How do you decide how many cash registers you need for a new Walmart store?

clarifying question

• Cashregiter - is a system which allows the users to make entry about the sale of product by scnanning it through a scanner and collecting the payment. Post payment collection either through, Card or UPI or any other method of payment, the software will make neccessay entry in the system , in current sceniors these are pos connected to a software  correct ? - Answer yes 
• Peak location what is foot fall daily ? - answer - not sure 
• the esitmation needs to be done for daily basis - yes
• Geo - india
• What is Operation hours - 
  • answer- 12 hours 

For the estimation, I am taking the assumption that an average Walmart neighbourhood store has a walkable area of 10000 square feet (Average neighbourhood store size: 40000 square feet)

My need of counters will be based on the bottleneck which would be the rush hours. I am assuming during these times, we will have one shopper every 20 square feet, making a total of 500 people. Another assumption is that average shopping time at these stores is an hour.

So, 500 people will be checking out every hour.

On an average a checkout (from scanning to payment and packing) takes 3 mins.

Every 3 mins, the number of people checking out = 500/20 =25

An average shopper is okay to wait for another 3 minutes in line before checking out.

So, the number of people checking out every 3 mins would be halved

≈ 12

A total of 12 counters would suffice for such a store

If you are opening a new Walmart branch, how would you decide how many cash registers are needed for the store?

I would collect the below data about other Walmart locations:

* Average number of customers per hour

* Number of SKUs for sale in the store

* Number of parking spots

* Number of additional locations in a 50 mile radius

 

Those metrics would form a profile. I would then attempt to match the projected metrics of the new location with the closest profile and determine the number of cash registers needed is the same as the closest match.

My formula would be Number of shoppers per hour/Avg. duration spent at the register

So everyone would leave in that hour.

First I want to estimate the number of visitors who do the shopping. per hour on a busy day.

Are there any other Walmart close by? No.

What is the demographics of the neighborhood cities in terms of Income Level? Let’s say mix.

I would estimate Walmart opens a Walmart in a peripheral of 150K- 200K people. there is a Walmart at every other three small cities.

Let’s say 200K. Average household number is 3. So 70K household. Since this is a mixed income area, Let’s assume 40% of the household would go to Walmart. 30K households.

Average number if visits per month would be 3 lets say. 90K visits per month.

3000 visits per day. Let’s say between 9AM and 9PM are the busiest times.

250 visitors per hour. Let’s say average waiting time in the queue is 15 mins + 5 mins at the register. So a total of 20 mins to check out.

250/20 = 13 cash registers.

 

Sanity check: As far as I can remember I see around 15 registers at a proper size Walmart or similar store.

You are new grocery store owner. How do you decide number of checkouts?

 

Note: Instead of estimating number of checkouts, this question would have been great if it focussed on designing a next generation checkout experience. 

Anyways …

 

Clarifications

1. Physical store?
2. Is it part of the brand? No. Just one (stand alone and not associated with brands)
3. Physical location? Boston
4. Any location specifics - just flat land
5. Competitors? 3 (already existing) +1 (new) ; makes total 4 stores to serve the community.
6. Store has online presence apart from physical? Yes
7. Sells typical groceries (veg, meat, diary, etc) 

 

Assumptions

1. Local population 650k
2. 4 stores serving the entire local population
3. Physical dimensions of the store 1000 x 100
4. Number of aisles 10+
5. Sells typical groceries (veg, meat, diary, etc). Not a specialty store

 

Visiting customers to the store

1. Under 18 (They are not yet a paying customer in this group considering CC). 
   1. But if we accept cash - 12 to 16 might be considered as customers as well
2. 18 - 65 (wide range but that’s OK for estimations). For purchase scenarios, we could decompose the segment into various groups. Further segmentation doesn’t help to decide number of checkouts.
3. 65+ (special consideration for checkout might be needed for this segmentation).

 

Calculation

To decide number of checkouts, young kids are excluded.

Total population to serve 640k

Excluding kids, etc (30%) = ~440k. To make the calc simpler, let’s say 400k customers.

Split evenly between 4 stores, the new store would have to serve ~100k (reducing this further based on number of people per house hold would have been better).

Assuming peak traffic is on Friday and Saturday ~ 20k on each day, 12k on rest of the 5 days.

20k to be served between 7am to 9pm = 15 hours implies ~1300 customers/hour.

