By
Nikki Baird, Managing Partner
Through
a special arrangement, presented here for discussion is an excerpt of a
current article from Retail
Paradox, Retail Systems Research’s
weekly analysis on emerging issues facing retailers.
Last
year, Retalix announced a new solution and partnership around pricing – one
involving Retalix’s price execution, KSS’s price optimization, and Willard
Bishop Consulting’s accelerated strategy program designed especially to
help mid-size retailers get a pricing strategy off the ground. A year later
at Retalix’s User Conference, one of the first retailers to take on the
solution, Super S Foods, provided an update on where they’ve been and where
they’re headed as they make their way along the pricing learning curve.
In
Super S’s case, they identified three pricing strategies. One was for highly
competitive markets where a Walmart or an HEB were close by; another for
geographically isolated locations (where Super S might be the only game
in town); and one for Hispanic stores.
The
operator of about 50 stores in south-central Texas started on its new pricing
path when it decided to centralize pricing in 1990. Although centralized
pricing improved gross margins by one percent, Super S became aware that
even more was possible when they started seeing that one price really did
not fit every store for the same SKU. They realized they needed tools to
help them create more granular prices
– both the analytics side of setting prices, and the execution side
of passing the right price for the right item to the right store. To do that,
Super S implemented Retalix’s Category Analyzer and HQ for price execution.
With
those solutions in place, Super S embarked on the second phase of the learning
curve – the part where their capability grew exponentially beyond
what they were capable of before. Just the Retalix implementation helped
the chain post seven percent comp store sales gains in 2006 and 2007. But
Super S found itself wanting to move forward faster than Category Analyzer
could support – for the analytics to produce a good result, it needed
12-13 weeks of history. This was too long to wait in a highly competitive
market, so Super S turned to KSS and its price optimization solution, which
provides daily insight into prices, and weekly insight into the price elasticity
for each item. With not anywhere near all categories rolled out, 2008 is
on track to beat even the results of the previous two years, with 12 percent
comp store sales on track.
On
the accelerated part of the learning curve, Super S is now learning that
their three price strategies don’t fit as well at the store level as they
thought – that there are actually opportunities to apply these strategies
at a category or even sub-category level, so that one store might have
all three strategies at work under one roof, as each category requires.
Other
retailers who have already gone through the full price optimization rollout
have started talking about better integration into other parts of the organization –
integration to inventory, actuals as well as planned, along with better planning
and optimization of promotions, bringing in marketing as well as merchandising. These
next steps are more incremental improvements on top of what is a significant
step-level change to pricing processes. That’s where the learning curve starts
to flatten out again – at least, until the next big innovation hits.
Discussion
Questions: How does the learning curve around pricing strategies differ
from other implementations at retail? What common mistakes are continually
occurring as retailers refine their pricing strategies to include category
analysis and price optimization tools? What are the particular challenges
to mid-size retailers undergoing price strategy changes?
[Author’s
commentary] An interesting note, and one that we’ve found to be true of
learning curves in general, is that it’s nearly impossible to skip a step
along the curve – there is no leap-frogging on a learning curve.
For pricing, the critical first step that you can’t skip is where you identify
your price strategy. I’ve heard time and again from retailers implementing
price optimization that they started down the path only to realize that
they had to step back and start over a little bit, because they realized
they didn’t have a price strategy to implement to begin with. It’s tempting
to think that price optimization will provide you with the analysis that
will help you identify what price strategy you should follow, but it simply
doesn’t work that way
– there isn’t one “magical” price strategy that a retailer should follow.
It has to be tied into your company strategy, the reality of your economics,
your competition, and the brand promises that you make to your customers.
It’s an objective, not a math problem.
BrainTrust
Discussion Questions
Poll
Loading …
The Super S case reveals a number of truths about pricing strategy, I think.
First, there is a world of difference between automating price management and price optimization. The latter is a modeling process that requires some fairly serious data processing. The former may be improved with IT tools too, but it incorporates much less insight into the process.
Second, true price optimization is inherently a granular process that could yield different best solutions for different stores, shopper segments, geographies, categories, trips, seasons, etc. It may also identify different sets of KVIs (known-value items) for each of these segments, adding a layer of complexity.
This level of insight has pitfalls, however. Such fine granularity will rarely be practical to apply in the field. And too-frequent price changes on KVIs may undermine shoppers’ perceptions–sometimes called the “price image” problem. So it becomes desirable to limit the number and frequency of pricing actions. Grouping stores based on shopper segmentation as Super S did is a good place to start.
Nikki’s closing point is a good one, I think. Price optimization is not a tool, it’s a business practice that happens to be enabled by technology. Organizations must define their objectives clearly and build disciplines around those, or they will find themselves optimizing prices to the constraints defined for the model, which may be worse than random.
Herb Sorensen, Global Scientific Director, Consumer and Shopper Insights, TNS North America, was quoted in yesterday’s discussion that “Supermarkets typically have 30,000 to 40,000 distinct items on their shelves, of which less than five percent contribute more than half the store’s sales. In fact, the typical household only buys about 400 distinct items in an entire year, many of those purchased over and over, month after month.”
Those facts are equally applicable to the pricing discussion. Regional, neighborhood or even store by store pricing does not have to be an overwhelming task. If a supermarket is carrying 40,000 items, it is likely that the pricing on 39,000 of them is largely irrelevant. There are key items that shoppers shop, those they buy regularly, those they know the price of, those that determine which stores they are going to stop at. The rest of the items, they will “pick-up” because they are there at their chosen primary supermarket.
Drug Chains and Convenience stores have proven that people will pay extraordinary prices for the convenience of grabbing an item quickly and easily. The supermarket is the same way. Most items in the store are not worth the shoppers’ effort to make a separate stop.
Certainly, data generated pricing models take time to implement and time to read and optimize. The greater the detail–that is, moving from hundreds of items to thousands of items–the more difficult it becomes. However, the real danger surfaces when the competitors also adopt similar pricing tools. If all follow the same basic strategy, the price changes will become neutralized. Therefore, the winners will be those operations who may be smaller rather than larger, and can quickly read, analyze and change the pricing dynamics.