This report outlines how Mobile Payments are working and/or are likely to work according to the following scenarios.

(A) Person to Person Payments from the User’s Point of View

Here the idea is that an individual would text some money to another individual to transfer some money to them. It is illustrated in the diagram below.

It is essentially a substitution for giving a cheque or cash. This ought to have some appeal as going to the bank to pay in a cheque is hassle for the recipient. Similarly, the obtaining and handing over of cash can be a pain for the giver, particularly if the sum involved is large. The other reasons why this payment method should be attractive is that:

• The speed of a telephone payment (e.g. using text a message) is probably okay for person to person payment.
• This supports remote person to person payment (e.g. while people are talking on a phone to each other) which cheques and cash are particularly bad at.

The diagram is deliberately vague as to what should get debited; some kind of account has to be involved. The “something” account could be:

• A telephone billing account in the Telco
• A bank account
• A credit card account
• An electronic wallet distinct from all the above.

(B) Person to Person Payments from the Bank’s Point of View

There are three models for providing this service, which is essentially one of the sender “pushing” money to the receiver via mobile phones.

• Off an Electronic Wallet
• Off the Telephone Company (Telco)Phone Bill
• Off a Bank Account (or Credit Card Account)

We look at each of these in turn, followed by a general look at security.

Off an Electronic Wallet

The Electronic Wallet would be the most simple to describe. the diagram below illustrates it:

 

It is based on a central electronic wallet system that is effectively moving money between user’s wallets by debiting and crediting records in a single database. The electronic wallet system is separate from the bank accounts (system). NatWest’s FastPay is an example of this approach (as are most email-based payment systems). “Real” money is transferred to and from the electronic wallet by a transaction with a real bank account (e.g. a debit card transaction or an automated direct credit).

The problems with this model are threefold:

• It depends on a single central wallet system. There is no standard for exchanging money between wallet systems. Thus either the entire banking industry coalesces on one system and supplier (unlikely) or the users find themselves having to have lots of wallets, one per system.
• Moving the money is a three-stage process. The money has to be got into a wallet, moved between wallets, and then moved out again. This is a bit user unfriendly (although not as bad as withdrawing cash from a bank, transferring it and then paying it in). The aim ought to be a one-stage process.
• The electronic wallet is a lot like a bank account, but without a number of the key features; namely: Credit facilities,Cash withdrawal and pay in and Regular payment capabilities

The Good features of this system are that it is easy to establish and combines well with e-mail based payments

Off the Telco Phone Bill

The idea here is that the service is provided by the Telcoms companies and the debits and credits appear on the customers phone account/bill. The diagram below illustrates how this would work:

 

Clearly the banks do not figure in this model! This is the model where the Telcos try to muscle in on banking, converting the telephone bill into a combined phone bill and electronic wallet. As a result the solution has some of the same features as the electronic wallet based approach.

One of the key differentiators from the electronic wallet is the existence of an inter-telco billing and settlement system. This in theory could form the basis of a mobile payment system that spans all telephone companies and hence all users (in Wallet section above there was a problem with a lack of standards for inter-wallet payments and settlement). However, this inter-telco billing system would need to be significantly enhanced;

• To cope with charges that are not call time based
• To reduce inter telecom settlement cycles down to a day or even real time

Telecoms companies could try and run the phone bill more as an account and less as an electronic wallet, i.e. not insisting that the payor has previously funded a payment from a bank account. In this situation the whole credit area of telcos would also need upgrading, either partnering with a bank or a credit agency such as Experian or Equifax.

Off the Bank(or Credit Card)Account

The idea here is that there is a new payment instrument that runs off the bank account. The diagram illustrates the approach.

Conceptually, this is the most attractive to a bank. However, it is not particularly easy to implement. If the accounts are both in the same bank then an inter account transfer process could be involved and in some banks this could be real time. If the transfer is between banks then the movement would have to be effected by a BACS direct credit, which would take 3 days (moving to one day in 2005). The debit could also be made to look like a debit card transaction to the bank account so in most circumstances the debit would look to be made from the sender’s point of view at the moment he calls. The recipient might get a message along the lines “You have received £30 from 1234 and it will credit your account in two days time”. This might be acceptable to users in that it is akin to getting a cheque, however, it is clearly not as good as cash from the recipient’s point of view.

Security and Mobile Phones

The most basic aspect of mobile phones is theft, with large numbers stolen every year. Clearly, if payments are to be made (particularly card/cheque sized payments) the users need to feel that a lost phone does not expose them to significant financial loss. There are a number of strategies that need to be followed to address this.

• Carry on the industry established process of upgrading phones and SIM cards so that they can be disabled once a phone is reported stolen.
• Train phone users to use PINs.
• Enhance the encryption in phones and over the airwaves.

Theft prevention by these strategies also helps the retailers’ concern with “non repudiation”, i.e. is the person with the phone the person they claim to be?

(C) Person to Merchant Payments from the User’s Point of View

Here the idea would be that a retailer would initiate an authorisation from the purchaser’s mobile phone, as illustrated in the diagram below, via a mobile device connected to or alongside the till (like a card reader for cards).

