Retail Audit

1
Retail AuditUses Abuses () ofPresented by
Candan Schabio
2
My sales are growing at 10 per annum !! Profits
are growing! I am doing very well, am I not?
600
Maybe not so
500
RETAIL AUDIT ENABLES ME TO ANALYSE MY
PERFORMANCE WITHIN TOTAL MARKET
400
337
300
269
213
167
200
130
100
100
161
146
133
121
110
100
0
1999
2000
2001
2002
2003
2004
MYSELF
OTHERS
3
THE RETAIL AUDIT PROVIDES

• A continuous, independent measurement of facts
• Retail sales to consumers
• Distribution in covered channels
• Stock levels and out-of-stock
• Pricing
• A bird eye look at channel development, trade
  concentration
• Insights to both sales, marketing and top
  management about the direction of the market
• Strategically (long-term analysis)
• category growth, sales importance of channels,
  segment analysis, annual review, product launch /
  segment entrance decisions
• Tactically (short-term analysis)
• pricing, distribution gains, opportunity gaps,
  success of new launch, cannibalization, key
  account plan

ALL, drilled down from manufacturer gtgt brand gtgt
SKUs
4
THE RETAIL AUDIT DOES NOT PROVIDE

• Consumer insights (i.e. can not answer WHY
  consumer has chosen that product, WHAT type of
  consumer buys that product and where, what are
  motives and preferences in consumption pattern)
• A 100 correspondence to manufacturer
  ex-factories (shipments data) because does not
  cover the whole market but only some and always
  retail channels.

• Retail audits should not be used as the "holy
  book", without which a decision can not be taken
• Retail audits should be used as a tool for making
  "better less risky decisions"

5
FIVE STEPS OF RETAIL AUDIT
5.Analysis interpretation Reports delivered to
clients
This product is a result of a process
4.Statistical expansion of sample to universe
3.Data collection Establish best system of data
capture
2. Design a representative sample How many of
each store type do we need to represent the
universe with a given margin of standard error?
1.Establish the Retail Universe Define store
types and gain information on shop numbers
turnover (retail census)
6
(1)

• Prior to setting up a retail audit Panel one
  needs to establish the Retail Universe.
• Available statistics hardly ever meet our own
  definitions or requirements
• THEREFORE..
• A retail census is ALWAYS undertaken by ACNielsen
• The estimation process should provide up to date
  universe information and measure the trend.
• Retail Census is conducted twice a year at Area
  Sampling (Rolling Census) to track these changes
  at continuity

7
RETAIL UNIVERSE - ROLLING CENSUS

• A census provides a SNAPSHOT of the retail
  universe at a SINLGE POINT IN TIME.

• This is NOT the case in developing economies,
  especially countries in a transitional stage.
  The Universe is changing very quickly!
• Thus the need for a continuous ROLLING CENSUS

8
(No Transcript)
9
Retail Panel is specially selected subset of the
total universe in such a way to reflect all
Universe characteristics.
Reported at 4 Standard Error

• Sample design ? The selection of a sample of
  stores on which a retail index is based involves
  the following steps
• Determination of total sample size (nationwide)
• Determination of total sample size per channel
  and per region
• Determination of panel stores by CELL (each
  channel within each region)
• Assignment of the specific shops to be signed up
  by field to the retail sample

Reported at 10 Standard Error
Not reported (i.e. small groceries in Zagreb)

• Only way to ensure full representation of the
  universe characteristics is the random selection
  process
• Each store in the universe has the same
  probability of being selected in the sample

10

• Electronic data collection
• Direct sales data from organized chains

• Data collected by the auditor (at SKU level)
• Forward stock Reserve stock
• Purchases
• Retail selling price

• SALES VOLUME Opening Stock Purchases -
  Closing Stock

11
Understanding the Audit-Cycle

• It is critical that one understands well the
  audit-cycle and WHAT it represents

February Report
February
March
January
1, 2,3,4....26,27,28,
1, 2,3,4....27,28,29,30
1, 2,3,4....27,28,29,30
Stores visited every 30 days
Thus, the shares ()in February report actually
cover a period of approx 60 days (i.e. Jan to Feb)
12
Understanding the Audit-Cycle

• It is critical that one understands well the
  audit-cycle and WHAT it represents

Therefore

• The nature of a retail audit cycle dictates a
  specific TIME LAG
• It would take some time to see the true picture
  of specific sales drives
• Especially true for market shares (Vol,Val)
• Price Distribution data is more dynamic and
  always reflects the reported period.
• A new product launch in APRIL will not appear in
  the findings until the May report. It will most
  probably only show-up in June.

13
Projection of raw data from sample to universe
preparation of final data
Reported Regions

• Data is expanded by CELL
• Each store within a sample cell has a projection
  factor representing the whole region in that cell
  (i.e. each Grocery in Central region is
  representative for total Groceries in Total
  Central)

Sup / Hyp
Large Groc
Small Groc


South


West






East





Central





Zagreb
Exp.Factor



Cell Universe



Cell Sample
Outlet Types
14
Problems limitations of retail audits

• Nothing is perfect. Audit is no exception
• Often, issues result from the fact that data
  users do not acknowledge the mechanics of retail
  audits

• Lack of invoices
• Changing retail structure
• Under-coverage due to other handlers not included
  on the retail audit panel (I.e. specialized drink
  stores)
• Limitations in pick-up of very regional products
  with low / narrow distribution (i.e. typically
  dairy products, beer etc)
• Parallel Imports/Exports
• Illegal products
• Institutional sales e.g. Army, Hospitals, etc
• Wholesaling, semiwholesaling activities

