As someone who has been in the web hosting industry for years it got me thinking, if people understood the economics of hosting companies, would their expectations change?
Hosting companies are largely like most others, where gross profit determines how many people they can staff for support among other things. A lot of hosting firms will have an average cost of goods sold (servers at data centers) of around 30%. That leaves 70% of each dollar they bring in to pay staff and everything else.
If the company is still trying to grow (and hopefully it is) then another 10% or so will go towards marketing expenses. After all, the cost to acquire a customer in hosting is actually one of the highest out there. Bidding on a keyword like “web hosting” can cost up to $50 per click with AdWords. So marketing and COGS (servers, infrastructure) alone mean that only 60% of each dollar they bring in is available to pay salaries and all other expenses.
Let’s say you’re paying them $10 per month ($120 per year) for managed WordPress hosting. You’ve been told by their marketing copy that they’re amazing, they have 24×7 support, phone support, and you’re expecting the experience to be like most other companies you deal with. You call GEICO, Anthem, AT&T, or Verizon and they’re not amazing but they do fix things eventually in your battle with them.
The difference is that you’re not paying them a paltry $10 per month, and their support agents don’t cost them nearly as much, since resolving an issue with your cell phone bill can be solved by someone making $40K per year, and doesn’t require a software engineer.
Let’s look at the mathematics briefly. If you’re paying the host $10 per month ($120 per year), and a good system administrator costs $100,000 per year including benefits, how many customers do they need just to cover that one employee?
The answer might surprise you. It’s not 833, since 833 x $120 = ~ $100K. That would assume that the company doesn’t have any cost of goods sold or marketing, but remember, the hosting company in this example has a COGS of 30%, marketing budget of 10%, and therefore only 60 cents of every dollar they bring in is available to pay staff.
Using the above assumptions, the real answer can be found with some simple math.
$100,000 = X * 120 * 0.6
$100,000 = 72X
1389 = X
By solving for X above, we find that it’s equal to nearly 1400 customers. So for every 1400 customers at $10 per month, a hosting firm can afford to hire a system administrator who actually knows what they’re doing. Or think of it like this, if 1389 customers pay $10/mo each ($120 per year), that’s $166,680 in annual revenue and after the 30% COGS (servers) and 10% marketing, the company is left with $100K to cover a salary including benefits.
But how can one system administrator, the type that can fix any crazy hosting problem that arises, handle nearly 1400 customers’ technical issues? The answer is they can’t.
This isn’t even accounting for the fact that a real business has other costs to use their gross profit minus marketing for, things like:
- Paying the founders or C levels
- Rent and staff
- Bandwidth costs and hosting infrastructure
- Complex analytics setup to handle, receive and process logs (Elastic)
- Transactional email services
- CDN integrations
- Support and ticket systems such as Intercom
We’re using this model because the math is easiest and simplest in these terms, with very few assumptions made.
When you move to $20 per month, the mathematics fall a little more into your favor, but still aren’t great. It would take roughly 700 customers to cover that same system administrator at that price point. Even at $50 per month, we’re still not looking that great with a ratio of 1 system administrator for every 278 customers. As you can see, until you get into the $100+ per month range, the economics aren’t even close to being in your favor, where you’re going to have a solid ratio.
Web Hosting Industry Lowdown
So if the economics aren’t in the customer’s favor when charging $10 per month, why do so many hosting companies do it? The answer is twofold. One is that they try to make a lot of money on upselling and hidden fees.
For example, things like website migration, SSL certificates, domain registrations, WHOIS privacy, ad revenue from parked domains, dedicated IP addresses, fake SEO help, website badges, and other stuff that is mostly junk. But those don’t always happen, as the people using these services are typically trying to save money and that’s why they’ve chosen them.
The second is what we call in the hosting industry “churn and burn”. Wait, that doesn’t sound good, what does that mean? It’s when you have a model where you know you’re charging too little with a bait-and-switch strategy, you know as a result your customer service is going to be terrible, you know people are going to be upset, you know you’re going to lose X customers per quarter, and you’ve designed your business model around trying to replace those X customers.
How does that actually play out? It’s pretty simple in hosting. You offer unlimited everything (except CPU or something hosts call “workers“), and then you throttle those clients/sites that push lots of bandwidth. The performance of these successful sites tank, they leave unhappy, and the host is happy because they can replace them with a site that gets little to no traffic. It’s business models like this that give the hosting industry a bad reputation.
So when you’re thinking about shopping for hosting, stop thinking in terms of paying the absolute minimum, and start thinking about what you’re really buying. Support and access to people with brains. You’re buying into a team, that will hopefully be there to help you along your journey.
Certainly, not everyone can afford hundreds per month for hosting, but on the flip side they should also set their expectations accordingly. They should expect their site to go down from time to time (since at $10/mo, you’re most likely sharing a server with way too many others) and expect that it won’t be resolved all that quickly. It’s just how the numbers work out.