Why almost every VPN is sold by the year

Acquiring a VPN subscriber costs money up front. Search advertising, affiliate commissions on review sites, sponsorships, app store placement: all of it is spent before the first payment. And it's spent on everyone who sees it, whether or not they go on to subscribe.

Whatever it works out to per paying customer, it left the bank account before that customer existed.

A long commitment recovers that spend immediately. Say you buy a prepaid year. It arrives as one payment that can be set against the cost of winning you, and what's left funds the next round of acquisition.

Now say you buy a single week. Your payment recovers a fraction of the same cost, and the provider waits to see whether you return.

The plan discounted hardest is the plan that repays its acquisition cost fastest. That's why the discount tends to grow with the term. It doesn't grow with anything about the product.

That's ordinary business. It isn't a trick, and it's worth understanding precisely because it's ordinary. The pricing page solves the provider's cash flow problem. It doesn't answer your question, which is how long you need a VPN.

The large number on a pricing page laid out this way is a quotient. It's the total charged for the term, divided by the number of months in it. It describes a rate.

The transaction is something else: one charge for the whole term, taken on the day you buy. The amount that actually leaves your account tends to live in the smaller type.

Two and three-year terms push this further. A longer divisor produces a smaller headline figure from a larger payment. So the plan advertised at the lowest monthly rate may carry the largest single charge.

Whether the total is stated nearby is up to the seller. Either way, at a glance you're comparing a rate nobody is billed with a rate somebody would be.

Suppose a page shows a monthly plan beside a two-year plan's per-month figure. The first is what you'd be billed each month. The second is one charge, taken today, divided by 24.

The correction is one step. For each provider, find the amount charged today and the number of months it covers. A genuine monthly plan and an annual plan's per-month figure are different kinds of thing. Set them side by side and a commitment looks like a price.

Where the first term is sold at an introductory rate, the second term renews at a different one. That's the standard price for the plan, and it replaces the promotional figure that won the sale.

The discount was part of the acquisition cost. It bought a customer, and nothing obliges it to apply for the life of the account.

The gap between those two numbers is the thing worth finding. Where it's disclosed varies by seller: the plan comparison, the small print, or the checkout screen.

If you note the introductory figure, you've noted the least durable number on offer. The price that describes what the service costs to keep is the second-term one. Rank providers on that basis and the order can change.

A year is also long enough to forget what you agreed to. The renewal arrives twelve months after a decision you made in minutes. It's charged to whatever payment method is still on file, and the first evidence of it can be the charge itself.

None of that is unique to VPNs. But a long term, an introductory rate and automatic renewal together are how a cheap purchase becomes a standing one.

The figure a subscription business is judged on is churn: the share of subscribers who don't renew. Several familiar features of subscription pricing act as levers on it.

Long terms reduce the occasions on which somebody can leave. Automatic renewal makes continuing the default and leaving an action. And a discount offered at the moment of cancellation can turn a departure back into a subscription.

Seen through that lens, some decisions that look like carelessness aren't. Think of a cancellation flow several screens deep, a reminder timed close to the renewal, or an offer that appears only once you try to leave.

Where you meet one of those, don't assume an oversight in an otherwise tidy product. The cheapest explanation is a defence of the number the business is judged on. That's our reading of the incentive. It isn't a finding about any particular seller.

One second-order effect is worth naming. Traffic is a real cost and the subscription is flat. So a subscriber who pays and connects rarely is more profitable than one who pays and streams.

The ideal customer of a flat-rate annual VPN is somebody who renews and barely uses it. That isn't an accusation about anyone's conduct. It's only the direction the incentive points.

Demand for a consumer VPN is often episodic. A trip. A fortnight of hotel and airport networks. A month working somewhere unfamiliar. One evening on a network nobody should trust.

The requirement has a beginning and an end, and often you know the end in advance.

An annual plan prices that requirement as a year. Say you need two weeks. You pay for the other fifty as well, and the advertised saving applies to a quantity you didn't want.

For a short need, even a deeply discounted annual plan can be the more expensive purchase. What matters is your total outlay against the period you actually use. The price per month is the wrong measure.

Cancelling is the other cost, and you pay it in attention, not in money. Eleven months later you have to remember a decision you took in minutes, find the right settings screen, and do it before a date nobody wrote down.

A forgotten cancellation is revenue the seller keeps. A yearly cycle leaves more room to forget than a short one.

Refunds depend on the seller and on the store that took the payment. Only Apple can refund a subscription bought inside an iPhone, iPad or Mac app. A Google Play purchase is different: the provider can refund it directly. With us, you ask in the app. On Android we refund a recent charge at once, and someone on our team reviews an older one. On iPhone, iPad and Mac the app opens Apple's refund form, and Apple decides.

We have no annual plan. Our cycles are $1.99 a week and $3.99 a month, and on Android there's also a $0.99 day pass. The monthly figure is the ongoing price. It isn't a first-term rate that rises later.

There's no second-term price to look up, because there's no introductory term. A seven-day free trial is attached to the monthly plan, and to that plan alone. The day pass and the weekly plan are paid from the start.

You can do the arithmetic between the three on the page. Two days at $0.99 is $1.98, so from the third day the weekly plan is the cheaper way to buy the same thing. Two weeks at $1.99 is $3.98 against $3.99 for a month, so from the third week the monthly plan is.

The day pass is for one or two days and loses its purpose after that. For example, pick it for a week-long trip and you pay $6.93 for what $1.99 covers.

Selling in these units is the worse business on every axis described above. There are more billing events for the same revenue. We spread acquisition cost across days, where an annual seller has a year. And a customer whose trip has ended may not return.

So why do it anyway? Because a good deal of the demand is genuinely short. If you reach for a VPN for one journey, you don't need the eleven months after it. An annual plan sells them to you anyway.

Here the argument runs against us. Twelve consecutive months at $3.99 is $47.88. An annual plan discounted hard enough to come in under that figure is the cheaper purchase if you want a VPN connected most days of the year.

In that case you're paying a premium here for flexibility you aren't using, and the saving from a long commitment is real. What any given provider charges for a year is something you'll have to go and check. This page can't tell you.

If you buy year-round, the short-cycle rate is simply what you keep paying. Our weekly and monthly plans renew until you cancel, just as an annual one does, so their flexibility only pays when you actually use it.

Our list of exits is short, and it changes as we add and retire servers. If there's no exit near where you need one, we can't serve you well at any price. Paying more doesn't close that gap.

What a short plan buys is the ability to stop, and that's a product with a price. The per-day cost of a short cycle is higher than the per-day cost of a long one, and it should be.

The question to settle before you open any pricing page is how long you honestly want the thing. Suppose the answer is a fortnight. Then the annual discount is irrelevant.

Suppose the answer is indefinitely. Then the discount is the point, and the sensible move is to find the best ongoing annual price. Don't buy flexibility that will sit unused.

Six checks cover most of what the layout obscures. You don't need to know how a VPN works for any of them. You can do them all from the public pricing page in a couple of minutes.