January 1984

In-Store ATMs: Steppingstone to POS

An Atlanta Fed survey found that 66 percent of major grocery and convenience stores have installed or plan to install automated teller machines. Retailers say the ATMs attract customers, reduce bad check problems and, perhaps most importantly, prepare consumers for point-of-sale terminals and debit cards.

Recently, many American consumers have installation of shared off-premise ATMs. They entered their local supermarket to discover an represent a significant step in the gradual automated teller machine (ATM) installed in displacement of paper checks. the front of the store. ATMs located off bank In order to probe this subject more deeply, premises appear to be spreading throughout the Federal Reseive Bank of Atlanta surveyed 35 shopping malls, supermarkets and convenience of the largest grocery and convenience store stores across the nation. In fact, the Florida chains in the Southeast. The results clearly Interchange Group—the forerunner of Florida's demonstrate that southeastern grocery and HONOR network—estimated that by 1986, convenience stores are moving aggressively to half of Florida's predicted 3,500 ATMs will be offer banking services to their customers. The located off-premise.1 Certainly, as indicated by survey found that 66 percent of major grocery substantial transaction volumes, consumers and convenience store operators either already find these ATMs convenient Beyond increased have ATMs or have definite plans to install customer convenience, however, other far- automatic teller machines on their premises. reaching implications may be drawn from the They feel that this initiative promises distinct advantages in attracting customers, reducing problems with bad checks, and expanding '"Igniting an EFT Revolution in Florida," Bank Network News, Vol. 1 their range of customer services. Furthermore, (February 8, 1983), p. 2.

31 FEDERAL RESERVE BANK OF ATLANTA

Digitized for FRASER http://fraser.stlouisfed.org/ Federal Reserve Bank of St. Louis January 1984

most of these organizations view the ATM as a The installation of in-store ATMs is signifi- steppingstone to the point-of-sale cash register cant beyond the spread of ATM cash dis- terminal and the debit card. pensers. ATMs in retail locations provide an Thus the survey confirms another evolutionary evolutionary steppingstone or "transition pro- step in the displacement of paper checks by duct" between cash acquisition and debit card electronic substitutes.2 Once customers accepted purchases. To the customer, it is only a small the ATM as a cash-acquisition device at their step from using a plastic card to acquire cash bank or other depository institution, banks and then groceries, in today's case, to using the began to place their proprietary ATMs in other plastic card to purchase groceries directly. locations, such as shopping centers and airports. From a technical standpoint, the ATM is trans- Another significant step involves the shift from formed into a point-of-sale terminal, the ATM proprietary to shared ATM networks, through card becomes a debit card, and the shared which account-holders at one depository insti- ATM network becomes a shared debit card tution can utilize the ATMs of other institutions. network. Yet market testing indicates that such With the formation of shared networks—cur- a transition will appear to be relatively minor in rently well under way—banks can multiply the the eyes of the consumer. convenience offered by their plastic ATM cards without having to purchase additional ATMs. Grocers and convenience store operators Typically, special service corporations estab- recognize this connection, our survey indi- lished by the participating institutions3 adminis- cates. Thus the ATMs in grocery and con- ter the shared networks. venience stores may be paving the way to widespread penetration and acceptance of debit cards, which in turn will become more prevalent in other retail establishments such as "To the customer, it is only a small gas stations and department stores.4 Potentially, step from using a plastic card to debit cards will displace a large number of acquire cash and then groceries to personal checks, because almost three times using the plastic card to purchase as many checks are written for retail purchases as are written for cash acquisition.5 groceries directly." Why are grocery and convenience stores so important in this evolution? Grocers cash a Grocery and convenience stores, meanwhile, tremendous number of checks; in many cases had been watching and waiting for a large base check value exceeds gross sales. They need a of ATM customers to develop. As long as ATMs quicker and cheaper means of negotiating were only accessible by a few people, or as long .such payments and of eliminating bad checks. as a grocer's (proprietary) ATM could only be Convenience stores are more concerned with used by customers of a single financial institution, eliminating currency in cash registers, thereby installation was not justified on anything other discouraging robberies. Both types of . stores than an experimental basis. As the ATM-cash provide a large number of widely dispersed dispenser became widely accepted, however, sites open for long hours. Furthermore, customer concurrent with the evolution of shared ATM traffic is already established and regular. The networks at the local level, grocers could install great majority of their customers are local, with ATMs with the expectation that a significant payments drawn against local financial institu- proportion of their customers would benefit. tions. Because of this combination of character- The survey described here indicates that grocers istics, grocery and convenience stores provide are indeed taking advantage of the new oppor- a good "testing ground" for retail point-of-sale tunity. transactions.

