Data Science and Machine Learning Approach to Improve Online Grocery Store Sales Performance

International Research Journal of Engineering and Technology (IRJET)

Volume: 09 Issue: 05 | May 2022 www.irjet.net

e ISSN: 2395 0056

p ISSN: 2395 0072

Data Science and Machine Learning Approach to Improve Online Grocery Store Sales Performance

Abhiram Nallamekala(11801535)1, Subbareddy Vanukuri(11801547)2, Dr. Om Prakash Yadav3

1,2Student, School of Computer Science Engineer

3 Assistant Professor (UID: 26121), School of Computer Science Engineer. Lovely Professional University, Jalandhar Delhi, Grand Trunk Rd, Phagwara, Punjab 144001. country

***

Abstract Data science, artificial intelligence, machine learning, and advanced computing techniques became the catalyst to expand the business and provided a venue to do electronic commerce to reach and interact with customers. There is a need to improve and enhance the sales performance for grocery Store The data science and machine learning approach with handling big data with the transparency of data records using computing techniques to enhance e commerce is one of the demanding areas of research these days. This paper discusses the data science and machine learning approaches to improve grocery sales performance. This paper can give an idea how can we achieve it. The method can also help in saving the time for customer who visits the store and helps him in tracking the groceries present in the store. It contains the data which is analyzed using data science where it has been designed in a way the data can be collected in faster way. operation can be done quickly.

grocery’sweuseinourreallife.Thishelpsthecustomersto buythegroceriesinfingertipswithaeasilyunderstandable user interface, any person very new to the website can understandthenavigationeasilyandcorrectly.Thewebsite ismadeusingPHP,HTML5Stackandusingwampserverto simulate the MySQL server and establish the connection between the frontend and backend. The user friendly interfaceisveryeasytounderstandbyanyuserwhovisits thewebsite

1.1 DATA SCIENCE, MACHINE LEARNING

1. INTRODUCTION

Thegrocerydeliveryisahugeopportunityforbusinessesto cashoncustomercomfortoforderinggroceriesfromhome. There’shugescopeofmachinelearninginthisspace.Alone in India there is Flipkart, Amazon, Jiomart, Swiggy, big basketandotherswhoarebettingongroceryservicespace. Online shopping is becoming increasingly popular for a varietyofreasons.Therearecertainlyexternalfactorssuch as rising fuel prices, difficulty accessing traditional stores and the problems that are often associated with supermarketsandothergeneralstorestocontributetothe increaseinonlineshoppinginterest.Consumerscanfindout moreabouttheproductanditsupdatesfromexistingusers. Whensomeonewantstobuyaproduct,theynolongerjust askfriendsandfamilybecausetherearesomanyproducts reviewedonthewebthatgivefeedbacktoexistingusersof theproduct.

Onlineshoppingsitescontainawidevarietyofhigh quality and soft quality products that keep in mind people. The projectisawebapplicationwhichisusedtoorderthedaily

The Data Science (DS), Machine Learning (ML) approaches are indeed required to scientifically develop methods,processes,algorithmswithsoftwareapplicationsto extract purposeful and timely information from user providedstructuredandunstructureddatacollectedfrome commerce websites. The expertise of data mining and big dataisalsousedtoanalyzebusinessandmarkettrendsona temporal scale. It also unifies data engineering, statistical methods, machine learning algorithms and programming procedures. The strong knowledge and expertise of mathematics,statistics,informationsciences,andcomputer science with artificial intelligence is required and much useful. “Machine learning” is the sub field of “Computer Science”i.e.itcomesundertheareaof“ArtificialIntelligence”, whichisthescientificstudyofalgorithmsdevelopedunder stochastictheorywhichiseffectivetoperformtaskswithout havingexplicitprograminstructions,basedoninferenceand relyingonpatterns.Themathematicalmodelofsampledata i.e.“TrainingData”isbuiltviamachinelearningmethodsto makedecisionsondataandpredictionsaredrawn.:

1.2 LOCALIZATION & ZIP CODE LEVEL CUSTOMIZATION

Thebigshiftinfoodretailishowthelocalfoodmovement hasstartedtochangetheassortmentlandscape.Increasingly moregroceryshoppersarereturningtotheirroots,seeking local food, supporting local farmers to reduce the environmental impact of transporting food from long distances.Assortmentsolutionsneedtoenablelocalstore managers to source, validate and test local products essentiallysupportingacuratedofferofallthatislocal and integratethiswithamoreefficientcentralassortmentwhich

IRJET | Impact Factor value: 7.529 | ISO 9001:2008

International Research Journal of Engineering and Technology (IRJET)

Volume: 09 Issue: 05 | May 2022 www.irjet.net

takesintoaccountthenumberofcustomerswhowillorder onlineandthereforedonotneedtochooseandpurchasethe product instore. In short, it requires a blend of the data driven, machine assisted recommendations and human approach to truly enable grocery retailers to deliver at a local level. For these reasons, the need for zip code level trackingofpriceandassortmentisparamountforahighly customized assortment. Retailers can no longer have a centralassortmentstrategyfortheentiremarket.