So let’s calculate number of checkouts needed for 1500 customers showing up at the same time.

Let’s say on an average 20 items are purchased. Each item need to be scanned and packaged.

Let’s say 1sec to scan and 4sec pack at checkout = 5sec for each item => 100 sec for 20 item

1500 customers with one checkout = 150,000 sec = 2500 mins = 41 busy hours

1500 customers with two checkout = ~20 busy hours

….

….

As a store owner if I commit myself to 4 busy hours for peak traffic, the store owner probably would need about 10 checkouts.

 

On the physical dimensions of the store (L = 1000ft; B = 100ft)

For aisles and shelves; 1000 x 75

For checkout stations; 1000 x 15 (assuming 1000 x 10 for clearance - walkway, etc)

Checkout lane dimensions 10ft x 15ft => 100 checkout lanes (max)

with 2ft distance, each checkout lane can hold 7 customers

At peak hour, only 700 customers could be served (even with 100 checkout lanes).

 

So distributing the customers to online check out is the best way (due to physical constraints)

Should I assume that its a normal size walmart store or a walmart super center? 

Do we know the kind of city that we are talking about?  Is it a highly populated city, is it located at a busy route, or is it outside of the city? - all this matters in terms of understanding the kind of traffic that you can come expect at the store

Do we know anything about the population of the city that we are opening the store in?

Should I do this for both manned registers and self-checkout?

 

You may choose whichever city you want and yes please do it for both manned and self-checkout

 

Assumption: 

Opened in a city like San Jose ~ roughly a 1M residents and I'll assume its a normal walmart store

I would look at the following to get started

 

1. Operational hours of the store

2. Frequency with which people come into the store

3. How quickly they shop and queue up for purchase

4. Processing time we are targeting with the customers

For a question like this, I think the best approach is to take the worst case scenario because that’ll be our bottleneck and thus dictate the number of registers we need.Example of a busy saturday:

 

Operating hours: 8 AM - 11 PM

 

8-11 AM: customer walks in every 3 mins = 20 customers an hour

11-5 PM: customer walks in every 30 secs = 2 customers every min = 120 customers/ hour

 

So lets take the 11 AM - 5 PM as our busiest time of the day for the calculation

 

Typically you'll have

Category A: Customers who pick up 5 items or less => take 10 mins to shop then line up

Category B: Customers who pick up 15 items and more => take about 20 min to shop and then line up

Category C:Customers who pick up 25-30 items => 30 mins to shop then line up

 

Lets assume:

10% of the customers who come in during 11 AM - 5 PM are Category A

20% of the customers who come in during 11 AM - 5 PM are Category B

70% of the customers who come in during 11 AM - 5 PM are category C

So that is:

12 customers / hour in Category A -> self checkout 24 customers/ hour in Cateoryg B  -> self checkout or manned registers

84 customers/ hours  -> queue up on manned registers

To account for some customers in Category B opting for manned registers, Im gonna split them and say half of them decide to do self-checkout and other half would opt for registers

So that's: 24 customers/hour -> self checkout 96 customers /hour -> queue up on manned registers

96 customers/hr x 6 hours = 576, so say 600 customers that would lineup for purchase over the course of the 6 hours from 11 AM - 5 PM.

24 customers /hr x 6 hours  = 144 ~ 150 customers that would lineup for self check out

For the customers who line up at the manned registers:

A Customer queues up every up every (20+30)/2 = 25 mins , but we pick the worst case, so say every 20 mins

600/20 = 30 customer will lineup per minute

 

Checkout takes on avg = 2 mins / customer / register Wait time = 29 x 2 = 58 mins (unacceptable, but this is assuming that we only have 1 register)

We want to reduce the queue length to reduce wait times, which means we want more registers 

Assume we want wait time to be < 3 mins

 

29 customers  x 2 min/customer x 1/register = 3

1/R = 3/(58) = > R = 58/3 = 19 registers

 

For self checkout:

150 customers and lineup every 10 mins = 150/10 = 15 customers/min

Avg checkout time on self check out  = 2 mins 

14 customers x 2 mins/customers x 1 = 28 mins

14 customers x 2 mins/customer x 1/R = 3 mins

 

R = 28/3  ~ 10 self checkout registers

Note : Wrote this before reading others. Took ~30 minutes to talk, record myself and write this. Clarifications :

A. Can we make assumptions on the number of stores around? (Other walmarts, target, costco, supermarkets?) → Assume typical case where they may be a target a few miles away but no direct competitors in the same block for example
 

B. Is this a walmart super center or a regular walmart fresh store?
→ Assume this is a supercenter
 

C. Location : Is this in a busy urban environment?

→ Assume urban population 

 

D. We need to understand the income ranges and the type of neighborhood. Do we have more information on this or the number of public/pvt schools around or area itself?