2For a more comprehensive description of this evolution, see "Displacing the Check." this Review, August 1983. See also "Payments in the Financial Services Industry of the 1980s" this Review, December 1982, "The check displacement forecasts for debit cards embodied in "Dis- especially quotes by Peter Merrill that "the financial services industry is placing the Check" are somewhat more aggressive than that of some now shifting into a second phase involving shared delivery systems." other observers because of the expected impact from retailers (p. 18-24, 41-42). 3See "Shared ATM Networks: The Nation and the Southeast," this Review, 5"Displacing the Check," Table 4, p. 32. December 1982.

JANUARY 1984, ECONOMIC REVIEW 32

Digitized for FRASER http://fraser.stlouisfed.org/ Federal Reserve Bank of St. Louis January 1984

Table 1. Stores in Some Stage of ATM Installation

Hours of Operation Transaction Types Handled

SUPERMARKETS

ATMs Already Installed Full-Line' Bruno's Inc., Alabama 24 Hours 24 Hours/Store Full-Line/plus Food Giant, Georgia Traveler's Checks Full-Line Jitney Jungle, Mississippi 24 Hours/Store Full-Line Kroger-Atlanta Division Store Hours Full-Line Kroger-Nashville Division Store Hours Full-Line Publix, Florida 24 Hours Full-Line The Red Food Stores, Tennessee 24 Hours/Stores Full-Line Sunflower Stores, Mississippi 24 Hours/Stores Store Hours Full-Line Winn-Dixie, Florida Finalized Installations Cash Withdrawals Albertson's, Florida 24 Hours Full-Line Grand Union, Florida, Georgia 24 Hours/Store Full-Line Pantry Pride, Florida 24 Hours/Store Definite Plans to Install; Lack Final Commitment Cash Withdrawals Bi-Lo Inc., South Carolina Store Hours Undecided National Supermarkets, Louisiana Undecided Cash Withdrawals Schwegman Giant Stores, Louisiana Store Hours Vague Installation Plans Cash Withdrawals Dixieland Food Stores, Alabama Probably Store Hours Undecided Food Town Stores, North Carolina Undecided Full-Line Harris-Teeter Supermarkets, Store Hours North Carolina Sub-Total: 17 or 63% of 27 Supermarket Chains Surveyed

CONVENIENCE STORES

ATMs Already Installed Full-Line Fast Fare Inc., North Carolina 24 Hours Full-Line Munford Inc., Georgia 24 Hours Full-Line The Pantry Inc., North Carolina 24 Hours Finalized Installation Plans Cash Withdrawals; Little General Stores, Florida 24 Hours/Store Barnett Bank Deposits Only Full-Line Except Deposits Shop & Go Inc., Florida 24 Hours/Store Full-Line Sunshine Jr. Stores, Florida 24 Hours Pilot in Texas 24 Hours Full-Line Except Deposits; National Convenience Stores, Cash Advances on Credit Cards (Shop-N-Go), Georgia 2 Pilot in Philadelphia Full-Line Except Deposits Southland Corporation (7-11 Stores), Store Hours/ Louisiana2 Usually 24 Sub-Total: 6 or 75% of 8 Convenience Store Chains surveyed TOTAL: 23 or 66% of 35 SURVEY PARTICIPANTS

'Full-Line: Deposits, withdrawals, balance inquiries, transfers between accounts ATM installation 2These two convenience store chains are not included in survey totals because they have not commenced in their southeastern stores

FEDERAL RESERVE BANK OF ATLANTA 33

Digitized for FRASER http://fraser.stlouisfed.org/ Federal Reserve Bank of St. Louis January 1984

Survey Results While most of the supermarkets and con- venience chains intend to install only one ATM Our survey, which was conducted in August, per store, the number of ATM locations varies included eight southeastern states—Alabama, widely. Two distinct patterns of installation Florida, Georgia, Louisiana, Mississippi, North emerge. First, there are those grocery and Carolina, South Carolina, and Tennessee. The convenience store chains in which a single thirty-five corporations included in the sample bank installs from one to ten ATMs. Without were culled from a list of the 200 largest exception, the bank owns these machines. supermarket and convenience chains in the Since only that particular bank's customers nation. Under the assumption that the largest may access the machine, it is a proprietary stores usually serve as industry leaders—in network; no one except those possessing the general they are the first to implement inno- bank's proprietary card may make transactions. vations—the survey is reflective of electronic In contrast, seven chains are installing ATMs banking activity occurring in southeastern gro- on a much wider scale, with anywhere from cery and convenience chains. fifty to over one-hundred included in the plans. Twenty-three of the thirty-five chains ques- These ATMs will be part of large regional tioned either have ATMs in their stores or plan networks comprised of many banks. Indeed, to install them within two years. Eleven stores the shared networks represent a large cardholder have already installed ATMs; six stores have base that makes massive implementation of signed contracts to do so within the next six ATMs economically feasible. Thus a dichotomy months. Three stores definitely plan to install exists between plans to install a few ATMs ATMs but lack final commitments. Three other chains intend to install ATMs within the next two years but plans remain somewhat vague at the moment. Furthermore, of the eleven stores "Twenty three of the thirty five without ATM installation plans, one is conducting an extremely preliminary investigation of the chains questioned either have topic. Also, two firms are national chains with ATMs in their stores or plan to ATM pilot programs in other parts of the country. install them within two years" Quite conceivably, ATM installation in their southeastern stores could follow successful pilots.

