1.0 IntroductionProviding maximum on-shelf-availability (OSA) and minimizing out-of-stock (OOS) are thecritical issues faced by retail grocery stores in a fast moving dynamic environment. They cancontribute to hidden cost for retail shop business, for example loss of sales and customersloyalty. However, the importance of preventing OOS can be easily overlooked. In their studies(Fernie and Corcoran, 2011) show that six of the seven retailers does not take action althoughthey had at least one process available to solve the OOS while only those who had performedindependent OOS/OSA research are pro-active in solving the issue. This shows the importanceof reducing OOS as a potential measure to increase sales is underestimated. Fernie and Corcoran(2011) have shown that House of Fraser was able to drive a 9% increase in monthly salesthrough liaising with store representatives on OOS/ OSA initiatives. Wide varieties of good isprovided can potentially improve sales, since different customers have own preference of sameproducts of different range. However, increase products variation causes complexity to thereplenishment of OOS. An increased in products variability will also reduce shelf space for fastmoving products. Therefore, the balance between high level of OSA, good product varieties andminimal back store inventory needed to be met. This article consists of survey at Tesco and Coop,issues that cause OOS and tools to provide high level of OSA.2.0 Survey at Tesco and Coop Myrtle StreetIn this section, high variation and unpredictable demand grocery shops in Liverpool areexamined. An OOS survey is conducted at Tesco and Coop along Myrtle Street. Ten OOS itemsare recorded for each store for similar period for example after 7pm each day. The data isrecorded for a total period of 5 days. The reason for selecting this two grocery shops is they aresimilar size (medium size) and are located along the same street which indicates they sharesimilar customer group, for example students and employees around the university area. OOSitems are categorized and the recorded OOS amount is tabulated. Table 1 shows the summary ofthe results.The result show the highest amount of recorded OSS are among fast moving items such asbakery food, food ingredients and ready meal. This scenario is observed for both Tesco and Coop.These OOS are fresh items which have close expiry date (within one week) and wouldexposed to the risk of wastage if the demand is low, however if there is a high demand, loss ofsales is induced. Some other items like fruits and vegetables also show some amount of OSSduring the survey period. The following section discusses the issues that might cause productsOSS.3.0 Issues of Supply Chain that cause the products to be out-of-stockThis section investigates the possible causes that lead to out-of-stock in Tesco. Figure 1illustrated in studies (Fernie and Grant, 2008, pp. 296) shows a total of 65% (shelf replenishmentin shop, retailer distribution centre and inventory inaccuracy) is accounted as retailerresponsibility to OOS issue. This indicates the retail company with correct implementation ofsupply chain management techniques can improve the OOS situation. This section focus on thethree major sources of OOS as highlighted in figure 1: Shelf replenishment efficiency, supplychain efficiency and operational accuracy. They are shown in section 3.1, 3.2 and 3.3.Figure 1: root cost analysis of retail out-of stocks3.1 Operational AccuracyIn order to achieve accurate demand forecast, it is necessary for retailer to maintain an accuratepoint of sales (POS) and current inventory data. Inaccuracy in these data causes incorrectforecast and can lead to OOS. Another problem is shrinkage due to thefts or poorly trained staffs,both of this can account to the inaccuracy sales data being transmitted to the supplier. Dataaccuracy is important because it ensures true customer demand information is being sent to theupstream of supply chain to prevent poor ordering practices that lead to OOS. Examples of poorordering are insufficient order, insufficient inventory to fulfil cycle demand before the nextscheduled replenishment or late order. Moussaoui et al. (2016) suggests that poor ordering can bedue to a variety of system failures such as order transmission problems and inadequate inventorymodelling with less-than-optimal reorder points. Therefore, without the full commitment of thestaffs, high level of operational accuracy cannot be achieved and this will lead to OOS problem.3.2 Shelf replenishment efficiencyExcessive backroom inventories are preventing transparency of inventory and will lead to OOS.Allocation of inventory in two different locations will increase operational complexity. StoringManufacturerdelivery