Ask anyone whether they understand mobile communication tariffs and you're likely to receive a mumble in return. It's not that people are completely clueless, just that most know that their notion of 'more minutes for more money' is not quite the full story.

To be fair, this is not a problem unique to mobile telecoms. As the glam kid of the class, mobile is singled out because people think about it more than fixed line or broadband. It's viewed more as a personal service than as a utility. Perhaps not quite cute and cuddly, but at least imbued with some flexibility and personality, unlike its grey sibling landline services.

Mobile also has one other important distinction - the risk of bill shock. It is possible to run up an eye-watering bill with a landline phone, but you have to try quite hard or be unlucky. Calling Mongolia, failing to properly disconnect the call and waiting for a few days would do it. But in the main mobile is where the high potential for misunderstanding occurs.

Let's say I want to save some money on car insurance, so ask for a few competing quotes. Insurance companies helpfully provide 'freefone' numbers to entice prospective customers to speak to them. After phoning round and receiving 4 quotes, I pick the cheapest and save myself £30. Ok, it took 90 minutes on the phone but thirty quid ain't bad. Problem is, more than half (£16.20) of my saving is wiped out by my mobile phone bill for those 4 calls! In that case why not call them 'pay through the nose fone' numbers?

In simple terms the problem is advice of charge. Consumers of communication services typically don't know how much a service will cost before they use it.

It has been this way ever since automated exchanges did away with the human interface of the operator connecting calls. The number you dial determines how much you will be charged. If there were, say, 3 types of call and the way to invoke each call type were different, then it would be reasonable to assume that people could memorise the costs and there would be no surprises come the day when the bill flutters softly on to the front doormat.

But we know that there are many types of call, all initiated in the same way and only distinguished by the pattern of numbers tapped into the device. If one call costs nothing because it begins with 02, and another costs £0.18 per minute because it begins with 08, and yet another costs £0.30 per minute because it beings with 070, how is the average consumer expected to know? Yes, all call rates are published on websites and in documents exchanged with the customer when they sign up, but how many study and understand these documents even if they have taken the trouble to locate them in the first place?

The usual retail experience is so different. As we are enticed to buy seductively packaged goods, prices scream out at us that we're saving, getting best value, more for less. It's almost impossible not to know how much something is going to cost because it's clearly labelled.

One solution might be to introduce a prefix which provides an audible advice of charge before the call connection is made. For example, keying 111 before the destination number would allow the caller to hear the unit cost and any call banding. It would be interesting to see how much such a feature would be used.

All this before we've touched on roaming, which is a notable example of free market failure and should probably be left as a topic for another time.

It all leads to distrust of providers by consumers. The uncertainty of call costs inhibits take-up of additional services and contributes to worsen supplier-buyer relationships.

In an attempt to simplify the market providers have come up with the notion of 'bundles', which include an allowance of commonly used services for a flat monthly fee. There is no doubt that this concept is widely understood. Bundled services are usually built up at discounted rates, so that a service which would normally cost £0.08 per minute would cost £0.05 per minute if the whole bundle allowance is used.

Unfortunately, although bundles have been generally accepted and enabled limited up-selling of services, they also raise complications. A simple 500 minute per month bundle for £25 is only a good financial deal for the consumer if the minutes actually consumed would have cost more to buy outside the bundle. The issue of right-sizing appears, where the balance between buying the slightly cheaper bundle must be balanced against the likelihood that some of those bundled services may not be consumed.

Right-sizing is further complicated by several additional factors. Only certain specific services are included within the bundle. In a way this goes back to the advice of charge problem above. How does a consumer know that the destination being tapped on to the keypad is in or out of the bundle which has already been paid for? Some services and destinations are allocated separate allowances, thus fragmenting the bundle and making it harder to work out which benefits are most needed. Further options to nominate particular destinations, such as 'friends and family' numbers, make it harder to work out how much bundle is needed, because these destinations now must be subtracted from the total usage of those service to right-size the customer. Some bundle benefits carry over between months, other expire at the end of the month.

What started out looking like a simple problem has become by stages rather complicated. Only the geekiest of consumers actually bothers to probe this murky world and attempt to work out the true benefit.

Now wouldn't it be nice if providers could right-size their customers, or trusted third parties right-size for customers not currently in a contract?

What about the original question, of why mobile tariffs are so complicated? That is, not explaining the complications but discussing the motivation for making them so. Is there a simple answer? Is the moon made of blue cheese (just think of the energy content if it was)? Such a discussion deserves an article of its own and should be a cracker.