→ Assume middle class neighborhood with decent mix of public and private schools

 

E. Does this walmart have self checkout? 

→ You decide → Assuming yes, will calculate.

 

High level Calculations :
 

Find the average number of people who shop at this store over the weekend (assuming weekends are busier and we want to ensure customer satisfaction even during busy hours)

 

Formula :

Number of people who shop on a given day = population of the area - number that shops at local grocery store - number of people that shop at whole foods/TJs - number of people who shop at target

 

Personally, I’ve seen myself go to target if I want to buy some prescription items or personal care products (CVS, Ulta Beauty and so on) 

 

People who shop at Amazon whole foods or Trader joes (Which are highly specialized stores) usually tend to be in the higher income groups

 

Most people would prefer to go to a walmart over a smaller grocery store since walmart has more selection and better deals/prices. And supercenter is almost like a wholesale location with furniture, gardening, groceries, electronics, games/toys and more. (one stop shop)

 

Assume that this walmart serves 10 square miles in a city that is not too dense but still urban (100 sq miles)

 

Population of said city = 2 M

Population per 10 sq miles = 2M x 10/100 = 200000 

 

Others = grocery delivery + online methods 

 

Pw = Total Population - Ptgt - Ptjs/whofd - Pgroc - others 

= 200000 - 200000x{ 15% + 5% + 8% + 20%}

= 0.6 x 200000 = 120000 

 

This would be the total number of people who would go to a walmart in general 

 

I would classify these as 3 categories of people :

People who go once in a couple of months → ½ per month = 40%

People who go regularly = 1-2/month = 30%

Powerusers once per week = 4-6/month = 30%

 

120000x0.4 /2 = 24000
120000x0.3 x 1 = 36000
120000x0.3 x 4 = 144000

Total per month = 204000

 

Total hours in a month = 30days x 10 hrs/day = 300

 

Total number of folks per hr = 2040/3 = ~700

 

This seems a bit high
I think I missed on thing here.
Number of people who would order through walmart online and pick up. 

10-20% → 700 - 140 = ~560 

This still seems high.
I think we need to consider folks who buy essentials and use the fast self check out. I would assume 30-40% of this would be those folks

Number of people to be checked out = 60% of 560 = 380 

 

1 counter takes approx 5 minutes per person

In 1 hr, one counter can serve 60/5 = 20

To serve 380 people in an hr, we would need 380/20 = 19 cash counters

Clarifying Questions:-

How many employees would we like to hire? if there is a cap on the hiring then that many cash registers only would be needed. For example, 5 employees means only 5 registers

What is the standard wait time needed for each customer? 5 mins?

Walmart would be launched in the Urban areas? Assuming it to be an urban area

 

Different Timelines for visiting Walmart by customers:

Weekdays

Weekends

National/Festive holidays

Scenario when we have one cash register:- let us assume 2 people come to the register at the same time, then one customer will wait for 5 mins while the other person is being checked-out. This clearly says that the No of cash registers is directly proportional to half the no of People visiting that particular cash register. Suppose, There are 30 people willing to checkout, then considering 5 mins as the maximum wait time, we should deploy 15 cash registers.

Now, We need to look at the store capacity & demographics of the city. A Walmart store can serve, 1lakh population/store. However, only 50% visits Walmart due to the availability of competitors. But not all these 50k would be coming on the same day/time, so equally distributing per week ~7k people would be visiting the store. Weekends or festive this number would spike up a lot around ~12k.

Now let us design our case for the maximum i.e peak days.

Now 12k people visiting Walmart, where per hour 1000 people would be in the store. Not all the people would do shopping & I am assuming every group/people would be completing their shopping within 60min(1hour).

People ready to checkout per hour= 1000/60= ~16 people

Now the number of cash registers required considering 5mins waiting per customer would be = 16/2 which is 8 registers.