to DC,30%Retailer DC toshop, 15%Shelfreplenishmentin shop, 35%Inventoryaccuracy, 15%Other, 5%inventories in the backroom substantially changes the nature of the optimization problem facedby retailers, which we call the backroom effect (BRE) (Eroglu, Williams and Waller, 2013). Forexample, products that have high demand velocity but with relatively little shelf space are verysusceptible to OOS. Backroom effect occurs when high inventory level obstructs items to belocated easily and lead to inaccurate ordering. This will prevent efficient execution of on-shelfreplenishment process. Moussaoui et al. (2016) claimed that backrooms are poor substitutes fordelivery trucks to replenish retail shelves and higher store inventory-levels exacerbate OOS duemainly to poorly reliable backroom operations. Eroglu, Williams and Walker (2013) listed twoother disadvantages of excessive backroom inventory, which are an increase in management costrequired to monitor the backroom inventories, and forgotten inventory in the backroom willcause inventory record inaccuracies (IRI). As mentioned earlier, record inaccuracies will preventoptimal reorder quantity to be achieved. Therefore, with high level of backroom inventories,OOS can be induced.3.3 Supply chain efficiencyPoor communication between upstream (suppliers) and downstream (retailer) can lead to supplychain inefficiency for example delivery delay, insufficient supplies of products and incorrectitems being delivered. As a result, OOS is produced. Moreover, a small change in customerdemand results in large variation in supplier production quantity. This sudden change createsvariance between actual customers demand and various suppliers’ production quantities alongthe supply chain (Bullwhip effect). This gives rise to a number of well-known problems: theaccumulation of excess inventories at certain times, followed by serious inventory shortages(Machuca and Barajas, 2004). Due to this inefficiency in supply chain, hazard of OOS will be anegative by product of Bullwhip effect even with buffer stock available.4.0 Tools to provide high level of on-shelf availabilityWith today’s advancement in technology, computerized systems are used to monitor the stocklevel or even automatically release the purchase order to the suppliers on a continuous basisinstead of daily basis. For example, an estimation of daily demand based on POS data is essentialfor everyday replenishment. The “item velocity monitor” developed by Date Ventures andProcter & Gamble predicts with 90 per cent accuracy the out-of-stock status for items that movefour or more times per day. This can provide a real-time signal to store managers and does notdepend on store inventory records (Corsten and Gruen, 2003). Advanced technology tools haveemerged to automate out-of-stock measurement and detection for example Sainsbury’s hasintroduced an automatic “shelf availability monitor” (SAM) to track the sales transaction date(rather than the inventory) for a store’s top 2,000 products and highlight items that may be out ofstock (Corsten and Gruen, 2003). For the restoration of replenishment system and supply chainefficiency, three OOS reducing tools are suggested in section 4.1, 4.2 and 220.127.116.11 Radio Frequency Identification (RFID)RFID system can register demand velocity and can be used to identify the fast moving item.RFID can also store data such as price, expiry date and weight. These data can be useful for thestore manager when making decision for reordering. The scanning is fast and efficient andlocation of scanned items can be identified in the backroom at any time. It is an economicalsolution to ensure data accuracy. Excess motion is eliminated when searching for the missingitem thus reducing waste in motion according to Kaizen principle. Corsten and Gruen (2003)provided an example that Procter & Gamble and SAP are jointly trialing the use of radiofrequencyidentification (RFID) transponders, according to them, “intelligent tags” will beattached to each stock-keeping unit, providing truly accurate inventory control.The implementation of RFID will be magnified with continuous participation of retailer to makereplenishment at frequent interval to meet dynamic customers demand. However, application ofRFID solely is insufficient. Maximum benefits of RFID may only be achieved if the retailer isable to make parallel decisions on optimal shelf space allocation and to implement replenishmentprocesses that are not performed periodically but are flexibly adapted to customer demand(Condea, Thiesse and Fleisch, 2012). The integrated nature of Vendor Managed Inventory (VMI)(discussed in section 4.2) required absolute data accuracy. Incorrect data can cause dominoeffect throughout