up and ask me any question there.

If you have an improvement idea I am happy to implement it. Its open source so you can also take the code and build your own version.

18

People sorry but you are not helping but trolling and complaining only.
Even if you are Eaten by bear you have two ways out.

19

you are litle bit not understand, you are not employ but freelancer. And instead of just tell you you will get now 50% from last month Storj try understand what is possible.

Storj workers are in same position, Boses left company with money and other should fix that sh*t.