Now sometimes there would be peak hours where the no of people willing to check out would nearly double i.e.= 16*2=32 people

In such cases, we would need 16 cash registers.

Based on the budget constraints, Manpower & their rotational shift policies anywhere between 8-16 registers should be deployed for a Walmart store.

Clarifying questions and assumptions

I’m going to assume this Walmart branch is opening in the US. Will it be opening in a rural, suburban or urban location?

 

I’ll do a top down estimation approach starting with some assumptions about population size:

 

1. Let’s assume that people won’t drive more than 10 miles to get to Walmart in a suburban setting

2. If there are ~2,000 people per square suburban mile, that would put our total population of potential walmart shoppers at  20K

3. Of those 20K shoppers, let’s assume 50% can actually get themselves to Walmart. The other 50% are too old or too young to drive, or don’t have a car or access to a bus to get them to the store. 

4. This puts us at 10K people who can get themselves to walmart to shop

5. Not all those 10K would be in the store at once, let’s assume walmart is open for 10 hours from 8am to 8pm a day and has the following hourly visit distribution:

   1. AM: 30%

   2.  Noon: 20%

   3. Afternoon: 20%

   4. PM: 30%

6. We should plan for enough registers to handle peak visit times so assuming 3K people in the store at once (10K * 30%)

7. Let’s also assume only ⅓ of those people are ready to check-out at the same time. So 1K people need to check out at once during peak hours.

8. Let’s assume that people should wait no more than 10 minutes in line and each check-out takes 2 minutes. This means that each line should have a max of 5 people in it at any given time

9. If there can only be 5 people in line at once, we would need 200 registers to serve 1K customers.

10. Sanity check: The last time I visited a suburban mega-store (Walmart Costco, etc) there were likely no more than 50 registers. My estimation seems far too high. Potential ways I went wrong in my assumptions are not considering frequency of visit for the 10K potential Walmart shoppers, not including a cut on people who don’t shop at Walmart, and low-balling the amount of time someone would stand to wait in line.

Clarification question:

1) is it Walmart supercenter or walmart neighborhood? - Assume is supercenter

2) Does cash registrar includes of people handling returns or customer service? - No, lets keep them separate

3) Do we need to estimate at particuar hour of day or weekday/Weekend? - Assume the avg. cash registrar needed.

4) Will Self-Checkout part of cash registrar ? No

I am going to estimate avg cash registrar needed in walmart supercenter which doesnt include returns/customer service counters and self-checkout

In order for me to estimate it right, i will estimate how many customers does an cash registrar handle on a typical day and then i will assume avg. total customers entering in walmart supercenter per day.

In order to provide estimation on avg customers handled by a walmart cash registrar, i will look at operating hours of walmart and segment it by traffic.

Typically walmart are open from 8Am to 11Pm = 15 hours of operation. 

I am assuming avg walmart cash registrar spends 3 mins per customers, so total 20 customers avg per hour.

I will assume at what capacity they are operating at to identify total customers handled by 1 walmart cash registrar.

Hour	Capacity	avg customer/hr	total Hours	total customer
8AM-11AM	50%	20	3	30
11AM-2PM	75%	20	3	45
2PM-5PM	50%	20	3	30
5PM-9PM	100%	20	4	80
9PM-11PM	65%	20	2	26
Entire day	70%	20	15	211

Lets assume on an average 1 paying customer (solo or group) enter every 15 sec, Hence total avg 3600 customers enter walmart supercenter for shopping.The above table estimate 211 customers handled by 1 cash registrar in a day.

out of 3600 paying customers, lets assume 60% of customers are willing to pay via self-checkout, that means 1440 customers used cash registrars.

Hence i get an avg of ~7 avg cash registrar needed to serve 1440 customers.

 

 

Here is my approach to this problem.

I assume the # of cash registers need to be enough to keep the average customer wait time to a reasonable timeframe, so as to not impact customer experience

so, let me just start off with a basic equation, and then start getting into my clarification questions and assumptions

1) basic equation: # cash registers = function [minimize (average customer wait time)]. 

I will flush out this equation more as I build out my assumptions and ask my clarification questions.

2) avg customer weight time is a function of 'avg checkout time' and 'queue length'. 

