Complexity of Using BI in Retail
Using business intelligence [BI] in a retail environment presents
a complex multi-dimensional problem.
We can best understand this complexity by looking at product and
sales management from different perspectives:
Most business users analyse by dimensions such as time, cost centre
and product or service.
In retail, there are thousands of SKUs across hundreds of outlets,
which need to be measured against weekly forecasts.
Operating in such a fast paced and highly competitive industry,
retail managers are increasingly demanding near real-time analysis.
The volumes of data needed to support this capability can be massive,
compared to a like sized business in another industry.
The multiplicity of data points scale in retail business intelligence
into billions of data cells.
For instance, daily sales data for 1,000 SKUs in 250 stores can
generate nearly 100 million data cells in a one year period. Calculating
and storing analytical reports results in data explosion, which
in turn can result in storage requirements increasing by a factor
of 10, with nearly 1 billion potential data nodes.
Obviously, not every store sells every item every day - however
the data problems cascade to needeing more processing power and
more sophisticated data management techniques.
Fortunately, the advances in massively parallel processing and
data warehouse technologies are now making analysis of these
vast amounts of data a practical proposition.
Retail is a very fast-moving environment. Product fashion cycles
have reduced from months to weeks, with many reatilers now operating
on just-in-time stock levels as low as 4 weeks of supply. 10 years
ago, 16 - 20 weeks stock cover was more the norm.
"Newness" is a key merchandising strategy requiring effective
and targeted decision support for managing old stock out of the
business. This can be the difference between making a profit or
In an effort to rationalise data analysis capabilities, retailers
are asking the question - what data should we store, at what level
and for how long?
This challenge of data relevance is fast becoming one of the most
significant issues in retail business intelligence.
The sheer volumes of data required by retail merchants and finance
executives to report which regions and products are performing best
and worst, and where the key actions need to be taken, requires
a new generation of interactive and dynamic analytics where the
software continually trawls data looking for exceptions.
This operationalised BI reduces the incomprehensible clutter of
data into specific suggestions for actions to gain competitive advantage.
Changes are not only evident in the retail business model. There
have also been significant developments in retail IT structures
that support business intelligence.
Radio frequency identification [RFID] is transforming the accuracy
of inventory data, and promises to remain one of the highest value
developments in retail business intelligence over the next few years.
Supply chains are extending world-wide, with Asia now contributing
for nearly 50% of global exports in the world’s clothing industry.
UK supermarkets sell fresh produce from Thailand, Ecuador and New
Zealand. These complex supply and cost structures must be closely
monitored and managed.
Such remote stock sourcing also means that decisions about short
life inventory investment must not only be finely tuned, they must
be made sometimes months in advance.
The level of predictive confidence required for this decision making
environment demands even more timely and accurate data, and it must
be presented in meaningful ways.
In summary, retailers are becoming ever more conscious of controllable
costs. Selling price deflation means constant pressure on margins,
with cost control often being the only way to maintain profitability.
This requires constant, effective analysis and alerting to any
exceptions to profitability rules.
Whilst collating and presenting relevant data may be time-consuming
and complex, the potential for bottom-line saving is significant.
With web based tools supporting collaborative demand planning,
local intelligence can be incorporated into centralised purchasing.
In addition, key performance indicators can drive supply chain
and stock management processes, with alert messaging ensuring action
is taken at the earliest possible moment to retain profit.
Retail business intelligence has evolved from reporting, to planning
and tactical action to business performance management [BPM], and
now operational performance management known as Operational BI.
At this micro level, predictive analytics is continuously applied
to a body of data and manipulates it using sophisticated algorithms
to provide dynamic suggestions for tactical action.
This can be used for optimising:
- Supply chain
Vendors using OBI typically report increased sales volumes, margin
percentage and stock turn – all at the same time. The bottom-line
impact of combining even small improvements in these key performance
indicators [KPIs] is potentially immense.
In spite of such compelling potential, it is important to remember
that technology alone cannot assure the success of a retail business.
However, Business intelligence is a tool, which when coupled with
sound business strategy and effective execution processes can propel
a retail chain to higher levels of performance and competitive edge
than ever before.
Next: Barriers to Retail
Back To Top
More Detail on Retail BI Strategy, Program & Technology
Index | BI Strategy | Complexities
| Barriers | Retail
IT Systems | BI IT Optimisation
| BI Evolution | Integrated
Retail BI | Implementation
| Business Benefits | RMS