supply chain. Therefore, the top priority is to make all staffs aware about theirresponsibility to record sales data accurately to ensure successful implementation of VMI. As aresult, retailer operational accuracy and OOS reduction can be achieved.4.2 Vendor Managed Inventory (VMI)VMI is one of the integration practices that can contribute to reduce inventory in the supplychain. Integration between supplier and retailer to collaborate on demand information sharingsuch that supplier share responsibility to ensure true demand is met and maximizing on-shelfdelivery. When there are fluctuations in demand, it is necessary to provide vertical integration ofupstream supplier in inventory management of the downstream retailer in supply chain. VMIprograms allow consumer demand information to be disseminated up the supply chain, thusmitigating upstream demand fluctuations due to the bullwhip effect (Escuín, Polo and Ciprés,2017).This synchronized system improves OSA by speeding up shelf restocking and timely delivery.Moreover, backroom inventory can be kept at minimal level. Employees can have better controlover OSA thus optimizing on-shelf replenishment to prevent OOS. However, during promotionalevents and sudden surge of demand, communication between vendor and retailer aboutproduction adjustment is required. Inventory sizing according to their demand velocity is utmostimportant in dynamic fast moving environment to provide guaranteed service level. This can beachieved with accurate anticipation of vendor through VMI. Overall, VMI may allow reducinglogistics and manufacturing costs, reducing overall lead-times, improving service level andreducing transportation costs (Escuín, Polo and Ciprés, 2017). Therefore, VMI helps to improveshelf replenishment efficiency.4.3 Electronic Data Interchange (EDI)With the accurate data provided by RFID, purchase signal is sent through EDI to the supplier.This signal carries instructions about the amount and list of items so that supplier can beginpreparing for the next delivery cycle. EDI speeds up order process thus shorten time from orderto products delivery through electronic business. EDI significantly reduce inventory and matchesdemand quantity to supply quantity by making smaller batches and higher frequency order.Corsten and Gruen (2003) confirm that batching orders disrupt the products’ flow to the shelfand causes the “bullwhip effect” throughout the supply chain. To address this, many retailershave already increased their ordering frequency, implemented EDI and internet ordering,introduced mixed truckloads, adapted minimum pack sizes, reworked delivery schedules andautomated ordering to break batches. As a result, bullwhip effect is reduced and OOS problem isminimized.Pull system can be enforced, for example based on POS data, the forecast demand can begenerated. Store manager can then make stock replenishment decision based on the dataavailable and their previous experience. Waste of overproduction can be reduced by just-in-time(JIT) techniques. Furthermore, Machuca and Barajas (2004) suggest the use of EDI not onlybenefits companies by reducing the costs of order processing, more frequent replenishment ofmaterial requirements in smaller batches reduces distortions in demand information, with asubsequent reduction in the bullwhip effect and the excessive costs it generates. Overall, theenhanced supply chain efficiency will lead to better control of OSA and therefore reducing OOS.5.0 ConclusionOOS and OSA are important measures in retail grocery business’s performance and are affectedby various factors that cause it to deviate from the equilibrium. Three sources causing OSS to thegrocery store are identified as data inaccuracy, backroom effect and bullwhip effect. Varioussolutions are being suggested to solve this problem for examples, maintaining accuracy of data,collaboration with the supplier and effective communication of purchase orders. The toolssuggested to rectify the problem are RFID, VMI and EDI. With their implementation, identifyingof OOS and reordering process can be automated by computerized system. These tools canfacilitate the highest level of OSA. Both retailer and workers must have commitment towardsdata accuracy and flexible shelf allocation according to customer demand patterns. The supplychain efficiency can be achieved thus enhancing the OSA while reducing OOS. Future researchercan seek for the potential improvements to the combinations of these tools to provide a supplychain optimization such that it’s function is not limited by only reducing the cost of loss sales butalso increasing the sales performance of the retail grocery store overall with possibly faster shelfreplenishmentsystem and customer check out time.