After ascertaining the number of supermarkets under a single bank's proprietary network and and convenience stores installing ATMs, we the placing of many ATMs participating in a compared various operational details. Of the shared network of regional banks. The situation in ATMs currently operating, all are accessible Florida serves as an excellent example of the twenty-four hours a day, as will be those ATMs latter alternative. currently being installed under contract. ATMs Florida's HONOR network became operational tend to be located in stores open twenty-four on September 30, 1983. Representing approxi- hours a day; this maximizes convenience for mately $50 billion in deposits, 120 financial customers. institutions comprise the HONOR network. All of the presently functioning in-store ATMs The combined total of 3.6 million ATM access offer a full-line of transactions—deposits, cash cards issued by the network approximates 75 withdrawals, transfers between accounts, and percent of Florida's card-holding base. Network balance inquiries. In addition, the American members will have access to 407 off-premise Express Money Stop ATMs located in seven machines, 125 of which belong to the Publix 6 Atlanta Food Giant supermarkets dispense tra- Teller Network. Beginning next June, Honor veler's checks. Of the stores still in the planning network members will share on-premise ATMs, stages, however, four indicated that the ma- creating a totally shared electronic environment chines will be for cash dispensing only. Ap- in Florida. In December 1983, two of Georgia's parently, some banks involved feel that it would not be economical to collect deposits Florida's Honor Racing to Grab the Network Lead,"Bank Network from such widely scattered locations. News, Vol. 2 (September 24, 1983), p. 7.

34 JANUARY 1984, ECONOMIC REVIEW F|

Digitized for FRASER http://fraser.stlouisfed.org/ Federal Reserve Bank of St. Louis January 1984

Table 2. Number of ATMs and Type of Network

Do You View This As a Steppingstone Number Network To Point-of-Sale?

Supermarkets

Albertson's, Florida 7 Shared Yes Bi-Lo, Inc, South Carolina Around 60 Shared Yes

Bruno's Inc., Alabama 2 Proprietary Uncertain Dixieland Food Stores, Alabama Shared Yes

Food Giant Georgia 7 Shared No Food Town Stores Inc., North Carolina Undecided Undecided Yes

Grand Union, Georgia & Florida Around 50 Shared Yes Harris-Teeter Supermarkets Inc., 1 - 10 Undecided Yes North Carolina

Jitney Jungle, Mississippi 1 Proprietary Yes Kroger-Atlanta Division 3 Proprietary Kroger-Nashville Division 6 Proprietary Yes

National Supermarkets, Louisana 3 Proprietary Uncertain Pantry Pride Inc., Florida 50-60 Shared Yes

Publix, Florida 100 - 500 Shared Yes Schwegman Giant Super Stores, Louisiana 11 Undecided No

Sunflower Stores, Inc., Mississippi 2 Proprietary No

The Red Food Stores Inc., Tennessee 97 Shared Yes

Convenience Stores

Fast Fare Inc., North Carolina 2 Proprietary Uncertain

Little General Stores, Florida 20 Shared Yes Munford, Inc., Georgia 1 Proprietary No 1 National Convenience Stores (Stop-N-Go), 81 Shared No Georgia

Shop & Go Inc., Florida 20 Shared Yes 'Southland Corporation (7-11 Stores), 200 Shared Yes Louisiana

Sunshine Jr. Stores, inc, Florida 1 Proprietary Yes The Pantry Inc., North Carolina 1 Proprietary Yes

'These two convenience store chains are not included in survey totals because they have not commenced ATM installation in their southeastern stores.