          - i.e. avg customer wait time = F(avg checkout time, queue length)

3) Assumption: I am assuming we will design for peak volumes (typically on weekends) when a customer is coming to Walmart to do bulk shopping for the rest of the week. so, let's assume that the avg person doing bulk shopping will spend around 5 min at the cash register

i.e. avg checkout time = 5 min

4) Clarification question: Is there an SLA on what Walmart considers a reasonable wait time for customers, without impacting experience. 

          - If not, lets assume that an average wait time of 10 min is considered reasonable, beyond which the customer experience is negatively impacted

    - i.e. desired average customer wait time = 10 min

we can generalize this, so desired average customer wait time =. y

5) Per point 4, we would like the desired average wait time for each customer to be y ( our assumption is  ~10 min). as we think about this, what this tells us is that for the people standing in queue, 1st person already being serviced, 2nd person waiting 5 min (avg checkout time), and next person is waiting y min, which is our desired wait time.

this means, that Walmart should aim to have a queue length based on the following equation

i.e desired queue length = [(y/ avg checkout time) + 1] 

     = (10/5) + 1 = 3 (for our example)

6) let's assume that the incoming and outgoing volume of customers during peak hours keeps the peak volume within Walmart constant. so peak volume = X 

7) of volume X, lets say 80% are actively shopping, and 20% are standing in line for checkout. so 

 - checkout volume during peak hours = 0.2*X

8) so to maintain this, our equation for # cash registers 

# cash registers = checkout volume during peak hours / desired queue length

So based on this, assuming volume at any given time during peak hours = 100 (i.e. X = 100), then

# cash registers = (0.2 * 100) / 3 = 6.66 ~ 7 cash registers

Sanity check:  any local grocery store has between 6 - 10 registers, depending on volume projections, so this passes the sniff test IMO.

This equation can flex depending on peak volume and desired wait time assumptions

So this is the first time I'm actually working through an analytical problem, so any constructive feedback is welcomed!

 

Step 1: Clarify scope

I would ask for location of branch (rural/urban/suburbs), does it include self check out? Acceptable wait time?

Assuming that we are looking at a Walmart in the suburbs, includes self check out.

 

Step 2: Create main equation

# of cash register = (Avg # of customers checking out at a given time during rush hour X # mins needed to checkout per person) / acceptable wait time

 

 

Step 3: Breakdown equation

Avg # of checking out = population X market share X % of CM shopping at rush hour

 

Key assumptions

1) We are looking at population within a 30 mile radius assuming that most people will drive 15-25 minutes for a supermarket

2) Self checkout will serve 30% of customers

3) Walmarts market share is around 30% among other supermarkets in the area

4) At the peak "grocery shopping" rush hour, 50% of households are doing grocery

5) Only one person from each household is grocery shopping at a given time. 3 people per household to account for singles/married/married with more than 1 kid

6) Acceptable wait time is 10-15 minutes at rush hour. We are calculating using rush hour standards to ensure that there is consistent customer experience even at the busiest time.

 

Step 4: Calculate the #s

(9,000 people within 30 miles radius /3 people per household) ( 30% walmart share) (rush hour shopper 50%) =450 shopping at walmart at a given time

Given that 30% of Customers will self checkout, cash registers only need to account for 315 customers

# of cash register = 315 customers/15 min wait time = 105 cash registers

 

Step 5: Sanity Check

I think 105 cash registers is way over, so I adjusted population to 6,000,walmart market share to 20%, rushhour shopper to 20%, and 40% to use self checkout. 

(6000/3)(20% walmart shares)(20% rushhour shopper)= 80 customers at walmart at a given time

# of cash register = (5min per customer X 80 customres)/15 max wait time = 16 cash registers

 

Assuming that we dont know Walmart store is located in which city , I am attempting using a Bottom-up approach rather than Top-down which is another approach by calculating total number of people that can visit a Walmart store per hour. For this we will have to calculate probable number of Walmart stores in that city . Instead we can use Bottom-up approach based on Peak -capacity

 

Assuming average size of a Walmart branch to be 10000 sq ft
We can assume 60% of branch will be occupied by goods, staff while 40% of branch can be occupied by shoppers = 4000 sq ft

Assuming 200 sq ft is occupied by Cash Counters which is 50 ft long and 4 ft wide and there is additional space of 35 feet for queue for each cash counter

Each Cash counter width can be assumed as approx 5 feet so we have space for 10 cash counters with 1 staff

Assuminng 100 ft length and breadth of the branch , for 1 person average space required is 4 ft so 25 people in a row can occupy in peak , assuming 25 such rows we can assume 2500 people can occupy in peak capacity

In USA there are multiple retail outlets like Walmart, Tesco, 7-Eleven etc. so assuming market share of Walmart to be 40% so 40% of 2500 = 1000 people will occupy at a given point of time in peak capacity. Remaining shoppers can be in waiting queues outside Walmart

An optimal wait time at a Cash Register as per Industry SLA should be < 5 minutes per person

In one cash counter which is 35 ft long, 10 people max can occupy at a given point of time. Assuming 10 cash counters, 100 people across all counters.

