Grey routes: why the cheapest SMS path is rarely the safest

Anyone comparing SMS pricing for a given destination notices the same thing: the quotes spread out far more than the underlying cost of delivery could explain. Some of that spread is commercial. Some of it is that the cheapest quotes are not describing the same product. They are describing grey routes, and the difference only becomes visible at the worst possible moment.

What makes a route grey

It helps to start from what a proper route costs. Reaching a subscriber means using an operator's network, and operators price business messaging deliberately, because it is a service sold to companies rather than a call between individuals. Any quote materially below that floor is either absorbing a loss, which nobody does for long, or avoiding the floor altogether. A legitimate application-to-person route reaches subscribers through a commercial agreement with their operator, who knows the traffic is business messaging and prices it as such. A grey route reaches the same subscribers by other means — traffic presented as something it is not, or delivered through an arrangement the destination operator has not agreed to. The message arrives, at least for a while, which is precisely what makes the practice persist. Nobody sells a grey route as such; it is sold as a competitive rate, and the rate is competitive because a cost that should have been paid has not been.

What you lose, and when you find out

The trouble with grey routes is not that they fail immediately. It is that they work until they do not, and their failure mode is abrupt rather than gradual. An operator identifies the traffic and blocks it, and delivery for that destination stops within the hour, usually with no warning and no fallback prepared. The consequences land where they hurt most, on the messages that mattered.

• →Sender identity is often rewritten or stripped, so the message arrives from an unrecognisable source and the recipient distrusts it.
• →Delivery receipts become unreliable, because the party generating them may not be the party delivering.
• →Longer messages are more likely to arrive split or out of order.
• →When the route is blocked, everything using it stops at once — including the authentication codes that keep customers able to log in.

How to tell before it matters

You rarely get told, and the vocabulary does not help: nobody offers a route described as irregular, and terms such as direct or premium are used loosely enough to mean very little on their own. What you can do is ask questions whose answers are hard to fake. Ask how the sender identity is handled for the destination in question, and whether registration is required there. Ask where delivery receipts are generated. Ask what happens to traffic when the primary path is unavailable — a partner who has thought about it will describe a fallback, a partner who has not will change the subject. And compare reported delivery against something you can observe yourself: codes entered, links opened, replies received.

It is also worth thinking about which traffic you are willing to expose. Many businesses conclude that promotional messaging can tolerate an occasional bad day, while authentication and transaction alerts cannot, and route them separately for that reason. Splitting traffic by consequence rather than by volume is a small piece of design that limits how much damage a single blocked path can do. The honest summary is that a rate which looks unusually good for a difficult destination is telling you something about how it is achieved. That is not an argument for paying the most; it is an argument for knowing what you are buying, and for asking the questions before the traffic matters rather than during an incident.