35 FEDERAL RESERVE BANK OF ATLANTA

Digitized for FRASER http://fraser.stlouisfed.org/ Federal Reserve Bank of St. Louis January 1984

Table 3. Motives Underlying ATM Installation

Convenience 2 3 Supermarkets1 Stores Total

Increase Customer Convenience 13 (48%) 7 (87%) 20 (87%) 7 (30%) Increase Customer Traffic 1 (4%) 6 (48%) Reduce Check Processing Costs 11 (41%) 1 (1%) 12 (52%) Reduce Check Volume 8 (30%) 1 (1%) 9 (39%) 7 (30%) Reduce Bad Checks 7 (26%) —

Reduce Cash Security Problems 2 (7%) — 2 (9%)

Reduce Labor Involved 1 (4%) — 1 (4%)

'Percentage of the 17 supermarket chains pursuing ATM installation. ?Percentage of the 8 convenience store chains pursuing ATM installation. 3Percentage of the 23 stores with plans for ATM installation.

largest financial institutions announced plans these in-store ATMs. For the most part, the to establish a similar network, and invited other Floridians immediately responded that they Georgia financial institutions to participate as acted in self-defense. In announcing its inten- charter members. tions to establish the Publix Teller network in In Atlanta, Kroger is involved in a pilot with 1981, Publix super-markets initiated a mad Heritage Bank. Within the past year, Heritage scramble to install in-store ATMs. Thus many Bank has established branch banks in three of Florida stores plunged into the pursuit of in- KrogeKs stores. These branch locations serve store electronic banking in an effort to remain as full-service banks, offering everything from competitive. loans to checking accounts. Generally branch Although competitive pressure rushed Florida's banks are viewed as "temporary and inadequate stores into installing ATMs, real benefits must either because they are too labor intensive and accrue from these systems in order to justify inefficient or they serve a limited customer such interest. In questioning the supermarkets base."7 Yet Bob Hodge, vice-president of •and convenience store operators, we found KrogeKs Atlanta division, reports that: "The they resoundingly answered that their major customers seem to be very pleased with the goal was to provide customer convenience. additional service. I presume Heritage is doing The grocery industry is traditionally sensitive enough business to justify labor and costs of to the needs of the consumer. Emphasizing the putting banks in."8 It is too early to reach any importance of the customer, the manager of definitive conclusions about the feasibility of financial services at Kroger explains "...our branch banks; however, installing ATMs accessi- highly competitive business is driven exclusively ble by all members of a shared network seem- by the consumer. That is unbelievably critical."9 ingly constitutes a more viable alternative. Twenty of the survey participants listed customer convenience as their primary, overiding concern. As Mike Ware of Little General Stores, a Florida Customer Convenience convenience chain, states, "We view the ATM Of greater interest than details of location as a convenience item and we're in the business of selling convenience."10 With the slim profit and operation are the objectives behind installing

'"Craig Gieler How Kroger Wants POS to Work," Bank Network News, Vol. 1 (January 25, 1983), p. 4. 9"Craig Gieler How Kroger Wants POS to Work," Bank Network News, "Bob Hodge, vice president, Atlanta division, Kroger Co., telephone Vol. 1 (January 25, 1983), p 4 interview, August 26, 1983.

36 JANUARY 1984, ECONOMIC REVIEW F|

Digitized for FRASER http://fraser.stlouisfed.org/ Federal Reserve Bank of St. Louis January 1984

margins characteristic of the grocery industry, Retailers cited four main reasons why in-store retailers seize upon any means of offering ATMs are not feasible in small towns and rural additional customer convenience and gaining a areas. First of all, many of the local banks do not competitive edge. have ATMs. Obviously, on-premise ATMs must Hand-in-hand with the notion of added exist before off-premise ones spread. customer convenience is the desire to increase Even in small towns possessing ATMs, trans- customer traffic in the stores. Retailers hope action volume sufficient to justify installing in- that customers, after entering the store to store ATMs cannot be generated. A Piggly- obtain cash, will purchase a few items. John Wiggly Southern spokesman explains: Polizzi of Florida's Shop-N-Go stores expounds on this: "One of the long-term objectives is to We're basically operating in small towns. increase average transaction size by being a We've had proposals (to install ATMs) but source of cash for the customer."11 Seven other as a matter of company policy turned them survey participants indicated that they expect down because of: (1) customer accep- that easier access to funds will spark impulse tance—customer contact is a major part of buying and strengthen sales. our business. ATMs are too impersonal; (2) not enough transaction volume; (3) minimal Because customer convenience is the primary bad check losses due to the local nature of motivation behind installing in-store ATMs, the business.13 supermarkets and convenience stores insist on Other chains serving small town and rural a shared network of regional banks before communities also cited low anticipated customer adopting a course of widespread installation. acceptance as deterring ATM installation. Even For instance, Carl Schauss of Mississippi's Jitney where customer acceptance is not a problem, Jungle Stores of America, comments that his Bob Hughes of North Carolina's The Pantry Inc. points out that "our type of market area— mostly rural areas—cannot generate enough transactions to make ATMs profitable."14 In- "Retailers hope that customers, store ATMs prove much more feasible in areas after entering the store to obtain supporting denser populations; in more crowded cash, will purchase a few items." areas ATMs truly do boost customer convenience. Because increased customer convenience seems to be the pivotal issue in successful off- premise ATM installation programs, marketing corporation has "no specific plans to install assumes an important role. According to a more ATMs (currently it has one in-store ATM) recent Synergistics Research Corporation study but a definite inclination to make plans in that of off-premise ATMs, 39 percent of the partici- direction except for the single bank mode pants expressd a desire to access ATMs at 15 existing in Mississippi. Without a shared net- supermarkets. If this is the case—if a demand work, ATMs are simply not feasible."12 For for in-store ATMs already exists—then proper electronic banking services to heighten con- marketing should ensure the success of in- venience significantly, the ATMs must be avail- store ATMs. Mike Ware describes the marketing able to a substantial portion of those in the plan of Florida's Little General Stores as "in- region possessing ATM access cards. stalling ATMs at key traffic arteries with high volumes... at areas with a high concentration of The topic of regional networks leads to another apartments and condominimums and no areas pertinent issue, namely that of the region with a predominance of older people. Older involved. Several of the participants claimed people are not concerned with convenience; that they were not interested in installing ATMs there is more of a difference in age than income because of the rural nature of their business.

