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

David Boxall david.boxall at hunterlink.net.au
Mon Jun 29 21:35:27 EST 2009

On Mon, 29 Jun 2009 at 18:59:46 +1000 Richard wrote:
> ...
> Without minimising the challenges of data quality, data matching, 
Given that each product is identified by a unique number (the bar code), 
what matching challenge?
> processing (since there's 20k -plus product lines in a large 
> supermarket, 
Which don't all change in price often. After the first mass download, 
data flows would be moderate.

Grocery Choice was to concentrate on only 1000 lines at the start, 
rising to 5000. Is going the whole hog really beyond the realm of 
> and many of them subject to store-by-store pricing), etc ...
Which is the point, I think. Why buy at Coles Mosman if you're going 
past Coles Neutral Bay, where what you want is cheaper?
> ....
> I'm not so quick to dismiss the Coles estimate,
If they're going to pay $8 million per year, then I'll consider coming 
out of retirement.
>  but I think it could and should have been done.
Agreed. I reckon our elected representatives have been conned.

Another point that arose in conversation: the accumulating mass of data 
would lend itself to mining, which could expose much about the 
retailers' behaviour. Another reason for their obstructionism?
David Boxall                    |  Dogs look up to us
                                |  And cats look down on us
http://david.boxall.name        |  But pigs treat us as equals
                                                   --Winston Churchill

More information about the Link mailing list