8/2/2019 The New Science of Retail (Financial Express) - Karthik and Ankur (Sep 08)

http://slidepdf.com/reader/full/the-new-science-of-retail-financial-express-karthik-and-ankur-sep-08 1/1

ETAIL in India is und oubted ly a growthstory. In various global rankings, Indiahas emerged as one of the most favoured

retail destinations. The considerable in-vestments made by large-scale domesticand international players have acceler-

ated the development of large, modernstore chains gradually superseding

the 12 million or so tradition al family-runretail outlets. Today, organised retailis estimated to represent only about 5-7%of Indian retail and yet is estimated togrow to about a quarter of sales in two tothree years.

Organised, western-style retailing hasbrought w ith it efforts to effectively add ressthe market and retailers have been tryingout a variety of product mixes and storeformats—super markets, hyper markets,malls etc. The early entrants have tried, oc-casionally erred and then tweaked theirsystems to improve their business models.The time is right for Indian p layers to takeadvantage of this high growth phase by

leapfrogging the more gradual evolutionof retail optimisation seen overseas. Bylearning from the most successful retail-ers in advanced economies, Indian retail-ers can benefit from the latest scientificstudy of consumer behaviour. One of thesesuccessful tools is footfall analytics.

What is footfall?

Footfall (literally a term represent ing thenumbe r of people who visit a store) is oneconventional measure of a store’s oppor-tunities to ma ke sales. Many retailers havetradition ally believed, quite wrongly, thatthe conversion ratio (the proportion of 

each day’s footfall, whichmakes a purchase) is directlyproportional to t he footfall it-self. The gatekeeper at thestores will merrily click on hisor her counter every time an-other customer sets foot intothe store and this goes to thestore manager as a measure of 

potential sales.Simply put, sales= footfall

Xconversion ratio Xaveragetransaction value (ATV).

Retail space is priced onthe basis of the footfall it re-ceives. Malls price the ir shops

on the basis of th e footall they receive dur-ing weekdays and weekends. Even standalone shops are sensitive to accessibilityby customers.

But, could mere footfall count be thetrue me asure of a store’s business? Someretailers today are happy to believe so; it justifies the price the y have paid for the re-tail space. But analysis seems to say thatgetting business is more than just gener-ating footfalls. The basic need of retailmanagement is to convert footfalls intosales. The tenet of ‘Know t hy customer’re-mains inviolable in retail. This is whatfootfall analytics tries to achieve.

Footfall analytics

Footfall analytics is a part of what is knownas ‘Scientific Retail Knowledge’. It studiesthe demographic and psychographic be-haviour of the retail customer and corre-

lates it with th e sales of the store. It begins,of course, with accurately measuring thenumber of customers that enter the store.Unobtrusive equipment placed at the storeentran ces achieves this end. To accuratelymeasure the traffic and automa tically logit on to th e data servers, 3D imaging tech-niques (infra red cameras) are used .

Here, in one example, the emphasis isnot on the number of distinct customers,but on the shopper groups. A family of four walking into a superma rket or a chil-dren’s clothing store, wou ld not be four cus-tomers, but one shopper group making a

family purchase.The data that the footfall equipment col-

lects is time-stamped be fore being sent tothe dat a servers. A computer logs the dat aand periodically transmits it to the centralhub. The communication channel couldbe anythin g from a modem to a GSM net-work. The central processing hub sits at th eservice provider’s office. This hub (themen and the machines) aggregates thedata, analyses it using sophisticated pro-prietary algorithms, and prepares the re-ports. Store specific suggestions are madeon the basis of these reports and sent b ack to the clients, according to the informationthey are looking to harvest; ATVs, staff-stretch, effectiveness of marketing cam-paigns, conversion rates etc.