13Larry Olsen, vice-president and treasurer, Piggly-Wiggly, Southern Inc. ,0 telephone interview, August 22, 1983. Mike Ware, vice-president of finance, Little General Stores, telephone '"Bob Hughes, vice-president of finance, The Pantry Inc., telephone interview, August 23, 1983. interview, August 19, 1983. "John Polizzi, Shop & Go Inc., telephone interview, August 24, 1983. 15 12Carl Schauss, executive vice president of finance, Jitney Jungle Stores "Study Shows Consumers Prefer Shopping Malls for Off-Premise ATMs, of America, telephone interview, August 26, 1983. Bank Letter, Vol. 7 (May 16, 1983), p. 7.

37 FEDERAL RESERVE BANK OF ATLANTA

Digitized for FRASER http://fraser.stlouisfed.org/ Federal Reserve Bank of St. Louis January 1984

as far as targeting ATM installation."16 Others the retail level are written to food retailers.18 In indicated plans to install in higher income 1981, supermarkets averaged 2,786 checks areas. All plan to install ATMs at the stores with per week, leading to a total of 4.2 billion checks the highest volume of traffic. that year. Furthermore, while the number of For those businesses with only one or two checks that grocery stores cash is rising, the 19 bank-owned ATMs, the bank involved assumes average purchase size is decreasing. responsibility for locational "on-the-spot" adver- Consequently, check processing costs are tising. Most supermarkets and convenience impacting retailers quite dramatically. Estimated stores embarking on major ATM installation check handling costs in 1981 averaged 45 plans intend to share advertising responsibilities cents. Thus a store typically spent $1,250 per with the network or bank owning the machines. week on check cashing. Since the average store In Florida, for example, both the HONOR nets $150,000 each week, checking costs ap- system and Winn-Dixie will advertise. In addition, proximated 0.83 percent of sales. For close to a Barnett Bank and NCR will help advertise for decade, supermarkets' net margins have been those stores joining their switch. Thus retailers, about one percent Check processing costs, bankers, and other network participants derive therefore, nearly equal the supermarkets' net mutual benefits from each others marketing margins.20 campaigns. It is very evident why retailers want to reduce the volume of checks cashed. Not only have retailers had to pay for the labor and processing Reduced Check Processing costs involved, they also have to pay a fee for Volume and Costs each check deposited. As the astronomical and somewhat inequitably distributed costs of check While customer convenience undoubtedly serves as the chief reason for installing off- premise ATMs, most grocery and convenience store chains hope reduced check processing "In 1981, supermarkets averaged volume and costs will result as a by-product. Indeed, this is a topic of pressing concern 2,786 checks per week....Estimated among retailers. A spokesman for Warehouse check handling costs in 1981 Groceries Management in Alabama says, "as far averaged 45 cents" as checks are concerned, something has got to happen because it's a burden and a costly one."17 Twelve survey participants listed re- ducing the costs of check processing/cashing processing continue to rise, there is a "new as a very important objective in ATM installation; •mood of retailers who insist that the checking nine cited the necessity of decreasing the huge burden has become so great that a new payments check processing volume (which greatly compli- process at the supermarket is in order."21 And cates turning checks into collected funds). in-store ATMs comprise one step along the way Besides accepting checks for purchases, to this goal. retailers also serve in a cash dispensing capacity In fact, supermarket ATMs represent some- after bank hours. For the most part, convenience what of a role reversal between banks and stores do not accept checks. The majority of retailers. Instead of stores absorbing processing the grocers surveyed, on the other hand, be- costs and paying the bank deposit fees, the wailed the extent to which they act as surrogate bank pays the store a rental fee for the space in banks in cashing checks. Twelve claimed to which the ATM is located. Grocery and con- cash many more checks than the local banks. venience stores receive rental fees; the owner In truth, it is acknowledged in the grocery of the ATM and the operator of the switch split industry that the value of checks cashed in many stores or chains exceeds the total annual '»"Craig Gielen How Kroger Wants POS to Work," Bank Network News, sales. Seventy percent of all checks written at Vol. 1 (January 25, 1983), p. 4. ''"Grocery Check Volume Soars, Reports FM I" Bank Network News, Vol. 1 (June 21, 1982), p. 1, 3. 20 ,6Mike Ware, Little General Stores. Ibid. 2 "Roger Dryer, executive vice-president Warehouse Groceries Manage- '"Craig Gielen How Kroger Wants POS to Work." Bank Network News, ment, telephone interview, August 22, 1983. Vol. 1 (January 25, 1983), p. 4.

