Self-service delivery of financial services has been a staple of the banking landscape for the past thirty years. The device known as the ATM has gone from novelty to necessity for most consumers. ATMs have proliferated from the bank on the corner to locations as unlikely and varied as churches, police stations, airports, bars, high schools, grocery stores, pizza parlors, and thousands of other sites. An episode of the TV series "Friends" featured an ATM prominently, as has the History Channel (on the Modern Marvels program, no less), and an ATM is scheduled to be one of the stars of a Hollywood feature called "The Barber Shop" coming soon to a theatre near you. You couldn't click on the television at midnight on Dec. 31, 1999 without seeing literally hundreds of news reporters using ATMs successfully after Y2K and assuring the nation that all was well.
So for all their ubiquity and assumed usefulness, would it not be more than reasonable to assume that the ATM has arrived? Perhaps by some measures you could call ATMs an unqualified success. However, a closer look at the quantitative data suggests that maybe we in the industry still have a way to go before we declare victory. In fact, I would call an overview of ATMs from an empirical standpoint the good, the bad, and the opportunity.
First, the good news. In or around 1995, self-service delivery in the United States crossed an important threshold. In that timeframe, financial institutions discovered that for the first time the number of transactions processed on ATMs exceeded those processed in a branch. The implications for an ATM owning institution's non-interest operating expense was favorable. Since ATMs cost per transaction weighed in at around $0.35 and the teller at $1.07, the more transactions that migrated to ATMs, the better for the P&L.; And indeed, self-service transactions have continued to grow. Since 1995, ATM volume alone has grown from around 8 billion transactions in the United States to over 11 billion. Also since 1995 the number of ATMs installed in the United States has nearly tripled. Add to that the recent phenomenon of transaction fee income from surcharges and things couldn't be better, right?
There is some bad news under the gloss. One of the key efficiencies that ATM deployers look for is maximizing the numbers of transactions per month that an ATM supports. And that metric has been in a nosedive since the introduction of surcharging. According to Bank Network News, the average ATM in the United States ran 6,580 transactions per month in 1995. In 2001, The same average machine was running nearly half as many transactions at 3,494 per month. Certainly, a major contributing factor in this decline has been the explosive growth of deployments. And this deployment growth has been fueled by lost cost (costs in terms of both acquisition and operation) single function units which require very few transactions to operate at break even. This trend doesn't bode well for the effort to add volume and utility to convenience ATMs.
Finally, let's review the opportunities that ATM owners and operators have at the present. At no time in the history of the industry have so many owners, vendors, deployers, and customers had so much choice with respect to potential and actual new transactions. While some of these new transactions are not so "new" as they are improvements on existing functions, consumers should find value in most of today's development activities.
The biggest change in ATMs will be related to deposit and payment transactions at ATMs. Consumers and deployers will notice marked improvements in deposit and payment transactions over the next several years. I believe the market can expect expansion of bar code reading features like those in use in South America today for bill payments.
This technology will allow consumers to credit accounts without enclosing checks in envelopes, instead they will simply read the bar code line from a statement or bill and directly debit their accounts. Likewise, deposits will not require envelopes. The way this transaction will be modified will be to allow consumers to place bare checks or and or bunches of cash directly into depositories. The notes will be automatically counted and processed. Likewise, the checks will be digitally imaged allowing for immediate positive verification and eventually allowing banks to clear images and substitute checks while eliminating the expense and difficulty of processing envelopes and their contents. This effort is currently the focus of the Federal Reserve Board's proposed Check Truncation Act. ATMs will be a prominent consideration in the implementation of streamlined deposit automation.
Today consumers think of ATMs simply as cash dispensers. Most signage for ATMs simply advertise that cash is available. When consumers become confident that an ATM can perform a deposit and payment transaction, then they will see many other types of transactions. I believe that adding functions is the best way to improve volumes and through-put on ATMs today. Adding transactions closely linked to consumers' financial needs like money order or other negotiable item sales, wire transfer, pre-paid card dispensing, and check cashing will add value to the ATM as a "destination" and improve the revenue generating potential for owners. But these types of services demand upgrades in the archaic ATM architecture to current standards from an application and platform standpoint. In addition, the vendor community will need to begin offering these services on multi-vendor platforms to ease the burden of deployment and support.
Financial institutions are keen to deploy deposit and payment automation in order to reduce their transaction costs. These institutions anticipate savings similar to the transaction savings create by the cash withdrawal in the ATM.
There may be some concerns about queuing and transaction times when discussions begin surrounding new types of transactions. The facts of the situation are that even the most used ATMs (expressed as tens of thousands of transactions on a monthly basis) sit idle for the majority of the day. My view is that our industry could and should be moving as many and as much transactions and volumes to the ATM channel to support consumer demand, drive additional revenue for the owners, and reduce comparable operating costs for deployers.
To demonstrate the volume of excess capacity in today's ATMs, perform the following calculations. If we assume that a typical ATM is available to public for just 16 hours per day, the average 30 day month has 28,800 minutes. If we also assume that the average ATM transaction last one minute, then we can assume that average ATM is utilized only about 12% of its normal business day (3,494 divided by 28,800). This calculation indicates a significant amount of idle capacity in the nation's ATMs.
As successful as the ATM has been over the past thirty years, I believe that clearly the best is yet to come in terms of quantity and quality for service delivery at ATMs. The ATM channel puts the action in transactions!