Datasciencehelpstheretailerstofindthebestplacesfor implementingnewstoresforsellingtheirproducts.Ituses the decisions of the customer concerning area, for this analysis,thereisalargeamountofdataisrequired.Suchas the customer data available online, market trends in that area,thelocationoftheothernearbyshops,etc.Usingdata science, geospatial analysis, and machine learning techniques,thisprojectaimstoprovideasolutionforthis problem and recommending the best neighbourhood for openingthenewstore.

A quick summary of problems which this grocery store addresses using data science and Machine Learning solutions:

1. maximizingthecustomerconversionratethrough advertisement and targeting customers to right grocery itemsorstoreswouldgeneraterevenueforowner.

2. Boostingthetraffictoaspecificstoreforvisibility alsoincreasesInstacartrevenue.

3. It would also need to know how many customers repeatcustomersarewhichpurchaseonregularbasis.

4 Thereishighvarianceinpredictingthetimeittakes to deliver an order based on various factors like weather, roadcondition,rushatstore,howlongwillittakeforstore topacktheorderandavailabledeliveryagentsetc.

e ISSN: 2395 0056

p ISSN: 2395 0072

5. Itisalsobuildingitsownmodelfortraveltimefor deliveryagenttodeliverordertospecificcustomer,asmost ofordersarerepeatordersandthereishighpossibility.

6. Recommending items dynamically to the user (Which he/she might be interested) would also increase conversion.

2. DATABASE TABLE SCHEMA

Schema 1: DatabaseSchema

3. ALGORITHM APPROACH

Someapparentchallengesasshowninbelowfigure.Thatthe onlinegrocerystoreisfacingare

1. Paymentsecurityissuesinonlinegrocerystore

2. Handlinglogisticsanddistributions.

3. Lowersalesconversationonsite.

4. Also,theWarehouseManagement.

Chart 1: AlgorithmFlowchart

• As, above figure no Payment security in online grocerystorehasalwaysbeenunderthespotlight.Although the payment gateways and processors have outgrown in terms of their subscriber base and average transaction

© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal

International Research Journal of Engineering and Technology (IRJET)

Volume: 09 Issue: 05 | May 2022 www.irjet.net

volume, there is still a deficiency of secure and stable securitysystemsespeciallyindevelopingmarkets.

• One of the biggest problems that e commerce businessesarefacingistheefficienthandlingoflogisticsand thedistributionsnetwork.Mostlogisticscompaniesarenot furnished to provide services to modern e commerce businesses.Therefore,manynewe commercestoresfailto meettherequirementsofhighefficiencyresultinginpoor on timedelivery.

• Theconversionrateismetricwhichiscalculatedas the percentage of the number of users who process the completepurchaseone commercewebsiteamongthetotal numberofvisitorsofthatsite.”Forexample,www.daraz.pk, a populare commercewebsitevisited by1000customers duringaparticularmonththatisMarchtobuyaperfume, butonly60peopleprocessedthecompletepurchase,hence thesiteconversionratebecame60/1000equalto6%.

• Anotherviewofasuccessful“SalesConversion”is oneinwhichtheuserfollowsthe“Call to Action”(CTA)and completesthestepsthatwererequiredtobeperformed.For example, to get more email newsletter subscribers for a brand,a visitorona websitemusthaveto enterhis email address and name. Each successful subscription would be countedasa“conversion”inthiscase.

• Thisisoneofthemostchallengingareasforeverye commercewebsite,especiallythenewonesinthemarket, that although receive decent traffic, some thousands of monthlyvisitorstothewebsite,butstillonlyaminorityof thosecompletethepurchaseandshowaconversion.