There is an accurate profiling (a timevs traffic graph) of the st ore traffic. A pas-sive count of footfall at the end of the da ygives mere seasonality. Hourly profilinghelps know the periods when the store re-ceives the bulk of its customers. This canhelp schedule the staff and can help man-agers determine whether it is better topay for more staff to cover busy periods orbe prepared to save money on additional

wages but sacrifice some sales. With accu-rate data, managers can see how staff training, for instance, can improve conver-sions. A ‘Happy hour ’ or other time-de-pendent policy can help improve trafficduring the dull hours or even wean awaya portion from the over-hectic hours.

Taking the analytics technology to thenext stage, int roduces the ‘Know th y con-sumer’part. Using the same equipmentthat you had so far been using for secu-rity—the closed-circuit te levisions, bu t in-stead of pa ssive security monitoring, ateam of experts does an active study. A ran-

dom sampling of the store traffic showssome distinctive consumer behaviour andhow th ey ‘shop’your store. From the videoclippings analysts can deduce the hot spotsof the store—the areas that receive mostattention and visibility.

And also what exactly makes thehotspot, a hotspot. Consumers react tostore layouts, ambience, lighting, prod-uct display—a host of controllable vari-ables. Once the peculiarities in consumerbehaviours of a population are mapped tothese controllable variables, the optimumlayout can be achieved. Though a lot of stores do eventually achieve it, the consid-eration with analytics is that it’s muchfaster and accurate—because it’s backedby hard, repeatable, accurate researchand analysis.

The results of analytic studies of eachstore can then be compared to others in thechain. The data generated by an electron-ics store in Mumbai, for instance, can becompared with the data from stores inPune, Chennai, Hyderabad and Delhi.This, in turn can be compared with the bas-ket of consumer electronics retailers. Thisbrings us to the concept of a retail index.

A retail index gives the relative traffic dur-ing a period across a group o f peer stores(stores with same or similarmerchandise/format/market sector).There would be different indices based onregion, merchandise, distance from citycentre (high street and suburban).

One such index is the SPSL (now partof Synovate) retail traffic index (RTI),which has been tracking traffic levels in theUK since 1998 and is the longest runn ingand most respected index of its kind. Itsprincipal role is to keep tra ck of clients’traf-fic levels relative to a comparative set.

They provide a means to gauge how busythe client’s stores are (in tra ffic terms) rel-ative to the rest of the UK market overtime. The reports show ind exed trafficrather than aggregate total traffic flowsso that each store contributes equally tothe overall client index value that isreported weekly. RTI tracks ‘pureretail’—only the customers entering thestores. This is because an a ggregate foot-fall in a mall or galleria, as counted by somerival companies, has less relevance to ex-act store figures.

Does footfall analytics help

sales?

Bespoke end-to-end systems helpclients achieve business growth in a vari-ety of ways—right from staff schedulingto optimising store layout and always witha measurable return on investment, some-times within mere weeks. For instance,for any retail store, assuming idea l trafficconditions and population d istribution,the catchment area (or area that generatescustomers) is a circle centred on the store.But this never hap pens—accessibility andconsumer psychology is what det ermines

the store’s footfall. For example, there wa sa mid-sized store in Bangalore, situatedon a busy arterial street. Around eveningtime, the heavy traffic on the road nearlycut off the customer tra ffic from the otherside of the road. The customer traffic,therefore, did not increase during the peak hours. Merely counting the aggregatetraffic at the store during th e day wouldnot bring forth this anomaly. Only a pro-filing of the hourly traffic can help diag-nose this problem.

Another example comes from a leadingfront runne r in organised retail. The com-pany started out with huge sized stores, butsoon realised that most of the a rea remainsunaccessed by the bulk of the customers.It then spent time working out the optimalsize of the stores and down sised many of them. Fast food chains as well have noticeddull business at some of t heir stores locatedin malls and then moved out in favour of stand alone shops.