The purchaser would authorise the payment by some kind of control on his phone and cause the amount to be debited from a “something” account (more detail below) and cause the transfer of funds to the merchant’s “something” account. A key distinction between the Person to Merchant and Person to Person situation is that the retailer wants to “pull” the payment from the purchaser. Whilst theoretically the payment could be “pushed” by the purchaser (like handing over cash), this would be highly undesirable from the retailer’s point of view.

• The amount to pay is already electronically available in the till in many cases and so manual re-entry of amounts to pay introduces delays and errors.
• The retailer can associate the payment with a specific purchase electronically in the till (important for dealing with fraud, complaints, returns, etc.) which would be much harder if the payment were “pushed”.

The appeal of this solution from the users’ point of view is very far from clear.

• Text messaging can be quite slow so if the purchases were made in a busy retailer, this would build queues and frustration (certainly in comparison to the alternatives of cards, cheque or cash). Perhaps more importantly the delay in current text messaging is unpredictable depending on a host of factors such as site location, other users on the network at the time, how the handset is held, etc. (There is a technical alternative that addresses these issues, but has other problems; see Ericsson Experiment below).
• To be attractive to retailers it would have to displace a more expensive payment scheme (cash, cheques, cards). Unfortunately many of the costs of these payment schemes are fixed (sunk) costs such as, cash tills, card readers, staff cash handling costs, and security van collections. The rest of the costs are variable based on bank charges and it is not clear that banks would make the charges attractive for mobile payments. (See How banks Could Cut Costs Using Mobile Payments.) Nor is it clear that Telcos could make the costs cheap enough to offer an alternative to the banks. (See Banks Have to provide Mobile Payments as a Defence.)

Thus, if the general case for mobile payments for Person to Merchants is at best unclear, if not downright poor, are there niches from the merchant universe that may be appropriate?

• The most obvious is the situation where the mobile phone is being used as an active agent in making the purchase (e.g. buying ring tones). This probably is a good place for making such a payment as the issues of delay are not relevant since the user is “in session” with the phone. However, this is not a very big market. In 2001 Forrester forecasted this might be as big as £10 billion per annum in payments worldwide, which might translate into £100 million per annum in the UK, which might earn one or two million per annum in payment fees. In practice the current business volumes seem very much less than this.
• Another potential area is where people are ordering things over the phone. This might be attractive to people who (rightly) feel nervous about giving out giving out credit card details over the phone. In this model the only information that the customer has to give is his mobile phone number hence he or she is not a risk. Also, since taking telephone orders does not imply immediate transfer of the goods to the recipient, the delays associated with the authorisation, confirmation loop seem acceptable.
• One final area where it is suggested that this technology might work is unattended sites such as car parks or vending machines. The rationale being that users do not need correct coins/cash and vendors do not have the high costs of collecting the cash. Unfortunately the issues of waiting and uncertain wait times seem just as relevant if not more so here as for busy shops.

Ericsson Experiment

A completely different idea being piloted by Ericsson the phone manufacturer and Eurocard in Sweden has the retailer talk directly to a “virtual” credit card in the mobile phone via Bluetooth technology. This bypasses the phone network and would give good reliable response times over the short distances between a purchaser and a till in a store or between a vending machine and its customer. Unfortunately this technology would seem completely unfeasible on a wide scale basis as it would require:

• all phones to be Bluetooth (most are not)
• all phones to have the special programming needed to talk to cash tills
• all cash tills to have Bluetooth and special programming (none are)
• established standards for communications between phone manufacturers and till manufacturers.(There are none).

(D) Person to Merchant Payments from the Bank’s Point of View

The diagram below illustrates how this might work if the accounts to be debited were Telephone billing accounts (i.e. the situation where banks play almost no part and the Telcos are trying to build a competitive position in the payments and Merchant Acquisition Markets).

It has a core payment mechanism between telephone billing accounts that is the same as for Person to Person payments. However, it has the important additional elements of:

• sending a request for payment from the till to the purchaser, and
• sending a confirmation to the till that the payment has been made.

In most retail situations this whole loop has to be quite fast to be acceptable (seconds or tens of seconds). With current phone (2G) technologies and coverage we do not believe the system can be made fast enough or predictable enough to be usable.

Even without this fundamental usability problem there are a number of issues with payments off the Telco Billing Account which are described in section (B) Person to Person Payment from the Bank’s point of View and which are not repeated here.

In the same section there are two other technical mechanisms for moving the money from the purchaser to the retailer:

• Off an electronic wallet
• Off a bank account

They are illustrated in the diagrams below:

In the same way as the payment mechanism based on the telephone bill has two problems, these also have two classes of issue.

• Providing a fast, reliable loop from the retailer device to the purchaser’s phone and payment system and back to the retailer’s device
• The problems inherent in the payment mechanisms and described in the section (B) Person to Person Payment from the Bank’s point of View also apply and are not repeated here.

An additional complicating factor is that Banks generate significant income from the existing payment mechanisms, (cards, cash, cheques) and so how it works economically for a bank is a complex question that we deal with in the Section on Mobile Payments Provide an Opportunity to Cut Costs.