[LINK] Grocery Choice - what's the problem?

Sylvano sylvano at gnomon.com.au
Mon Jun 29 19:26:34 EST 2009

On Monday 29 June 2009, Richard wrote:
> David Boxall wrote:
> > If I'm not mistaken, the supermarkets that would have been involved in
> > Grocery Choice use scanners. For that, they'd need back-end databases of
> > products and prices. Once set up, transmitting part or all of that data
> > to Grocery Choice should cost zero (or near to it). Coles reckons it
> > would cost $8 million a year to do it twice per week (8 million/104 =
> > $76,923 - what were they planning to do? Tattoo prices, one-by-one, on
> > the buttocks of forked-stick messengers?).
> Without minimising the challenges of data quality, data matching,
> processing (since there's 20k -plus product lines in a large
> supermarket, and many of them subject to store-by-store pricing), etc ...
> But if memory serves me correctly, the data is already collected by
> Nielsen from the scanner records. So methinks that the data collection
> issue is not as insuperable as it was made out to be.

Yes, they do. 

But note the following commentary in a Woolworth's submission to the ACCC last 
year, which highlights some of those data issues that can exist.  

 refer to the following 208 character URL: 


Further, AC Neilsen acknowledges these limitations.

The suggestion that Woolworths and Coles collectively have an 80% share of the 
market is limited for the following fundamental reasons:

*􀁸 AC Neilsen scan data is generally based only on dry packaged groceries. 
This category is but a narrow component of the grocery industry to which the 
Grocery Price Inquiry’s Terms of Reference relate. Furthermore, given the 
range of product categories that the Issues Paper states are of interest to 
the Inquiry and the different retailers participating in the sale of different 
categories of products, it is not reliable to assume that the category of dry 
packaged groceries can be used as a “proxy” for the entire grocery industry;11

*􀁸 AC Neilsen notes that scan data is collected for only a limited range of 
products, and in some cases is based on estimates (ie HomeScan data); 

* Given that AC Neilsen’s ScanTrak data for Woolworths, Coles, Franklins and 
some of the independents is based on point of sale (POS) data, and AC Nielsen 
itself states that POS data is not suitable for measuring retail share of 
trade, it is difficult to see how these data can be cited as being definitive 
of any sort of “market share”; and 􀁸 

* AC Neilsen’s data is obtained only from some outlets, resulting in a 
tendency to over-report the “share” held by those that are more represented in 
the survey.

> I'm not so quick to dismiss the Coles estimate, but I think it could and
> should have been done.
> RC
> > I obviously have a lot to learn. Perhaps someone on Link can enlighten
> > me.
> >
> > <http://www.smh.com.au/opinion/how-grocery-choice-was-ushered-to-the-grav
> >e-20090628-d17b.html?page=-1> How Grocery Choice was ushered to the grave
> >
> >     * Phillip Coorey
> >     * June 29, 2009
> >
> > The Government might have taken some stick at the weekend for abolishing
> > the Grocery Choice website, but it pales against what it would have
> > experienced had the project gone ahead.
> >
> > As critics of Friday's decision stamp their feet and demand inquiries,
> > there are sighs of relief within the Government and cries of "I told you
> > so" from elsewhere.
> >
> > Grocery Choice was due to be launched on Wednesday.
> >
> > Ideally, shoppers would go online, compare the prices of thousands of
> > products at various supermarkets, and then head off to where the basket
> > of goods added up to be the cheapest.
> >
> > In reality, it would have been nothing like that and the Government
> > would have been inundated with complaints between now and the next
> > election about the website providing false information.
> >
> > Grocery Choice was a spin-off of Kevin Rudd's pre-election empathy with
> > voters over the cost of living.
> >
> > A Rudd government could not mandate lower grocery prices but it would
> > enhance competition, went the mantra.
> >
> > (The Government has committed to introduce unit pricing and change
> > planning and foreign investment laws to allow more players like Aldi
> > into the market).
> >
> > After the election, the website concept began as Grocery Watch. It was a
> > disaster, so the Government gave the task to the consumer watchdog,
> > Choice, with $13 million to make it happen.
> >
> > Choice was billing the site as giving consumers up-to-date information
> > on 1000 products, rising to 5000.
> >
> > Inside the Government, it was soon realised the information was never
> > going to be up-to-date but out-of-date and inaccurate.
> >
> > The supermarkets contended the IT systems to provide such information on
> > an instant basis do not exist.
> >
> > Woolworths, like almost every other supermarket chain, was not rushing
> > to embrace Grocery Watch. Woolies has about 800 supermarkets in
> > Australia. Prices for goods vary from location to location, depending on
> > the suburb the supermarket is in and whether a supermarket is in a price
> > war with a nearby competitor.
> >
> > The prices for some goods, especially fresh produce, can change in the
> > same day and vary from suburb to suburb.
> >
> > Woolies offered to provide, twice a week, an average price for products.
> >
> > For example, on Fridays it would tally how many cans of Coke had been
> > sold in recent days and how much money had been received for that Coke.
> > It would then divide the number of cans sold by the money received and
> > release an average price.
> >
> > That average would be higher than what a can of Coke cost in some
> > outlets and lower than in others. It would also be old information.
> >
> > Others had different problems. IGA, for example, is a franchise. A small
> > local operator would have neither the technology nor the time to funnel
> > through to Choice on a regular basis the prices of thousands of goods.
> > There were also legal ramifications to consider should the pricing not
> > be exact.
> >
> > Ultimately, Aldi and FoodWorks were the only supermarkets prepared to
> > try and make it work.
> >
> > Choice argues the big supermarkets could have provided real-time price
> > information for each location using the data scanned in at the checkout.
> > Choice says the cost would be negligible. Coles argued that just
> > providing twice-weekly average pricing would have cost it $8 million a
> > year.
> >
> > Whatever the case, the Government knew that in the end it would wear the
> > opprobrium. Chris Bowen had been the minister charged with trying to
> > make Grocery Choice work. Upon his elevation to cabinet this month, the
> > new Consumer Affairs Minister, Craig Emerson, found himself the
> > recipient of the hospital pass.
> >
> > Emerson, alarmed at the brinkmanship, called all the players to his
> > office on Friday to gauge whether they were in or out.
> >
> > Word of the meeting was sent out Wednesday. Choice didn't show up,
> > citing it was too busy putting the finishing touches on the website.
> >
> > At the meeting, it became apparent that even if the website was launched
> > Wednesday, its information would be neither complete nor accurate.
> >
> > Prices would have asterisks attached indicating they were averages only
> > but the shoppers would not notice that. They would blame the Government,
> > not Choice or the supermarket.
> >
> > Michael Jackson was dead and, Emerson concluded, so was Grocery Watch.
> > Emerson knew he would be accused of trying to hide bad news behind
> > Jackson, so he issued the press release as soon as possible on Friday so
> > it made the evening news. He was canned anyway.
> >
> > Grocery Choice was an example of a big promise which was always going to
> > be difficult, if not impossible, to deliver.
> >
> > -- Quote ends --
> >
> > I wonder whether many shoppers would take advantage of the data on
> > Grocery Choice. Putting the information in the public domain might,
> > however, motivate the supermarkets to keep prices down. Perhaps that's
> > the real problem.
> _______________________________________________
> Link mailing list
> Link at mailman.anu.edu.au
> http://mailman.anu.edu.au/mailman/listinfo/link


Gnomon Publishing

More information about the Link mailing list