Taking 5 minutes as wait time for 1 cash counter per shopper, so in 5 minutes we can process bill for minimum 10 shoppers. Assuming an average of 2.5 minutes we can assume 20 people in 5 minutes

So to handle peak capacity 10 cash registers are needed to be active

Peak Occupancy = 10 cash registers

Normal Occupancy = 60% therefore 6 cash registers

Low Occupancy = 20-40% so 3-4 cash registers can be active

# of registers = # of people go to Walmart in rush hour / # of people a register can handle in rush hour

# of people a register can handle in rush hour = suppose a person checkout takes 3 mins, a 3-hour rush hour would handle 60 customers/register

 # of people go to Walmart in rush hour:

Suppose we are in a city with 1M population and the walmart has a market occupation of 20%, and there are 20 walmart stores in the city,  then 10K will be the customer of this walmart.

Then we want to calculate # people come in rush hour:

Among these people, I assume 40% are active customers who come to Walmart once a day, 30% are the regular customers come once every 3 days, and 30% come once every 7 days. Suppose 60% people go to supermarket during rush hours:

4 K * 1 *60% = 2.4K

3K * 3/7 * 60% = 0.7K

3K * 1/7* 60% = 0.25K

Total in rush hours: 3.3K

3.3K/60 = 55 Registers

If we allow 2 people waiting/counter  - 55/3 = 18 Registers

 

Possible Improvement: 

1. Market share of Walmart

2. % of people going shopping on rush hour 

 

 

 

 

Assumption it’s a typical size walmart.

 The Footfall of the store will be more on weekends and less on weekdays.

Even on weekdays will be more in the evening than in on weekdays

Lets design for weekends and evenings when there is more footfall rather than for weekdays

 

Assumption

1. Lets assume in an hour a local walmart sees a footfall of 1000 people
2. Out of1000 people ; there is a family , singles , couples - assume an average size of 2 and they will be shopping together hence 500 unique shoppers
3. Lets assume a conversion rate of 80 % , hence a conversion 400 Unique shoppers [ 20 % will go without purchasing anything]
4. Since Walmart is a Superstore the average basket size > 1 for this ; for the purpose of this exercise lets assume a basket size of 20
5. Lets assume they are equally distributes ie means 400/60 ~ 7 shoppers in a minute walk to the counter everyminute

 

 

Cash Register.

 

1. One Scan takes 6 seconds in an average[assumption] , hence one cash register can scan 10 items in a minute or he can take care of a customer in every 2 minute
2. If input == throughput hence the customer walking in should be equal to customer out hence it should be equal to 7*2 ==14  registers [assuming no wait time]

 

 

 

Walmart is a big box retail store with huge number of footfalls during the peak times.

Cash registers could be of 3 types:

1. Manned cash registers
2. Auto checkout registers
3. Express checkout Cash registers for people with less than 5 items

Lets focus on calculating the number of manned cash registers using the below formula. This formula can be used to calculate the number for all three type of cash registers by dividing the customers who take one of the above three options in to three % groups.

Since the number of cash registers should result in a stressfree experience for the customers during the peak times, the number should be determined to sustain the peak time threshold number of customers.

The core values to consider to determine the number of cash registers for the store would be the below points:

1. Average checkout time taken to serve per customer - Considering the time taken for billing and payment, it takes about 5 mins per customer

2. Number of customers looking for checkout per the average checkout time - 

Lets assume 50 people walk-in for every 5 mins to wallmart and out of them 20 people join the line for manal registersper every 5 mins, we could have about 20 people joining the checkout lines

3. Permissable wait time per customer - Lets assume average permissable wait time is 5 mins.

# of customers that can be served per cash register = (Time per checkout + permissable waiting time)/ time per checkout = (5+5)/5 = 2

# of manual cash registers required = number of people joining the line/# of customers that can be served per cash register= 20/2 = 10

In a similar way, number of express cash registers and automatic cash registers can be calculated using the average time required for checkout, % of users choose these options and permissable wait time.

