On the impact of inventory accuracy improvements on sales Christoph - - PowerPoint PPT Presentation

• 1 -

ECR Brussels – October 25, 2018

Christoph Glock, Yacine Rekik, Aris A. Syntetos

On the impact of inventory accuracy improvements on sales

• 2 -

About us

■ Christoph Glock

l Professor Chair: Technical University of Darmstadt, Germany l Specialises in Inventory Optimisation and Warehousing Management

■ Yacine Rekik

l Professor Chair: EM-Lyon Business School, France l Specialises in Inventory Optimisation and Tracking (e.g. RFID) Technologies

■ Aris A. Syntetos

l Panalpina Chaired Professor: Cardiff Business School, Cardiff University, UK l Specialises in Statistical Forecasting, Demand Classification & Inventory Optimisation.

• 3 -

Background and objectives

■ Inventory inaccuracies: major issue in retailing and apparel industry. ■ Physical stock is (typically) less than what we think it is.

l Most reasonable assumption in retailing. Generally, stores are negative in terms of

stock.

l Thus, reconciling inventories may only lead to an increase in sales. l (We will see later that positive stock discrepancies are also possible, still leading

though to reduced sales!)

■ The problem has been established; we are not here to argue for its

existence.

• 4 -

Background and objectives

■ But rather:

l Assess the implications of the problem, or rather the implications of fixing the problem (phase 1); l Assess alternative ways of fixing the problem itself (phase 2).

■ Phase 1: What is the impact on (increased) sales if inventory accuracy is increased

by x%?

l How does inventory accuracy develop over time after a stock take? l Is there an optimal number of stock takes? Do too many stock takes negatively influence

inventory accuracy?

l What exactly constitutes this problem of inventory discrepancies?

■ Phase 2 (upon clearly establishing the implications): what are the strategies to be

employed (algorithmic driven, new identification technologies, counting, etc.) to fight the route causes of the problem?

• 5 -

Error free (r,Q) inventory policy

Safety Stock L1 L2 t Reorder Point: r Order Quantity Q Q Lead Time

• 6 -

Safety Stock L1 L2 t Order Quantity Q Q Lead Time Expected error free stock level

Impact of errors on the (r,Q) inventory policy

Reorder Point: r This is the visible stock behavior: POS (Real Demand) + Replenishment This is the invisible stock behavior: POS (Real Demand) + Replenishment + Skrink (Ghost Demand) Actual stock level subject to errors

• 7 -

Safety Stock L1 L2 t Order Quantity Q Q Lead Time Wrong

• rdering

timing and quantity

Impact of errors on the (r,Q) inventory policy

Lost Sales are more frequent Reorder Point: r And more importantly, they are not detectable

Errors act as “ghost” demand decreasing the stock level without generating a revenue: Inventory is controlled based on some visible wrong information, whereas sales are satisfied based on some correct but invisible information

• 8 -

Experiment I: original / ideal experiment

Sales (e.g.) 12 weeks before count Sales (e.g.) over 12 weeks “Test” “Control”

Stock Count

Sales 12 weeks after the inventory records have been “trued up”

No Stock Count

Sales 12 weeks after

Stock Count Stock Count Stock Count Stock Count

Stock Counts in the beginning and at the end of the experiment in both Test and Control stores

• 9 -

Experiment II: retailers with frequent periodic stock audits

“Test” “Control”

Stock Count Stock Count Stock Count Stock Count Stock Count

A comparison of sales on a same period of time, with different stock audit tasks in Test and Control stores In addition, several cases of Experiment I can be deduced from Experiment II

• 10 -

Experiment III: weekly stock audit

Week 1 … Week 12 Stock Audit Computer and Counted Physical stock during the audit are contrasted with what they should be (last counted physical stock + stock input – stock

• utput)

Computer Discrepancy: what was found vs. what is shown in the computer before the audit Physical Discrepancy: what was found vs. what should be in the stock given last week’s audit and stock movements

• 11 -

Empirical analysis

■ Currently working with 8 Retailers across Europe:

l NDAs have been signed and we are in various phases with regards to data transfer and analysis; l 4 Grocery retailers (supermarkets), 2 Apparel retailers and 2 other; l Customised reports to be produced for all participating retailers.

■ Initial results:

l 4 Grocery retailers (a, b, c and d) l +600,000 SKUs l 80 stores examined (40 test VS. 40 control) l We are still working on the best possible way to present descriptive statistics without

unintentionally disclosing individual retailer information (and identity)

l We present the (initial) results for the 4 Grocery retailers.