• when it comes to online advertising such as FacebookAdsorGoogleAdvertising,specificallyforthee commerce businesses,sales conversion isoneofthemost importantmetrici.e.,“KeyPerformanceIndicator”(KPI)to analyzetheuserbehavior,engagements,andpurchases.All these metrics assist in determining the final return on investment(ROI)fromonlineadvertisingforthecompany.

OnesimplycannotignoretheROIfactorasitisthejudging criteria to decide whether a marketing channel should continuetobeusedforbrandawarenessornot.Let’ssaya smallamountof$80foradaywasspentonadsthatresulted in total of 16 sales on the e commerce website with an averagecartsizeof$150.

Thiscouldbeconsideredasaprofitablemerchandisingor advertising campaign that gives the return on investment andthecompanywouldcontinuetoleveragethismarketing channelwithhigheradsspent.Alternatively,however,the samee commercecompanywouldtakeiftheyspent$1000 forsevendaysonthesameadcampaignbutfailtogettheir return on investment. This is where the concept of improvingconversionsrise.

e ISSN: 2395 0056

p ISSN: 2395 0072

Usually, warehouse management staff do not possess360°visibilitytotheirinventorystock.Theyusually facedifficultiessuchasrunningoutoftherequiredstockat the very last minute at the time of an immediate demand fromcustomers.Thisfallacydirectlyimpactsinlettingthe company’scashflowfall.

4. METHODS TO IMPROVE CONVERSIONS AND SALES

Improving conversion rate, a metric generally called “Conversion Rate Optimization (CRO)”. These are data drivenapproachesalongwithaseriesofproceduresthrough whichanywebportal,WebApporanAdvertisementcould beimprovedandoptimizedtoachieveincreasedsalesand maximum revenue with better usability and high on site userengagement.“Sales”metricisoneofthemostimportant metrics highly needed to be improved to achieve higher conversions.

Numerous data science and analytical techniques are implementedone commercewebsitesthatareproducing data,thatlaterproduceconsiderableresultswithoptimized metricsfordecisionmaking.Someofthetechniquesarestill underexperiments.

Chart 2: MethodologyFlowchart

PERFORMANCE (A/B) TESTING:

I .A/Btestingisaneccentricwayforframingthebest online promotional and marketing strategies for any business”.Itisa methodto selectthe design,content,and functionalityofa webpagethatseemsmoresuccessful for conversionfactorbywebsitevisitors.Thistestisalsousedto test website copy, sales emails, search Ads, and so on etc. Thevariationincomponentsorelementsofapagethatmay affecttheactionsorbehaviorofsitevisitorcouldbetested easily.

II. TheiterativeA/Btestingismoreeffectivetofindan improvedversion.Therearetwo typesofA/Btesting: (1) Client side and (2) Server side, with client side being the morecommon.Client sidetestinginvolvesshowingthesame version of a page to every website visitor and then using

© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified

International Research Journal of Engineering and Technology (IRJET)

Volume: 09 Issue: 05 | May 2022 www.irjet.net

JavaScript to make changes within the visitor’s browser beforethevisitorlookstotheresultingpage.Ontheother hand, Server side testing is applied when the webserver showsvisitorswithdifferentpagevariations,alteringthem ontheserverbeforetheyaresenttothevisitor’sbrowser. Nomodificationstothepagearepermissibleatthebrowser. A/B testing can be applied in the cases where an e commerce owner wants to compare: (1) Various page layouts, (2) Navigation organization, (3) Headline effectiveness and content, (4) Website photography and product images, (5) New visual styling for a page, and (6) Newpricingstrategyordifferentpromotionsandoffers.

III. Someoftheresearchtoolsusedareincludedas(1) Analytics,(2)Heatmaps,(3)UserTests,and(4)Surveys.The detaileddescriptionofresearchtoolsiselaboratedinTable below.

Sr No. Tools Description

1. Analytics Identifythemostpopularpages,CTR,traffic sources,pageswithahighbouncerate,and pathsthroughthesite

2. Heatmaps Discover exactly where and how people navigatethewebsite Seewheretheyclick, wheretheypause,wheretheygetconfused, etc

3. UserTests

Personallywatchusersnavigatethewebsite todiscoverwherepointsoffrictionoccur. Proper user testing will also include the testing subjects speaking out loud about what they are experiencing as they completetasksonthewebsite.

4. Surveys

Directly ask the customers about their shopping experiences Did they encounter anyproblems?Whatwouldtheyliketosee improved?