 

Assumption 1: It is a Supercenter which on average sees 5000 people at it's peak time. 

Hence 5000/12 hrs = 416 people per hour

Assumption 2: There would be a 50:50 division between manual registers and automated registers  

Assumption 3: Every hour the target for the manual register is 15 customers and automated register is 20 customers 

Assumption 4: 416 people are equally divided between the manual and automated registers 

Manual register: 213/15=14

Automated register: 213/20=10

Total registers required = 24

What are the assumptions regarding the size of the store were made? --> this will give idea about max number of shoppers expected, kind of SKUs they would be buying which would give rough idea of billing time, payment mode, offers they can avail during billing etc.

 

Assume max shoppers for billing can be 100. So counters required would be 

num_counters = (max_shoppers_in_store/max_wait_time) * avg_billing_time

max_shoppers_in_store = 100

max_wait_time = 30 mins (assumption)

avg_billing_time = 5 mins (assumption)

 

num_counters = 100/30*5 = 17 roughly or may be 18 for sanity check.

 

Now variations can exist like how many shoppers would be

 < 5 items so express counters are required.

regular customers so regular counters required

self check out shoppers.

counters above would have different billing times and depending on shoppers' distribution among the above categories will impact number of counters required.

 

Clarifying questions:

What are the assumptions regarding the size of the store were made? --> this will give idea about max number of shoppers expected, kind of SKUs they would be buying which would give rough idea of billing time, payment mode, offers they can avail during billing etc. Assume every shopper would like to wait 15-20 minutes in billing queue so max number of shoppers expected at a time, total billing time at one counter, that divided by 15-20 minutes will give counter numbers needed. 5-10% adjustment for extraordinary delays etc.

Clarifiying Questions:

Does Walmart have any maximum registers allowed per store

Is this Walmart a supercenter or a regular store?

Assumptions:

Average time spent per person at the cash register is 10mins

The surrounding area of the store has a population of 300,000

Working hours of Walmart 8am - 8pm

Number of cash registers required = # customers/day * time spend at the register / Total working hours

 

Calculation

Let's say 10% of the population come to Walmart  = 30000

In a day number of shoppers would average to 30000/30 = 1000/day (This might be high on the weekends and low on weekdays hence average it)

1 person spends 10mins at the register

1000 spends total 10,000mins

The working hours 720mins

so the number of registers required would be 10000/720 = 14

 

Few Questions and Assumptions: 

Question: 1) Any specific category of neighbourhood to reference to? Ans. No

2) Do we have access to Walmart's existing store data for assumption? For instance: Average No. of Shoppers visiting over weekends/No. of Housholds in the neighbourhood? Ans. Yes, You can assume few things. 

Assumptions: 1) Self Checkout = Assisted Checkout = Fast Checkout - I am not taking in account any difference in the average checkout period for the simplicity of calculation. 

2) Store has to be equipped to take the Peak hour capacity while deciding what would be the average no. of People waiting in the line.  This will help me to weed out the tedious calculation for weekdays/weekends/and then average hours. Also, operationwise it makes sense to consider Peak hours for better customer service.

3) I am not taking in account any Return department in account.

Now, No. of Cash Register (CR) = Average No. of Concurrent Shoppers waiting for their turn to checkout (CS) per 5 mins / (Acceptable No. of Shoppers per Cash Register at any point (ASC) ~ 12 ) = 125/12 = ~ 12 / 13

Calculating Acceptable No. of Shoppers per Cash Register at any point (ASC) = 5 mins to for 1 checkout so in one hour = 60/5 = 12.

Calculating Concurrent Shoppers (CS) = (Shoppers visiting on Peak day and Peak hour in 1 hour (SH) – 3000) / Average time taken by Shopper to browse, drop and walk with cart towards checkout (AT) – 1 hr) ) X Concurrency factor (50%, as an assumption). = 1500/ Hour = 1500/60 = 25 per minute X 5 = 125

Calculating Shopper Visiting every hour (SH) = (No. of Household in the neighbourhood – 60,000) X (% of Household visiting Walmart on Peak day ~ 50%) / (Total No. of Hours Walmart operates on Peak day ~ 10 hours) = 3000

Calculating Average time taken by Shopper to browse, drop and walk with cart towards checkout (AT) = (No. of items picked on an average by shopper ~ 50) X (Average time to Pick ~ 1 min) X (Average walking time ~ 10 mins) = 60 mins

Calculating No. of Household in the neighbourhood = Total Population of the neighbourhood / Average no. of people per house. = 60,000 (made assumption)

Tradeoffs/Points that I might have missed:

1)      Average checkout time taken as 5 mins which might be more.