38 JANUARY 1984, ECONOMIC REVIEW F|

Digitized for FRASER http://fraser.stlouisfed.org/ Federal Reserve Bank of St. Louis January 1984

the transaction fee. Often, the transaction fees installing ATMs. Convenience stores experience are volume-related; that is, the fee per transaction minimal bad check losses because they cash depends upon the amount of monthly trans- few checks. Grocery stores, however, experi- action volume. For instance, in Florida's HONOR ence enormous bad check losses. Given this network, if there are fewer than 500,000 trans- fact, it is somewhat surprising that only nine actions per month, a withdrawal costs 61 cents. companies reported any type of electronic 40 cents will go to the ATM owner; the switch check verification systems. itself will receive 21 cents. Withdrawals cost Several other potential benefits were men- only 55 cents if more than 1.5 million monthly tioned in association with in-store ATMs. Two transactions occur. Balance inquiries and transfers of those surveyed cited increased security of between accounts cost 20 cents less.22 The cash as a result of decreased cash exposure. bank of the customer whose account was Convenience stores in particular have a problem debited pays these fees; it may in turn pass the with hold-ups. And one participant felt that in- charge on to the customer. But the retailer store ATMs would reduce the labor required does not pay anyone. for check cashing and processing. Furthermore, if the supermarket or conve- nience store owns the machines, it will receive A Steppingstone to POS the transaction fees and begin to generate profits. Publix Supermarkets in Florida however, Thus, increased customer convenience is is the only chain in the survey to own its own the primary motive for installing in-store ATMs. ATMs. For the most part, the retailers are Stores hope that a reduced check volume will content to let the banks or third party networks produce various benefits. These, however, are own the ATMs. Yet if the banks or networks do only the short-term objectives. In the long run, not provide sufficient customer coverage, stores installing in-store ATMs may be an intermediate will purchase their own machines. For instance, step in a changing retail payment mechanism. in Mississippi, where the banks have thus far In-store automated teller machines represent not collaborated, Carl Schauss of Jitney Jungle one stage in the transition to electronic point- Stores says "owning our own machines is a of-sale (POS) registers. major possibility if the banks don't get with it Ultimately, most retailers hope to have POS (network formation) themselves."23 Apparently, registers that electronically debit the customers ATM ownership is not critical to the retailer bank account at the check-out station. When unless inadequate network coverage exists. asked if they viewed in-store ATMs as a stepping- Although reducing the check volume is a stone to POS, seventeen of the survey participants hoped for by-product of in-store ATMs, chains responded yes. Some, such as Harry Wade of stated that they doubt the number of checks Winn-Dixie, even implied that ATMs were cashed will decrease. Joe Letvelter of Pantry- being installed solely as an intermediary step: Pride sums up this attitude: "we don't feel that "ATMs are being installed as a training ground ATMs will reduce check cashing... From talking to get customers to use debit cards. They 25 to NCR and Publix(in Florida), the check volume (ATMs) are a necessary evil." Others said that doesn't decrease... the costs associated with ATMs "give us definitive experience in cash- 26 check processing remain the same."24 Certainly free types of transactions." John Polizzi of this is contrary to the expected results. It is too Shop-N-Go Inc. sums up the general attitude: soon to determine accurately what effects in- ATMs are going to work in certain locations store ATMs actually have on check volume. Yet but won't have the impact that POS will. even if they have zero impact on the number of Ultimately POS will do away with check checks cashed, they will still be of value in cashing. ATMs are simply another service we 27 preventing bad check losses. offer our customers. Obviously, strong support exists for the eventual Seven of the survey participants listed re- implementation of POS registers. duction of bad checks as an objective behind

"Strada, "Converting FIG's Promise Into Power," Bank Network News, "Harry Wade, director of work methods, Winn-Dixie, telephone interview, August 22, 1983. Vol. 1 (May 11, 1983), p. 5 26 "Carl Schauss, Jitney Jungle Stores Ray Ayers. vice-president of real estate, Grand Union Company, tele- 24 phone interview, August 22, 1983. Joe Letvelter, cash manager, Pantry Pride Ine, telephone interview, ''John Polizzi, Shop & Go Inc. August 23, 1983.

