Additional Comments by Senator Xenophon

Additional Comments by Senator Xenophon


1.1             This inquiry was established to investigate the establishment, management, operation and closure of the GROCERYchoice website.

1.2             A key 2007 Labor election promise, the grocery price comparison website was intended to provide consumers with up-to-date information regarding supermarket prices around Australia.

In announcing the website, then Opposition Leader, Kevin Rudd, said that:

If Labor forms the next government of Australia, the ACCC will be directed to publish a periodic survey of grocery prices at supermarkets for a typical shopping basket, including family staples like biscuits, bread and baked beans.[1]

1.3             The Australian Competition and Consumer Commission (ACCC) was given this responsibility as an extension of its role in ensuring transparency in the marketplace and to create greater competition between market players.

1.4             $12.9 million over four years was budgeted towards the website, with a total $3.6 million spent within its first year by the ACCC. Subsequent costs for lawyers' fees, data collecting agencies and payments to CHOICE bring the total expense of the GROCERYchoice website to $7.693 million, with an estimated contingency of $700,000. 

1.5             The website was launched on 06 August 2008, covering 500 products in 600 supermarkets across 61 regions.

1.6             The website's introductory page read:

GROCERYchoice is an Australian Government initiative. It provides practical grocery price information to help consumers find the cheapest overall supermarket chain in their area. With the large number of grocery items available at each supermarket, consumers often find it difficult to determine which retailer offers the cheapest prices overall. GROCERYchoice helps consumers compare the general price levels of supermarket chains in their area.

1.7             However, a number of key issues plagued the website from the start. The data was not real-time and therefore was out of date for consumers before it was even uploaded on the website; the 61 regions into which Australia had been divided were too large and had no relationship to consumers' shopping behaviour; there were discrepancies between the size of the supermarkets included within the survey; and, the data provided was based on total cost of a secret basket of goods that consumers may or may not be purchasing.

1.8             Between its August 2008 launch date and the end of the year, 'hits' or visits from consumers GROCERYchoice had fallen from over 446,000 in September to under 105,000; suggesting that grocery shoppers found the website of little value.

1.9             The decision to transfer operations and management, and funding, of the website to Treasury occurred in November 2008, after it was agreed that the ACCC was not able to undertake certain required changes and was in a difficult position given its regulatory role.

1.10        Treasury outsourced the website's operations on 5 January 2009 to consumer agency, CHOICE, who said they would re-design the website and address issues faced by the ACCC, including providing additional features to add value to the grocery price information; more regular price updates across more supermarkets; better price information including unit pricing and personalised baskets; consumer interactivity; and, increased transparency.[2]

1.11        CHOICE worked to develop an improved version of the website for launch on 1 July 2009; however, GROCERYchoice was closed down on 26 June 2009, without discussion with CHOICE, by decision of the new Minister for Competition Policy and Consumer Affairs, the Hon Dr. Craig Emerson MP. 

Consumer behaviour

1.12        Consumers are driven to shop at a particular outlet not only by price but by a myriad of reasons.

Consumers can take price information on the website into account with a range of other factors important to them when determining where to shop, including the location and accessibility of supermarkets to them, the quality of fresh produce, the product and produce range, and their assessment of the quality of service.[3]

1.13        According to Associate Professor Frank Zumbo from the University of New South Wales, shoppers in metropolitan regions tend to travel no more than five kilometres to purchase supermarket products.

Prof. Zumbo—The GROCERYchoice website as operated by ACCC was broken into 61 regions which were very large regions of Australia. In the Sydney metropolitan area it went from one side of the Sydney metropolitan area to the other side, some of the regions, and the reality is that consumers operate on a three to five kilometre radius. That is the market definition used by the ACCC. It is commonly accepted that consumers will not travel more than those three to five kilometres in the metropolitan area. So the GROCERYchoice website was not local in its approach, it was too general.[4]

1.14        Further, Associate Professor Zumbo stated in his submission to the Committee that:

Consumers already have a "general" feeling about grocery prices. What consumers want is specific pricing information that they can use to identify the cheapest products in the cheapest individual supermarket in their local area.[5]

Implementation of GROCERYchoice

1.15        Understanding of consumer behaviour is crucial to configuring an effective consumer-focussed website; however, it's understood no modelling was undertaken by the Government, Treasury, the ACCC or CHOICE to ensure GROCERYchoice would be relevant to the Australian public:

Senator XENOPHON—Has any modelling or analysis been done by Treasury in terms of the benefit to consumers from giving consumers full price transparency?