2)      Walmart is having 50% traffic. This might be an over estimation.

3)      Concurrency factor that I took as 50% might be actually less than that.

 

Clarifications questions:

 

1- How big is the Walmart store (is it supercenter? neighborhood?)

2- What number of people are we planning on serving daily?

3 - Are we focusing on a register that needs an associate, or do it by yourself cashier?

 

1. My assumption is that we are thinking of opening a Walmart Supercenter

2. I also assume that the repartition between the regular cashier and the automated cashier is 50/50

3. I assume each regular cashier has per goal to serve 15 people every hour.

4. Automated cashier  25 people per hour 

5. I also assume that a Walmart supercenter gets at least 5,000 visits a day where each person buys at least 5 items.

6. let's assume Walmart supercenter are opened 10 hours a day

 

Calculation would be : 5000/ 10 = 500 people per hour 

 

250 people per hour for regular cashiers = 250/15 = 17 cashiers.

 

250 people per hour for automated cashier = 10 automated cashiers.

 

• Let’s say on a weekend around 10,000 people come to walmart .

• Out of which 2000 are just window shopping.

• Out of the left 8000 customers some come alone but some come in a group of two and sometimes the whole family joins in for shopping.

• If we take an average of 2 customers per billing, there will be an average of 4000 customers going to the billing counter.

• Let’s assume that walmart open at 11 and closes at 9, so it’s open for a total of 10 hours.

• Let’s assume that day hours see less rush and evening hours see more rush . so let’s assume that last 4 hours get 50% of the traffic.

• Hence total number of billing done in last 4 hours is 2000, i.e 500 per hour.

• On an average a billing takes 5 minutes , so the total number of billings that can be done in 1 hour is  60.

• So for doing 500 billings in one hour we will need 8.33 i.e 9 counters

 

Lets assume that we are looking for maximum rush that can be handled at checkout. Preferably this time would be on weekends when people purchasing in store are maximum.

We first need to calculate chekcout time.

(A) Checkout time  = scanning time per item * number of items to be scanned + payment time + receipt generation time + packing time 

Looking at formula above we need to calulate average checkout time. 

We'll be needing basket size for this

(B) Basket Size = Average number of itmes in the basket

(C)Average Checkout Time  = Scanning Time per Item * Basket Size + Payment Time + Receipt Generation Time + Packing Time ..... ..... .... .... From A & B

Lets assume that customer an wait maximum 15 minutes in line

Hence 

(D) Max(Wait Time) = 15 minutes = 0.25 Hours

Now we'll calculate how many customer can be accomodated in maximum wait time

(E) Number of customer accomodated in Checkout line = 0.25 Hours / Average checkout time ..... ..... .... .... From C & D

Now we need to find out total number of people checking out in 0.25 Hours

(F) 0.25 Hourly Traffic at Checkout =  Daily Store Footfall * Conversion Rate / Total Active Store Hours (Assume 10 AM to 10 PM) * 0.25 

Now, to calculate number of checkout registers needed, we will divide traffic by accomodated customers

(G) Number of Checkout Registers needed = 0.25 Hourly Traffic at Checkout / Number of customer accomodated in Checkout line ..... ..... .... .... From E & F

However, this approach should not be used in isolation.

We should always have one Multipler say M to the number of checkouts needed.

This can be calculated using

• Variability of demand/checkout in various hours
• Geography and Customer Preferences (Silicon Valley Customer will prefer lowe wait time due to busy schedules compared to other states)
• Competitive Landscape
• Survey Responses (Customer Rating especially on Checkout Experience)
• Type of Store Format (Small/Big... People tend to prefer lower wait time in small stores as basket size is lower)
• Distance from City (Farther the store higher tolerence to wait time)
• Penetration of Indirect Competition in the locality (E-commernce and Home Delivery Providers)