15
Problems limitations of retail audits

• A store promotion in specific group of stores,
  which are NOT INCLUDED in the panel, will have no
  effect on the data

• Heavy "PIPELINE FILLING" will be picked up much
  slower by the audit

THE FLOW OF GOODS FROM MANUFACTURER TO RETAILER
TO CONSUMER
ACNielsen Universe
16
Retail Audit Coverage
Kiosks
Petrol Stations
Drug Stores
Total Market (example)
For each market and each category within the
market channels and their importance can differ
to a bigger or a lesser degree.
4
1
11
Open Markets
2
Groceries
CashCarry
65
7
On-Trade
10
17
Retail Audit Coverage
Kiosks
4
Drug Stores
11
ACNielsen covers only Groceries, Drug Stores,
Kiosks. Therefore maximum coverage retail audit
can achieve is 80.
Groceries
65
18
Coverage Analysis
Coverage analysis compares manufacturers
shipment data with research sales data to
establish the pick-up or coverage of
particular brands.
Average coverage during this period 87
19
Coverage Analysis - Months Moving Total
20
Coverage Analysis - Moving Annual Total
21
Issues around Coverage Analysis

• There are some issues associated with coverage
  analysis
• pipeline delays between shipments and sales in
  Retail Outlets
• proportion of shipments being sold outside the
  Retail Universe
• brands which have very skewed distribution
  pattern (e.g. rural or particular region focused)

• Product cycle New, Established, Declining
• Time-lag due to audit methodology
• Parallel imports and exports
• Non invoiced goods
• Brand-share and distribution
• Consistent brand and period definitions

22
Product Lifecycle - New Product Typically
Under-covered
Pipeline filling
23
Product Lifecycle - Declining Product Typically
Over-covered
Pipeline emptying
24
FACTS -DATA TYPES
(5)
Data delivery Analysis
Statistical expansion Processing

DATA collection
Design a representative SAMPLE

• Sales and Share (volume and value)
• Price
• Product Availability (numeric and weighted)
• Distribution
• Out-of-Stock
• Net distribution
• Purchases (by retailers)
• Stocks and Forward Stocks, Days Supply
• Average Sales per Shop (Handling)
• Adjusted Average Sales per Shop
• Sales per Point of Distribution (SPPD)
• Share in Handlers (some databases)

Establish the UNIVERSE
calculation
25
Where do I begin to analyse data? How to continue?

• It is impossible to look critically at all
  numbers in an audit report
• Learn to be selective and pin-point numbers that
  look "odd" or "different

• Start at the total market level and then drill
  down
• Do not try examine all issues at once, but
• rather examine each in depth, and
• then go back to the beginning and start again
  with the next issue

• Do not jump to conclusions based on a single
  period's readings
• Usually one can talk of trends after 6 months
  data
• Retail Audits are all about TRENDS

26
Example identify the growing / declining channel
and look at your performance within those
channels

SHARE OF TRADE FOR MINERAL WATER


60

1.9

50
Am I on the right channel? What about my
competitors? If I am loosing, who is gaining?
Different pricing strategy? A special
distribution gain? Etc etc
Growth to C
-
stores


40
from Supermarkets


30

20

10

0




Supermarket
Supermarket
Small
Total

Horeca




gt50,000 sq ft
lt50,000 sq ft
Grocers
Convenienc
e






Yr Ago
5.4
20.1
5.2
17.2
52.2






Yr to Date
4.7
18.3
5.1
17.9
54.1

1
27
SALES AND SHARE QUESTIONS

• What is the size of the category?
• How important is each segment, manufacturer,
  brand?
• What is their strength by trade channel?
• How are my sales and share performing over time?
• How did they react to
• Change in price?
• Change in distribution?
• New product introduction?
• Change in competitive action?

28
Brand X SHARE GAIN LINKED TO PRICE REDUCTIONS
Total Croatia
Price
Share
29
NUMERIC VS WEIGHTED

• Numeric Based on percentage of the stores in the
  universe (store count)
• of all stores in the universe that handle the
  product
• Weighted Based on category sales
• of product category sales (value or volume)
  accounted for by stores handling product

DISTRIBUTION QUESTIONS

• What percentage of the stores are handling our
  brands vs. our competitors?
• What are the opportunities and where are they?
• Can I increase my sales by increasing the
  distribution of the best selling products and
  decreasing the distribution of slower moving
  items?

30
EFFICIENCY OF DISTRIBUTION

• Brand A Brand B
• Num Dist 80 Num Dist 40
• Wtd Dist 76 Wtd Dist 64
• Efficiency of distribution wtd/num distribution
• Brand A 76/80 0.95
• Brand B 64/40 1.6
• Brand B is more efficiently distributed It is
  getting to 40 less stores but only missing 12
  of the business!

31
DISTRIBUTION EXAMPLE
SKU As volume growing as distribution increases
Groceries
32
Conclusions

• Retail audit is the only powerful tool
  contributing to less risky sales and marketing
  decisions, yet its mechanics needs to be well
  acknowledged.
• What channels are covered, what are not, why
• Are there any category specific issues, any local
  market implications which may interfere with
  retail audit methodology (i.e. any emerging
  channel, which is not covered but already
  important? Change in tax applications? Parallel
  imports, across border sales, incidence of grey
  market etc etc).
• This is only possible by the maximum utilization
  of the 5th Step, Analysis, through client x
  agency interaction.