39 FEDERAL RESERVE BANK OF ATLANTA

Digitized for FRASER http://fraser.stlouisfed.org/ Federal Reserve Bank of St. Louis January 1984

List of Stores Surveyed Louisiana

Alabama 19. Schwegman Giant Super Stores New Orleans * 20. Southland Corporation, Midsouth Division - 1. Bruno's Inc., Birmingham Baton Rouge 2. Warehouse Groceries Management - Gadsden 21. National Team Company (National Super- 3. Dixieland Food Stores - Geneva markets) - Harahan (National Headquarters in 4. Delchamps, Inc. - Mobile Rosemont, IL) Florida Mississippi 5. Winn-Dixie - Jacksonville 22. Jitney Jungle Stores of America - Jackson 6. Publix - Lakeland 23. Sunflower Stores Inc. - Indianola 7. Pantry Pride Inc. - Fort Lauderdale North Carolina * 8. Shop & Go, Inc. - Mango * 9. Little General Stores - Tampa 24. Food Town Stores, Inc. - Salisbury * 10. Sunshine Jr. Stores Inc. - Panama City 25. Harris-Teeter Supermarkets Inc. - Charlotte 11. Albertson's - Orlando (Southco Division * 26. The Pantry Inc. - Sanford Headquarters) 27. Ingle's Market, Inc. - Swannanoa Georgia 28. Lowe's Food Stores - Wilkesboro * 30. Fast Fare, Inc. - Henderson 12. Great Atlanta & Pacific Tea Co. - National South Carolina Headquarters in Montvale, New Jersey

13. Food Giant/Big Apple Supermarkets - Atlanta 31. Bi-Lo Inc. - Mauldin * 15. Munford, Inc. - Atlanta 32. Community Cash Stores, Inc. - Spartanburg * 16. National Convenience Stores (Stop-n-Go) - Tennessee Decatur (National Headquarters In Houston) 33. The Red Food Stores, Inc. - Chattanooga 17. Grand Union Company (Colonial and Big Star 34. Malone & Hyde, Inc. - Memphis Food Stores - Decatur) 35. The White Stores, Inc. - Knoxille

18. Kroger - Atlanta Division 36. Kroger Company - Nashville

'Indicates convenience store chains. The two different divisions of Kroger are counted as one company, leading to a total of 35 survey participants

While most retailers concede that a tran- instance, one supermarket chain in Texas found sition to POS is inevitable—indeed, they wel- that it could "save one-sixth of labor costs at come such a transition—few have made any check-out due to the time saved by direct- concrete moves in that direction. In fact, out of debit transaction."28 And, without a doubt, this 35 surveyed, only three have any definite plans is the most effective way to prevent check regarding POS. Seven claimed to have vague fraud or bad check losses. Even more important plans. are the savings in check handling costs. Naturally, the same benefits resulting from reduced check volume after in-store ATM instal- 28"Texas POS Rekindles a Fiery Network Feud," Bank Network News, VoL lation will also accrue from POS registers. For 2 (June 25, 1983), p. 4

40 JANUARY 1984, ECONOMIC REVIEW F|

Digitized for FRASER http://fraser.stlouisfed.org/ Federal Reserve Bank of St. Louis January 1984

Currently established POS systems charge the 75 percent saturation of the debit card- transaction fees averaging 20 cents. The super- holding market believed essential for an eco- markets and convenience stores do not have nomical POS system. Extensive local networks, any transaction costs but must pay for installation in conjunction with increasing consumer aware- and maintenance.(lt should be noted that where- ness of electronic payment systems, are paving as most in-store ATMs are bank or network the way for electronic point-of-sale. Thus, the owned, almost all stores plan to purchase their current installation of in-store ATMs—in fur- own POS registers.) According to a 1980 survey, thering the development of an electronic infra- the handling cost of a retail cash transaction is structure and in familiarizing customers with 45 cents.29 Thus POS is proving to be more automated banking services—is a key factor in economical than cash. the upcoming transition. The advantages to retailers of POS systems We found a definite trend toward the instal- are obvious. But we must not forget that the lation of in-store ATMs in supermarket and grocery industry is consumer-driven. POS will convenience store chains in the Southeast. The be implemented if and only if the consumer primary motivation lies in increasing customer finds it beneficial. Therefore, the quicker check- convenience, which retailers hope will increase out times of POS registers may be a decisive customer traffic through their stores. Con- factor in its favor. sequently, grocery stores are beginning to insist According to a 1983 Food Marketing Institute that local banks cooperate in forming regional survey, 39 percent of consumers place greatest networks; only in this manner will a given priority on a quick check-out Low prices rank store's ATM be available to a sufficient quantity of second.30 A typical POS transaction takes ten consumers to increase convenience significantly. to fifteen seconds, cash transactions require 27 As long as the banks prove fairly cooperative seconds, and checks require even longer. Thus in establishing these shared networks, the super- POS registers transact payment with maximal markets and convenience stores should have speed. At Dahl's, an Iowa supermarket that pioneered the use of POS registers, the added convenience alone converted 10-15 percent "Evidence suggests that a of check writers to POS without any promotion or financial incentives. This 10-15 percent com- substantial portion of consumers pensated for installation costs.31 Thus consumers will opt for quicker check-outs. face a tradeoff between increased convenience Those unwilling to give up float will or relinquished check float Evidence suggests have to be lured with economic that a substantial portion of consumers will opt incentives and proper marketing for quicker check-outs. Those unwilling to give strategies." up float will have to be lured with economic incentives and proper marketing strategies.

