AMA — I sell offshore hosting. Ask me the things I'd rather you didn't.

What would you change about how customers evaluate providers? Assume you're not selling to them.

Stop reading the abuse policy and start reading the network page.

The abuse policy is marketing. It costs nothing to write, it binds nobody, and everyone in the segment has broadly the same one. What it cannot fake is who provides the transit, whether the ASN belongs to the company or someone else, and how long that ASN has existed. Ten minutes with a looking glass tells you more about whether a host can keep its promise than a year of reading terms.

Second thing: ask for a location's owned-or-resold status in writing before you order, and keep the reply. Nobody does this. It takes one email, it's a perfectly normal commercial question, and the answer — including no answer — is the most informative thing you'll get from a sales team.

Third, and this is the one that'll cost me money to say: for most people here, the right choice is one of the three or four names that have been doing this for over a decade, not the cheaper thing with better marketing. FlokiNET and PRQ have survived rounds of pressure that removed everyone else selling the same promise. That's not a feature list, it's evidence, and there's no substitute for it.

If your requirement is genuinely just a box that won't fold on a template letter, buy from the boring old company and stop reading forums about it.

Noted that the guy selling hosting just told the thread to buy from someone else. Fair play.

The looking-glass advice is the most actionable thing in this thread and it will be the least followed.

My friend, why nobody in this business have server in South America? Is real question, we have money also.

Honest answer: transit cost and legal predictability, in that order.

Bandwidth into and out of South America has historically cost several times what it costs in Frankfurt or Amsterdam, because there's less of it and it's owned by fewer people. On a €5 product that difference isn't a margin problem, it's the whole product.

The second reason is that the value proposition of this segment is jurisdictional, and it depends on a legal environment that is slow, predictable and unfriendly to foreign complainants. Some South American jurisdictions are genuinely fine on that axis. But "fine" isn't enough — you need a decade of demonstrated behaviour before anyone will trust their business to it, and nobody's had the patience to build that.

It'll happen. The transit economics have improved a lot in five years. It just needs someone willing to be unprofitable for three years to establish the record, and this industry is not full of patient capital.

How much of your customer base is on a plan that doesn't do what they think it does?

More than I'd like, and it's the thing about tiered product lines that I find genuinely uncomfortable as a seller.

At my scale it's maybe one in eight. People buy the cheap plan because it's the cheap plan, and the cheap plan is on resold capacity where I can't protect them, and they read the company's abuse position from the front page rather than the plan page. I catch some of them at signup when the category is obvious and move them up, usually at the same price for the first term because it's my product design that confused them.

AbeloHost gets named for this constantly on this board and I'd defend them slightly: they publish the tier difference clearly. The problem is that customers shop on price first and read second, and that's true at every host with more than one product line, mine included.

If you take one thing from this AMA: the plan is the policy, not the company.

Has your own risk appetite changed over the years, and in which direction?

Narrower, and I think that's the normal trajectory rather than a personal failing.

When I started I'd take almost anything that wasn't on the two hard-no list, because I needed the revenue and I hadn't yet had the experience of a transit provider asking me a question in a certain tone. After you've nearly lost a location over one customer, you start pricing that risk into who you accept, and pricing it in means saying no more often.

The uncomfortable implication is that every operator in this market gets more conservative as they get more established, which means the companies with the best track records are also the ones most likely to turn you down. The ones that'll take anything are the ones that haven't been burned yet, and you're buying the same lesson they haven't learned.

If a client needed capacity in a specific jurisdiction that no reputable operator serves, what does that conversation actually look like from your side?

Reminder that this is an AMA, not a marketplace. Two DMs reported. Take it elsewhere.

Winding this up before it turns into a year-long thing.

Two questions above I never answered. One I can't answer without describing my own supply chain closely enough to identify it. The other one I've read four times and I don't have anything useful to say that isn't a guess, and there's enough guessing in this industry.

What I'd want anyone to take away, since a few people have asked me to summarise:

1. Every policy in this market is a claim about a contract you can't see.
2. The plan is the policy, not the company.
3. Operating history is the only unforgeable signal, and it costs more for good reason.
4. If a price seems impossible for the jurisdiction, it's resold, and resold means somebody else's tolerance sets yours.

None of that is what I'd put on my own website, which is roughly the point of doing this.