Categories
Mining Business

A hosting renewal is a negotiation most owners treat as paperwork. What the offer really contains, how to judge the rate, and the leverage you actually hold.

Every hosting agreement has an end date, and most of them quietly roll past it. The provider sends a renewal notice, the rate has moved a little, and the owner signs because the machines are already racked and hashing. The whole thing takes ten minutes.

That is the most expensive ten minutes in the hosted mining calendar. The renewal is the one moment when the commercial terms of the relationship are genuinely open, and it is the only moment when both sides know exactly how the last term actually went. Treating it as paperwork hands the entire negotiation to the other side.

The renewal is on a clock, and the clock is in your contract

Before anything else, find the notice provision. Most hosting agreements renew automatically unless one side gives written notice inside a defined window, and that window is often narrower than people expect. Miss it and the term renews on whatever terms the agreement specifies for a renewal period, which may not be the terms you would have accepted.

The window matters even if you fully intend to stay. A renewal conversation where the provider knows your notice deadline has passed is a different conversation from one where it has not. You do not have to leave to benefit from being able to.

Work backward from that date. If notice is due ninety days before the term ends, the useful work starts well before that: pulling your own performance record, deciding what you want changed, and asking the provider for their proposal early enough that there is time to respond to it. A renewal offer that lands two weeks before the deadline is not a negotiation, it is a deadline.

What a renewal offer actually contains

The headline is the rate. Almost everything that determines what you pay is underneath it.

The rate and what it includes. An all-in number and a power-plus-fee structure can produce very different bills at the same headline figure, because they treat overhead, infrastructure charges and pass-throughs differently. Compare like for like, and compare at your actual consumption rather than at nameplate.

Term length. A longer term usually buys a better rate. It also removes your ability to react for that period, which is worth something in a market where hashprice moves and machine efficiency moves with it. There is no universally right answer here, but there is a wrong way to decide it, which is to take the longest term because it has the lowest number on it.

Minimum commitments. Some renewals introduce or increase a minimum monthly charge, a committed load, or a take-or-pay floor. These change your risk profile materially and they tend to appear in renewal terms more often than in first contracts.

Pass-through mechanics. If power is passed through at cost plus a fee, the fee is negotiable and the cost is not, but the way the cost is calculated and evidenced certainly is. The right to see the underlying utility data is worth asking for.

Curtailment terms. How much notice do you get, how is curtailed time treated on the invoice, and is there any cap on hours. For a site with real curtailment exposure, this clause can outweigh a cent on the rate.

Service levels. Uptime commitments, how downtime is measured, what the remedy is. If the last term produced downtime you were unhappy about, the renewal is when that gets addressed, not later.

Judge the rate against your own record, not against a market price

There is no published spot price for hosting. Rates vary by site, by power contract, by machine density, by how much of the facility is already committed, and by what the provider thinks you will accept. Asking around produces anecdotes, not a benchmark.

What you do have is a year of your own data. Before you respond to a renewal offer, work out what you actually paid per kilowatt-hour delivered to your machines over the last term, including every line item, divided by what the fleet actually consumed. That is your real rate, and it is frequently not the rate on the first page of the contract. If you have never done this, checking a hosting invoice line by line is the place to start.

Then set that against delivered performance. Uptime you measured yourself, not uptime you were told. Curtailed hours. Machines that sat dead waiting on a technician. If you keep an asset register with serial-level history, this is the moment it pays for itself, because it converts a complaint into a record.

A renewal conversation built on your own numbers is a different conversation. It moves from “this feels expensive” to “we paid this much per delivered kilowatt-hour and lost this many machine-days, and here is what we want to change.”

The leverage you actually have

Owners routinely assume they have none, because moving is painful. The pain is real and it is worth being honest about it: de-racking, crating, freight, downtime in transit, and intake at the other end all cost money and time. If you have not priced that, read what an orderly move at the end of a term actually involves before you use it as a threat, because a threat you cannot afford to execute is not leverage.

But the provider’s position is not as comfortable as it looks either. An occupied rack generates revenue; an empty one generates nothing and still carries its share of the facility’s fixed cost. Replacing a customer takes time, and the replacement arrives with unknown machines, unknown payment behavior and an intake process to work through. A paying customer with a clean record and known load is worth keeping, and the provider knows what it cost them to find you.

That asymmetry is smaller than you think in both directions. It is enough to negotiate with, and not enough to make demands with.

Ask for things that are not the rate

Providers have limited room on price and considerably more room on terms. If the rate will not move, these often will.

Notice periods, in both directions. Curtailment notice. A defined response time for remote hands requests, and clarity on what is included versus billable. Payment terms and what happens on a late invoice, which is worth settling calmly in advance rather than discovering during a dispute, as anyone who has watched a suspension over an unpaid invoice can confirm. Firmware policy, and who is allowed to change what. Reporting: what data you get, in what format, how often.

These concessions cost the provider little and change your operating experience considerably. They are also easier to agree to than a rate cut, because they do not set a precedent the provider has to explain to every other customer.

Questions worth asking before you sign another term

What has changed at the site since the last agreement. New customers, new load, new equipment, changes to the power contract. A facility that was comfortably within its envelope two years ago may not be now, and a site that has been sold past its real capacity will show it in curtailment and in thermal performance long before anyone admits it.

Who owns the business now, and who will be operating it for the length of the term you are being asked to sign. Ownership in this industry changes hands regularly.

What the provider’s own power position looks like going into the new term. You are not entitled to their contracts, but a provider who will not say anything at all about the shape of their power cost is telling you something.

If the agreement is substantially rewritten rather than extended, read it as a new contract, because that is what it is. The clauses that matter in a hosting agreement do not become less important the second time.

When to split rather than renew or leave

The decision is not binary. Renewing part of the fleet and placing the rest somewhere else is often the most rational outcome, particularly if the current provider is adequate rather than good. It tests an alternative at manageable cost, it gives you a real comparison next time, and it reduces the concentration risk that comes with having every machine in one building. That is the case for using more than one host, and a renewal window is the natural moment to act on it.

It also changes the renewal conversation, because a provider negotiating to keep most of your fleet is in a different position from one negotiating to keep all of it.

When to walk

Walk when the record is bad and the explanation has not changed. Repeated downtime without a root cause, invoices you cannot reconcile, requests that go unanswered for days, or a site that is visibly beyond what its infrastructure supports. A renewal does not fix any of that; it funds another year of it.

Walk when the proposed terms shift risk onto you without compensation, such as a minimum commitment introduced at the same time the uptime guarantee is weakened.

And walk if the provider will not put the operational commitments in writing. A provider who is willing to promise a response time on a call and unwilling to write it into the agreement has told you exactly how that promise will be treated when it matters.

The renewal is the relationship

A first hosting contract is signed on a brochure and a conversation. A renewal is signed on evidence: a term of real invoices, real uptime and real responsiveness. It is the only point in the relationship where both sides are negotiating with the same facts.

Owners who prepare for it the way they prepared for the original search end up with better rates, better terms, or a better provider. Owners who sign it in ten minutes find out at the next renewal what that cost them.

If you are approaching the end of a term and want to talk through what you are being offered, or what it would take to move part of a fleet, get in touch.

Explore Rax Mining

Categories