Table 1 : ToolsandDescription

Retaining Buyers using Email Marketing

I. Accordingtothestudyonreturningcustomeraftera successfulsale,acustomerhasa27%chanceofbuyingfrom thesamee commercewebsiteagain.Iftheymakeasecond and third purchase, they have a 54% chance of making another i.e. fourth purchase, hence adding value to the business. Today, to get the most out of “customer lifetime value(CLV)”,ane commercewebsiteneedstoretain,upsell, and cross sell its customers. To do that in a world of increasing population and market competition, they must stay top of mind. Email marketing is still the top way in whichmostpeopleareeasilyreachable.Thesignificanceto retaincustomersinawaythatencouragethemtoreturnand

Impact Factor value: 7.529

e ISSN: 2395 0056

p ISSN: 2395 0072

refertotheproductorshoppingwebagainistomakethem feelliketheyarepartofabrand’sfamily,wherecustomarily andsocially,thefamiliesshouldalwaysstayintouch.

II. Thetimelyemailingtothecustomerwiththerelease ofabrandoranypromotionalsalewithrelevantcontenton regularbasisbuildstrustandimportanceofthatdemandin their running life. Customer will surely recommend that productorbrandtoothermembersoffamilyorfriendfirst psychologically if they get themselves satisfied with such dealingofcare.Itcosts5to10timesmoretoacquireanew customer, than the efforts required to retain an existing customer as a returning customer. Every time a customer returns, it is more likely for him to return and buy again since returning customer build businesses. The retention ratecouldbemaintainedandincreasedbyemailcampaigns, push salesetc.thatincreasetheROI.

OPTIMIZING CHECKOUT PROCESS

I. Optimizing the entire checkout process is a core technical work that requires a lot of testing at the developer’send.Forexample,ane commercewebsitehasa separatepaymentpagewhere84.71%ofthevisitors’traffic landtobuyaproduct.Ifweincreasethispercentageto90%, that will result in generating 461 more orders and an additional USD 87,175 per month. If we take a close look, that’s a surprising rise in revenue of 23.94%. Although it seemslittlecount,consideredagaininrevenue.

II. Similarly, another good technique to apply is to maketheshoppingcartandproductsaddedtoitalongwith the“Checkout”or“OrderNow”buttontoremainvisibleatall thetimeuntiltheuserclickssomewhereelse.Also,changing the“Checkout”buttoncolortosomethingthatstandsout mostlyyelloworgreen isalsoknowntoworkandcouldbe anotheradvantage.

Visibility

Throughoutthebuyer’sjourney,theshoppingcartfeature shoulddisplayalltheproductsthatareaddedtothecartand showthetotalcostofthebasket.Thereasonbehindthisisto avoidcardabandonmentsonthecheckoutpage.

Control

Thecustomermusthavethecontroltoeasilymake anychangestotheproductquantityorevenremovethem.

Here, another amazingly profitable strategy that manye commercebusinessownerstodaystilldon’tknow aboutisCartAbandonment. Itisverycommontoseethat mostvisitorsonawebsitewhoclickonaproducttoadditto the shopping cart, usually don’t complete the checkout processandleavethewebsitewithoutmakinganypurchase. Thisheavilyresultsinagreatlossofsalesandrevenue.Asa solution, if we do not let the cart contents expire and

9001:2008

International Research Journal of Engineering and Technology (IRJET)

Volume: 09 Issue: 05 | May 2022 www.irjet.net

leverage the technology of cart abandonment email reminders and re targeting ads to bring those potential customersbacktothesamecheckoutpagecandramatically increaseconversionsandrecoverlostsales.

Thelastphaseofthecheckoutoptimizationistweakingthe payment fields. If the customer has entered the shipping informationfirst,onlythentheymustbeabletogetintothe billingpart.Thismulti stepformprocessisusedtopushthe customertowardscommitmentandconsistency.Thelogic behindthisisverysimple whenahumanbrainstartsdoing something,itfocusesthepersontoeventuallyfinishit.This littlepsychologicalhackcouldsignificantlyaddupsales.

INCREASING THE USER TRUST

 Security and online privacy are some of the most importantconcernsformanyofuswhoaredoing businessactivitiesorevenbuyingactivitiesonline. Customers,especiallyincountrieslikePakistan,still are afraid to use credit or debit cards online to avoidanymishapsorpotentialtheft.Therefore,an e commerce website of any kind needs to ensure the use of end to end encryption from the origin servertotheclient sideviaSSLandTLScertificate. Ifane commercewebsiteisalreadyusingallthe requiredsecurityprotocols,itisstillimportantto tellvisitorsaboutitbydisplayingsitesecurityseals and PCI compliant labels prominently on the paymentorcheckoutpage.