little inclination to own the ATMs or to operate Conclusion the electronic switch. Their concern appears not to be with controlling the payment system, In the foreseeable future, the implementation but rather with modernizing traditional pay- of electronic point-of-sale registers will trans- ment mechanisms. form the retail payment mechanism. Numerous Indeed, the traditional retail payment system grocery stores are installing electronic scanning has become a burden. Supermarket chains equipment at the check-out station, thus greatly spend exorbitant sums on check processing facilitating the conversion to electronic POS costs; the sheer volume of checks supermarkets registers. In addition, the ongoing formation of process renders the system inefficient. Across shared regional bank-card networks is creating the board, grocery industry management agrees on the necessity of reducing the check volume.

""Cost of a Cash Transaction Put at 45 Cents," EFT Report, Vol. 6 (January Many of the survey participants hope that 3, 1983), p. 5. customers will obtain cash from in-store ATMs, 30" Do Shoppers Want POS? FMI Study Gives Clue," Bank Network News, Vol. 2 (June 25, 1983), p. 5. cutting down on the number of checks written 3 '"Why Iowa's POS Test Remains in a Pilot Mode,' Bank Network News, both for groceries and for cash at courtesy Vol. 1 (February 8, 1983), p. 6.

41 FEDERAL RESERVE BANK OF ATLANTA

Digitized for FRASER http://fraser.stlouisfed.org/ Federal Reserve Bank of St. Louis January 1984

desks. However, even if the ATMs do not The spreading electronic infrastructure, accom- substantially reduce the volume of checks, panied by the formation of large regional net- they acclimate the consumer to the idea of works, provides the backdrop for the coming electronically debiting his checking account at of electronic retail payments. The grocery indus- the grocery store. In the future, debiting accounts try—experiencing acute problems with check by an electronic POS register may not seem processing—will undoubtedly be the first to such a sudden, drastic change. make the transition to electronic POS, just as it Hence, in-store ATMs, while generating bene- has been the first to implement wide-scale fits in and of themselves, ultimately derive ATM installation. Other retailers will convert their greatest value as a steppingstone to POS. more slowly. Electronic POS registers will dramatically reduce Generally, the size of the retail business is the volume and costs of check processing. POS directly proportional to the volume of checks also is an effective way of minimizing bad cashed, which in turn directly relates to the check losses. In addition, electronic POS registers need to convert to POS. In other words, larger greatly speed up the check-out process, which retail businesses, such as major department seems to be the most critical factor in improving stores, will likely find it advantageous to convert customer convenience. Thus, electronic POS to electronic POS terminals as fast as possible. holds multiple benefits for retailers and con- Small specialty stores, on the other hand, may sumers. never experience any problems with the existing Indeed, many retailers eagerly welcome the check-collection system. Thus, the rate at which appearance of electronic POS systems. As Leo the transition to electronic POS occurs will vary Conlan of South Carolina's Bi-Lo Stores empha- greatly. Yet, the burden of check processing on sizes, "banks and retailers want to encourage some retailers has become so great that a new consumers to quit using checks... It is about retail payment mechanism seems to be in order. time it (POS) is coming to this part of the Electronic point-of-sale appears to be the most country."32 Some legal matters remain to be likely candidate; indeed, the growing popularity clarified, but for the most part, POS technology of in-store ATMs indicates that the transition has been refined and seems about to come of may already be underway. age.

— Helen Stacey "Leo Conlan, Financial Vice President Bi-Lo Inc., telephone interview, August 19, 1983. and William N. Cox

42 JANUARY 1984, ECONOMIC REVIEW F|

Digitized for FRASER http://fraser.stlouisfed.org/ Federal Reserve Bank of St. Louis