• 12 -

Results

PART I: GENERAL INSIGHTS (apply to all retailers)

• 13 -

Result 1

Independently of the experimental setting, inventory record inaccuracies are an important issue for all participating companies: across all grocery retailers, between 46% and 73%

• f the audited SKUs are subject to inaccuracies even if a stock take is performed very

frequently (as in Experiment II, or even each week as in Experiment III)

• 14 -

Result 1

Independently of the experimental setting, inventory record inaccuracies are an important issue for all participating companies: across all grocery retailers, between 46% and 73%

• f the audited SKUs are subject to inaccuracies even if a stock take is performed very

frequently (as in Experiment II or even each week, as in Experiment III)

73,33% 62,15% 46,44% 63,25% 0% 10% 20% 30% 40% 50% 60% 70% 80%

Retailer a Retailer b Retailer c Retailer d

Stock Inaccuracy Level per Retailer

• 15 -

Result 1: inventory record inaccuracies constitute an important issue

44,34% 37,80% 25,05% 39,16% 26,67% 37,85% 53,57% 36,75% 28,99% 24,35% 21,39% 24,09% 0% 10% 20% 30% 40% 50% 60%

Retailer a Retailer b Retailer c Retailer d

Negative Accurate Positive

It is not only a matter of shrinkage: positive discrepancy is not negligible and generally is caused by Information System (IS) manipulations and errors

• 16 -

Result 2

Inventory record accuracy leads to an increase in sales turnover between 2.36% and 4.68% at the retailers.

2,94% 2,36% 2,70% 4,68% 0,0% 0,5% 1,0% 1,5% 2,0% 2,5% 3,0% 3,5% 4,0% 4,5% 5,0%

Retailer a Retailer b Retailer c Retailer d

Sales Increase with better stock accuracy

• 17 -

Retailer a – sales increase of 2.94% in the Test stores comes from:

2,75% 8,61% 7,07% 0% 1% 2% 3% 4% 5% 6% 7% 8% 9% 10%

Fast_Mover Middle_Mover Slow_Mover

4,86% 3,79% 4,41% 0% 1% 2% 3% 4% 5% 6%

High_Discrepancy Middle_Discrepancy Slow_Discrepancy

0,67% 4,54% 6,01% 0% 1% 2% 3% 4% 5% 6% 7%

Accurate Negative Positive

• 18 -

Retailer b – sales increase of 2.36% in the Test stores comes from:

• 0,53%

3,89% 2,21%

• 1%
• 1%

0% 1% 1% 2% 2% 3% 3% 4% 4% 5%

Accurate Negative Positive

2,92% 3,20% 1,16% 0% 1% 1% 2% 2% 3% 3% 4%

High_Discrepancy Middle_Discrepancy Slow_Discrepancy

2,26% 0,62% 0,23% 0% 1% 1% 2% 2% 3%

Fast_Mover Middle_Mover Slow_Mover

• 19 -

Retailer c – sales increase of 2.70% in the Test stores comes from:

2,32% 2,58% 2,12%

0,0% 0,5% 1,0% 1,5% 2,0% 2,5% 3,0% High_Discrepancy Middle_Discrepancy Slow_Discrepancy

0,98% 1,66% 3,62%

0,0% 0,5% 1,0% 1,5% 2,0% 2,5% 3,0% 3,5% 4,0% Fast_Mover Middle_Mover Slow_Mover

0,66% 2,82% 2,35%

0,0% 0,5% 1,0% 1,5% 2,0% 2,5% 3,0% Accurate Negative Positive

• 20 -

Retailer c: inaccuracy level per category

■ SKU Categories subject to both positive and negative discrepancies

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%

B A K E R Y F R O Z E N F O O D P R O D U C E N E W S / M A G S / O T H E R G R O C E R Y P R O V N C A B I N E T P O U L T R Y P R O V N C O U N T E R M E A T H O U S E H O L D P L A N T B A K E R Y H E A L T H & B E A U T Y H O R T I C U L T U R E W I N E S & S P I R I T S F I S H C O U N T E R H A R D W A R E T O B A C C O C L O T H I N G B r a n d M O B I L E

Inaccuracy Level per Category Inaccuracy Level Negative Discrepancy

■ Can we infer that the sales increase offered by better stock accuracy is proportional to

the inaccuracy level?

• 21 -

Retailer c: sales increase per category

7,07% 5,74% 5,54% 4,58% 4,47% 3,44% 2,76% 2,61% 2,02% 1,79% 1,37% 1,18% 1,18% 0,98% 0,98% 0,66% 0,10%

• 0,35%
• 0,59%
• 1%

0% 1% 2% 3% 4% 5% 6% 7% 8%

W I N E S & S P I R I T S N E W S / M A G S / O T H E R H O R T I C U L T U R E B A K E R Y H E A L T H & B E A U T Y C L O T H I N G T O B A C C O F I S H C O U N T E R P L A N T B A K E R Y H A R D W A R E F R O Z E N F O O D H O U S E H O L D P R O D U C E P R O V N C A B I N E T G R O C E R Y P R O V N C O U N T E R R e t a i l e r B r a n d M O B I L E P O U L T R Y M E A T

Sales increase per Category

0% 10% 20% 30% 40% 50% 60% 70% 80% 90%

B A K E R Y F R O Z E N F O O D P R O D U C E N E W S / M A G S / O … G R O C E R Y P R O V N C A B I N E T P O U L T R Y P R O V N … M E A T H O U S E H O L D P L A N T B A K E R Y H E A L T H & … H O R T I C U L T U R E W I N E S & S P I R I T S F I S H C O U N T E R H A R D W A R E T O B A C C O C L O T H I N G B r a n d M O B I L E