 A few proven techniques to let customers know thattheirdetailsaresecureare(1)Usingdifferent background color for payment form preferably lightgreencolor,(2)DisplayingSSLcertificatelogo or a green padlock icon, and (3) An additional written statement with the SSL logo such as: “Paymentsecuredby256 bitSSLEncryption”.

 In general, most users are not very technically awareandthereforetheyprobablyarenotawareof terms like SSL or HTTPS secured, so telling them about their data security in plain terms is also a goodidea.

RESULTS & DISCUSSION

Foreverytaskwedoandfordefectwefix,foreveryfeature weaddweaneedtheresult.TheresultcangiveIaoutcome oranideatofix,thisthinghelpedverywellinthisproject andleadtoaperfectoutcomeasexpected.Theresultsplaya maintoboostupyourideasandhelptoreachthegoal.The managedplanhasbeenplayedmainroletoachievethegood results.

Discussionhasbeenmadeweekly,thementorhashelpedus ingivingsuggestionsandapproachtosolvetheissueanduse different techniques to achieve the results. The proper

e ISSN: 2395 0056

p ISSN: 2395 0072

communicationcanhelpthediscussionmoreinterestingin thetechnicalterms,thesyncupsarethemainthingsforthe issue discussion and solving them using better approach finalizedafteradiscussion.

CONCLUSION

Concludingtheresults,weseethatpeoplearenothappyand consideritagoodthingsometimes.Peoplesufferbecauseof itsbadeffectsandareveryhostile.Theychooseaccordingto tradition methods over this new technology. As the populationgrowscompaniesandmarketsitgrowsdayby day.Sopeopleplacethreatsofcheating,fraud,transactions, etc. In fact, people do not really think that they can be trustedpurpose.So,wecanclearlyconcludethattraditional shopping is better than online to buy. This paper helps in expandingbusiness.Savesalotoftimeforclient'scustomers whoevervisitsthestore.Merchantconvenienceandsaving merchant’stimeforothercustomers.Knowledgegainedby thestudentsbybuildingaprojectandinteractingwithteam members and mentor can help in improving the communication skills and completion of project helps in improvingdomainknowledge.

Machinelearningmethodsarehighlyusefulwheregeneral computing methods are less effective. But it is worth mentioning that machine learning can be used as one method for data analysis. This is because, in essence, machine learning can be thought of as a data analysis methodthatbuildstheanalyticalmodelautomatically.Itcan help identify and uncover patterns found within the data. Further,theseintelligentandsmartapproachesstrengthen the enterprise to lead the market more effectively with timely product marketing, timely introducing product promotions and provide a better quality of services to potentialcustomers.

FUTURE WORK

In future we can add more enterprises and classes to increasethesecurityoftheprojecttimeyoucanusesome thirdpartyAPStotriggertheemailsandnotificationswhich canbesenttotheuserandalsoaboutthedeliveringitcan alsointegratedotherpaymentthingswhichareenablingthe net banking internet upi and credit and debit card transactionsalsointerestedinhavingsomespecificwallets wecanalsointroducethewalletsinthefuturetotheproject. Apartfromthepaymentswhichwecanalsohadadelivery trackingwithonlineGPSwhentheproductisoutfordelivery so that customer can easily able to track like some of the deliverysystemszomatoandswiggy.

Also,wecanaddsomeauthenticationthingswhileloginin andandalsointegratedsomeGoogleAPIforauthenticating usingtheGoogleauthenticatorandalsowecanintroduceso manythingstoimprovisethesecuritywhichcanbeupdated in the future in latest. The way US designing can be improvisedmoreinthecomingdaysitrequiresagoodteam

© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008

International Research Journal of Engineering and Technology (IRJET)

Volume: 09 Issue: 05 | May 2022 www.irjet.net

withgoodsetofknowledgeonthedesigningwheretheycan helpinimpressingthedesignsandmakelookandfeelvery well.Theauthorscanacknowledgeanyperson/authorities inthissection.Thisisnotmandatory.

REFERENCES

[1] The Virtual Online Supermarket: An Open SourceResearchPlatform ExperimentalConsumer Research,April21.