Mr Martine (ACCC)—Not that I'm aware of.[6]

This questioning was followed up by Senator Barnett during the Committee hearing:

Senator BARNETT—Do you have any evidence to suggest that it would work? There must have been some document, some modelling undertaken, some information and research undertaken by the department so as to say to the government, ‘Yes, this is going to work.’ Where is that evidence? Can you identify it for us?

Mr Chisholm (Treasury)—In response to questions on notice we provided some documents which had been developed by CHOICE which were made available to government, to the best of my recollection, in August-September 2008 setting out a proposal for a CHOICE-run GROCERYchoice website.

Senator BARNETT—I am not talking about the CHOICE—

Mr Martine (ACCC)—You are talking about the original—

Senator BARNETT—The original one, taken by the ACCC.

Mr Martine—Treasury obviously at the time—it would have been in the lead-up to the 2008-09 budget—provided the government with our advice. In terms of the question of whether we undertook detailed modelling, not being there at the time, I guess it would have been that we did not, and there would have been no detailed economic modelling undertaken on it. That is probably fair to say.

Mr Chisholm—To the best of our knowledge, no.

Senator BARNETT—You are saying that there is no evidence, no modelling, that you just sort of came up with the idea. You do not just spend $13 million over a four-year period and put it in the budget and say, ‘We hope it is going to work.’ There must have been some basis for that. Can you please advise the committee of the reasons for it and of the evidence or modelling you relied on to put forward a budget proposal of some $13 million over four years.

Mr Martine—If you are talking about quantitative economic modelling, where one looks at a proposal and undertakes an assessment of its impact on consumers or prices, anything like that, then the answer in this case is no, we did not undertake quantitative analysis. Just sitting here today, I would probably scratch my head trying to work out what sort of quantitative analysis you could undertake.

Senator BARNETT—What analysis did you undertake, Mr Martine?

Mr Martine—Like all budget proposals, in fact any proposals the government is considering, we provide the government and the relevant ministers with our views and advice on the merits or otherwise of the proposals under consideration. In terms of the original GROCERYchoice website that the ACCC were to run, we did provide that advice.

Senator BARNETT—I think it would be fair for one to assume that, if there were no modelling or business plan—and there is evidence that apparently that was not undertaken—some might argue that it is policy on the run or ill-conceived.[7]

1.16        The Committee was advised by the ACCC in its Answers to Questions on Notice, provided on 13 November 2009, that:

modelling of effectiveness of a policy proposal would form part of the policy decision for Government.[8]

Further, the ACCC stated that:

decisions on policy and the parameters of a project are made by the Government with advice from appropriate government departments.[9]

1.17        Subsequently, the commencement date provided to the ACCC by the Government was a policy parameter[10] and as such it appears not once that the ACCC questioned or discussed with the Government as to whether it could be extended. This means that adequate planning and preparation for the website was not able to be done.

1.18        Bids to collect the data were received from five companies, including Retail Facts, who was successful in the tender, and Informed Sources.

Retail Facts' quote for $4.669 million (excluding GST) was the second cheapest of the quotes, but was $2.694 million greater than the quote provided by Informed Sources, of $1.975 million (excluding GST).

1.19        The ACCC's explanation of awarding the tender to the more expensive quote of the two companies during the Committee hearings was:

Mr Cassidy—...while the Informed Sources tender was lower in price than the Retail Facts tender, which we accepted, we were under some time pressure to get the GROCERYchoice website up and running and we did have some doubts as to whether Informed Sources was going to be able to deliver, particularly on the data collection side, within the time frame we were operating in.

Senator BARNETT—What time pressure were you under, Mr Cassidy?

Mr Cassidy—The government was keen for the website to be up and running as soon as possible.

Senator BARNETT—How soon?

Mr Cassidy—We were working with an indicative time of having the first collection done so it could be released in early August.

Senator BARNETT—When was this discussion? You are talking about a six-week period to get it up and running.

Mr Wing—By the time the contract was let, there would have been about six or seven weeks to do the surveys.

Senator BARNETT—So the government gave itself a self-imposed deadline to require it to be established within that six-week period. Did you advise them of the obvious cost differential? Did you advise the government of the implications of their push to rush this forward and to have it up and running so quickly?