Inaccuracy Level per Category Inaccuracy Level Negative Discrepancy

• 22 -

Retailer a: sales increase per Category

50 100 150 200 250 300 ALCOHOLIC… WATER AND… BREAD FRUIT VEGETABLES SWEETS BULK PRODUCTS YOGHURTS PACKAGED… YOGURT,… CIGARETTES BUTTER,… TEXTILE FRESH MEAT CHEESE, CHEESE SNACKS COFFEE-TEA-… PASTRY SALADS AND… CAKE SHOP HYGIENE… KITCHEN… CHEESE CANNED MILK, CREAM BULK READY… CHEESE DETERGENT FISH AND FISH… CULINARY… CLEANING… CLEANING HELP PACKAGING BISCUITS BODY CARE COFFEE-TEA MEAT COSMETICS DESERT FLOWERS PAPER ICE CREAM INSECTICIDES MILK, BUTTER,… FOOD ANIMALS CHILDREN'S FOOD BREAKFAST PICKLES, EGGS PARAPHARMAC… OVER CLEANING… FOOTWEAR… CHILDREN'S… HAIR PRODUCTS LEISURE… RESTROOM… SEASONAL… READY MEAL… READY MEALS CHILLED READY… SPECIALTIES FRESH JUICE HOUSEHOLD

Count of High and Middle Discrepancy SKUs per Category

High_Discrepancy Middle_Discrepancy

• 1,00%
• 0,50%

0,00% 0,50% 1,00% 1,50% 2,00% 2,50% CIGARETTES FRUIT VEGETABLES WATER AND SOFT… BULK PRODUCTS ICE CREAM MILK, CREAM BREAD DETERGENT FRESH MEAT HYGIENE PRODUCTS PAPER YOGHURTS CULINARY PRODUCTS CLEANING PRODUCTS COFFEE-TEA-COCOA BUTTER, MARGARINE,… OVER RESTROOM PRODUCTS CHEESE, CHEESE CHEESE SPECIALTIES BREAKFAST READY MEALS MILK, BUTTER,… CHILDREN'S PRODUCTS CLEANING HELP CLEANING ACCESSORIES COSMETICS FOOD ANIMALS CALLING CARDS LEISURE PRODUCTS FRESH JUICE HAIR PRODUCTS BODY CARE FLOWERS TEXTILE MATERIALS INSECTICIDES CHEESE HOUSEHOLD ECO PASTRY COFFEE-TEA FOOTWEAR PRODUCTS EGGS MEAT PICKLES, DESERT CHILLED READY MEALS FISH AND FISH… CHILDREN'S FOOD SNACKS PACKAGED MEATS SWEETS READY MEAL PACKED CANNED KITCHEN PRODUCTS YOGURT, DESSERT SALADS AND… ALCOHOLIC BEVERAGES BULK READY MEAL

Sales increase per category

• 23 -

Extra material / analysis PART II: SPECIFIC INSIGHTS (apply to some retailers)

• 24 -

Retailer d: service Level deterioration due to IRI

■ Service levels achievement is very sensitive to stock inaccuracies:

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% . % 1 . % 1 2 . 5 % 1 6 . 6 7 % 2 . % 2 5 . % 2 8 . 5 7 % 3 3 . 3 3 % 3 7 . 5 % 4 2 . 8 6 % 4 5 . 4 5 % 5 4 . 5 5 % 5 7 . 1 4 % 6 2 . 5 % 6 6 . 6 7 % 7 1 . 4 3 % 7 5 . % 8 . % 8 3 . 3 3 % 8 7 . 5 % 9 . %

Service Level PH Discrepancy Level

Achieved Service Level as a function of the PH discrepancy level

84,89% 61,46% 37,24% 93,02% 80,45% 43,47% 90,91% 85,68% 33,73% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Fast Mover Middle Mover Slow Mover

Service Level Sales Class

Service Level Achievement as a function of the Sales and Discrepancy Classes

High Discrepancy Middle Discrepancy Small Discrepancy

• 25 -

Retailer d: some SKUs end up as ‘ghost SKUs’- no sales during several weeks

■ Some SKUs faced a situation during some weeks where the computer stock record

was positive whereas the counted physical record showed a stock level equal to zero.

■ With a positive computer stock level, the replenishment process does not trigger

an order from suppliers leading to a physical stock level equal to zero in the forthcoming week and consequently leading to lost sales.

■ Each week, approximatively 10% of SKUs faced this situation and it could be

verified that the EPOS for these SKUs is approximately equal to zero in the forthcoming week.

■ 6% of SKUs faced this situation each week and a big majority of them end up as

‘ghost SKUs’ without any EPOS signal during the 12 weeks experiment.

• 26 -

Next steps and opening up the discussion

■ Complete the analysis for the rest of retailers; ■ Produce customized reports for each of the participating retailers; ■ Synthesize the results in the form of a white paper to be hopefully useful for the

retailing industry (and sectors within it). FOR DISCUSSION

■ What are the implications of our findings for your organisation? ■ How could you operationalise our insights? ■ What would you like to see addressed in Phase 2 of the project? ■ How do you envisage a benchmarking exercise in this area?

• 27 -

Extra material / analysis

APPENDIX