[2] H.Saleem,K.B.Muhammad,A.H.Nizamani, S. Saleem and J. Butt, "Data Science and Machine LearningApproachtoImproveE CommerceSales PerformanceonSocialWeb,"InternationalJournal ofAdvancedResearchinEngineering&Technology (IJARET), ISSN: 0976 6499, Scopus Indexed, HEC Recognized"Y"CategoryJournal,vol.12,no.4,pp. 401 424,2021.

[3] MachineLearning:Algorithms,Real World ApplicationsandResearchDirections,[2021],Iqbal H. Sarker ,SN Computer Science volume 2, Article number:160(2021)

[4] 'Data Science and Machine Learning Approach to Improve E Commerce Sales PerformanceonSocialWeb,April2021|12(4):401 424DOI:10.34218/IJARET.12.4.2021.040

[5] Supermarketsofthefuture:deployingAIin yourgrocerystore,ByNadejdaAlkhaldi,Innovation Analyst,PublishedonMay7,2021

[6] DataScienceUseCasesinRetailIndustry| MayurBadole May,2021

[7] 'How Zomato Uses Machine Learning | Zomatosaw 3ximprovementinits'featurestore' servicecomparedtolastyear|ByAmitRajaNaik| PublishedonAugust20,2021inOpinions.

[8] AreAITechnologiesPavingaNewPathfor FoodRetail?|2020|India

[9] GroceryPredictionusingMachineLearning, ISSN 2321 3361 © 2020 IJESC, Neha Kumari, DepartmentofCSE,VITUniversityChennaiCampus, India.2020

[10] How ML/AI is revolutionizing grocery deliveryservice|RaviChandra|Aug2020,India

[11] 'How Data Science and AI are changing Supermarketshopping',RichardFarnworth[2020], Adelaide,Australia.

e ISSN: 2395 0056

p ISSN: 2395 0072

[12] Anatomy of Understanding Consumer BehaviorWhenGroceryShoppingintheNewNext 2020.

[13] J.Leek,"Thekeywordin"DataScience"is notData,itisScience,"SimplyStats,2013.[Online]. Available: https://simplystatistics.org/2013/12/12/the key word in data science is not data it is science/. [Accessed2019].

[14] 'HowGrocersAreReimaginingtheFuture WithAI?'|ByAmandaBaltazaronJan.,2019.

[15] Online Grocery Shopping and Consumer Perception,Feb2018

[16] A Study on Consumers acuity towards onlinegroceryshopping,December2017

[17] Online grocery shopping: the key factors influencing consumers’ purchase intention, May 2017

[18] Web Content Analysis of Online Grocery ShoppingWebSitesinIndia,March2016

[19] Park,A.C.W.,Iyer,E.S.,Smith,D.C.,Iyer,E. S.,& Smith, D. C., 2016. The Effects of Situational FactorsonIn StoreGroceryShoppingBehavior:The RoleofStoreEnvironmentandTimeAvailablefor Shopping. Journal of Consumer Research. 15(4),422 433.

[20] Retail,G.T.,&Survey,C.,2015.TotalRetail 2015 (February). Statistica. (2016). ECommerce. Retrieved June 26, 2016, from https://www.statista.com/markets/413/ecommerc e/U.S.CensusBureauNews.(2016).Retrievedfrom http://www.census.gov/retail/mrts/www/data/pd f/ec_current.pdf

[21] Tribune., 2015. By 2020: Pakistan’s ecommercemarkettosurpass$1billion.Retrieved May 23, 2017, from https://tribune.com.pk/story/1016706/by 2020 pakistans e commerce market to surpass 1 billion/

[22] Wharton.,2015.OnlineGroceriesinIndia: WillConsumersBite?RetrievedJuly9,2016,from http://knowledge.wharton.upenn.edu/article/onlin e groceries in india will consumersbite/

[23] Online Grocery Shopping: An exploratory study of consumer decision making processes, Joana,March2013.

© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page232

International Research Journal of Engineering and Technology (IRJET)

Volume: 09 Issue: 05 | May 2022 www.irjet.net

[24] Pozzi, A. (2013). The effect of Internet distributiononbrick and mortarsales.TheRAND JournalofEconomics,44(3),569 583.

[25] Pozzi,A.(2012).Shoppingcostandbrand explorationinonlinegrocery.AmericanEconomic Journal:Microeconomics,4(3),96 120.

https://www.businesstoday.in/latest/trends/relian ce jiomart opens online grocery service in 200 cities key things to now/story/404916.html

Impact Factor value: 7.529

e ISSN: 2395 0056

p ISSN: 2395 0072