After p rice and accessibility, variety is

the most important consideration in aretailer’s list. Do products get stocked ou ton peak occasions, are too many varietiesconfusing the customer, what are theproducts on wh ich the customers expecta lot of variety? Analytics maps the salesdata from the electronic point of sales(EPOS) counter and maps it to the con-sumer traffic detected by the footfallequipment. Data might, for example,show that a certa in brand of shaving foamis being favoured at a men’s store when t hecustomer enters as a couple. Algorithmwould be able to suggest whether thebrand would generate good sales duringgift occasions. This would lead t he retailerto over stock this brand during occasionslike Valentine ’s day or Diwali.

Customer motion studies can help storelayouts an d store sizes. By visually study-ing the cu stomer throu gh security CCTVs,analysts can determine what layout wouldbest attract th e customers. Which productsshould be shelved at the ent rance for im-

pact buying and which ones should in theaisles so that th e customer can study andfeel it at leisure? If too many areas in thestore remain unvisited, does that call fordownsising the store or sub-letting conces-sions, improving the lay-outs and ambi-ence or expanding the ranges? Onlyproper analysis can tell.

Where does this lead to?

The competition can on ly get fiercer as or-ganised retail grows. Retailers need pre-cise tools in hand th at will allow them totarget individual consumer bases and in-vest with greater confidence. As this indus-try grows, it will become increasingly im-perative for retailers to adopt the‘analytics’approach. Radical innovation inIT processes helped India leapfrog to be-come one of the biggest IT destination s of the world. Given the current scenario,adaptat ion of a similar attitude towards re-tail science could catapult India to theglobal leaders of retail too.

Karthik Ramamurth y is associatedirector & head, Syno vate Business

Consulting, Mumbai, and Ankur 

The real challenge for retail management is to convert footfalls into sales

The new science of retail

F I R ST S T O P F O R M A R K E T I N G , A D V E R T I SI N G

4 THE F INANCIAL E XPRESS

TUESDAY,SEPTEMBER

BW

◗ BYINVITATION |KARTHIK RAMAMURTHY& ANKUR HAZARIKA

Malls pricetheir shops onthe basis of the

footfall they

receive duringweekdays and

weekends

The pharmaceutical industry in Indi ais poised for double-digit grow th.Signal ing this is the increasedoutsourcing of clinical research toIndia and contract manufacturingoffering cost-effective sourcing forglobal companies. Marketingconditions tailored to localconditions will propel higher growthin revenue of small and medium sizecompanies, according to aCII-Interlink white paper on “GrowthAgenda of pharmaceutical industryin India” , tabled recently.In the current scenario, Genericscontribute 7% of the brandedformulations market and 3% of thetotal pharma market at Rs 2,173crore. Branded formul ationsconstitute the bulk of the market atRs 31,073 crore.

Take a

chill pill

      S      T      A      T      I      S      T      I      C      S

      S      P      E      A      K

Growth rate of acute vs chronic therapies in Indian pharmaceutical market

Therapy 2007 2006 2005 CAGRAcute 11% 18% 8% 12%Chronic 21% 17% 11% 16%

Indian pharmaceutical marketShare of acute vs chronic therapies in Indian pharmaceutical marketTherapy 2005 2004 2003Acute 76.6% 77.1% 77.9%Chronic 23.4% 22.5% 22.1%

Value of brands in theIndian pharma market

       A     c     u      t     e

       C       h     r     o     n       i     c

       T     o      t     a       l

        2        3  ,

        3        4        7

        7  ,

        6        9        2

        3        1  ,

        0        3        8

Composition of the pharma industry

2007 2015*

Health Ins 0.14

Exports 44%

Branded

formulations 48%

Generics 3%

Hospitals 5%

Brands .5Mktgefficiencies 1Pricing 1Rural Market 2Middle class 2

Current growt h 13

Formulations

branded g enerics 31,038

Exports 28,703

Hospital market 3,073

Generics (7% of market) 2,173

Source: ORG IMSMarket I ntell igence Report 2007 and Int erlink Know ledge CellGraphic by

Year: 2007,value in Rs cr