Mr Wing—No. It was a policy and we had a budget so we just ran within that.[11]

1.20        The ACCC went on to state that it did also not believe Informed Sources would be able to prepare a field force team within the six-week timeframe, as compared to Retail Facts who already had a field force in place.

However, Informed Sources did advise the ACCC that it would indeed be able to prepare a team within the required timeframe:

Senator XENOPHON—Sure. But weren’t you satisfied at the outcome of that subsequent supplementary meeting with Informed Sources that they would be able to deliver within the time frame that was requested?

Mr Pearson—We had no doubts about their integrity, their process, the fact that they could do work, because they are doing an extremely good job for us now with petrol. The risk was too high for us, because if they could not do it—

Senator XENOPHON—So are you saying that the reason that Informed Sources did not get the job, even though it was $2.7 million cheaper, was based on a belief, notwithstanding that you have said that they have a track record of doing a good job for the commission and, I think, they were also responsible for undertaking the mammoth task of price monitoring after the introduction of the GST?

Mr Pearson—They have done a wonderful job for us. In fact, we signed another a contract with them either in June last year or June this year. We extended our petrol monitoring for two years. It was purely within that time frame and the fact that we believed that they did not have the staff ready to put on the ground ....


Senator XENOPHON—Can I go back a step. The ACCC’s relationship with Informed Sources goes back how many years—since the introduction of the GST?

Mr Pearson—A fair while.

Senator XENOPHON—About a decade?

Mr Pearson—Yes.

Senator XENOPHON—Have they ever let you down before in terms of their commitments to the ACCC?

Mr Pearson—Not that I am aware of.

Senator XENOPHON—I suggest to you that the answer is no. Someone’s track record would be important to you in terms of assessing a tender and assessing their ability?

Mr Pearson—Definitely. We would not have put as much work into assessing them, we would not have invited them back and we would not have asked the supplementary questions if we did not feel that on one level they were technically capable of doing the job. We went to those extra steps because of that. There is a notion that we did not go a bit further. We could have just looked at the initial proposal and said it is going to be a problem because they did not have the people on the ground. We could have gone straight to Retail*Facts. But we did not—and that was because of our relationship. So that did come to bear and we did look at it.

Senator XENOPHON—Although they previously had an unblemished record with the ACCC, you were holding back because you did not think they could deliver. That is effectively what happened, that is the reason why they did not get the tender, but I still do not understand the basis on which you made that decision.

Mr Pearson—We just did not believe they would be able to get sufficient well-trained staff on the ground by the August date. That was our concern. We had to weigh up the fact that we had one potential operator who can confirm they had those staff and another one that said they could train the staff in time. We did not believe that time frame would be sufficient.

Senator XENOPHON—I think this goes to the nub of it: if a successful tenderer does not deliver the goods for the ACCC, there would be contractual consequences. There would be a breach of contract and you could potentially pursue damages against an entity that does not deliver the goods. Also, that would presumably put a big black mark against them in terms of any further work with the ACCC.

Mr Pearson—Yes, without a doubt.

Senator XENOPHON—Here is an organisation that has a long track record of providing massive surveys, not just on fuel but also on groceries, for the ACCC in the past. They have never let the ACCC down before.[12]

1.21        On the question of cost, Senator Barnett during the Committee Hearings asked the Chairman of the National Association of Retail Grocers of Australia,
Mr John Cummings, a retailer himself, how much a private company would likely have paid for the same work to be conducted.

Senator BARNETT—What do you think would be more realistic?

Mr Cummings—I would have thought that it would be well under $1 million.

Senator BARNETT—To undertake the work?

Mr Cummings—Absolutely. And that is not only collection of data but also interpretation of the data and then setting price files.


Senator BARNETT—Their initial bid was $4.669 million and subsequently it signed up at, I think, $5 million-plus. The Informed Sources tender was $1.975 million, and you are saying that in fact it should be about $1 million or less to undertake that sort of work.

Mr Cummings—That is what I think private enterprise would be expecting to pay.[13]

1.22        Mr Cummings also advised the Committee that private enterprise would also likely demand regular auditing, spot checking and reviews to ensure best practice; however, no audits of GROCERYchoice were undertaken while it was operating.[14]

Senator BARNETT—Despite the fact that the contract referred an opportunity for an audit, such an audit was never undertaken. What is your view about that, and should such a contract ever have been signed up?

Mr Cummings—We were always concerned about that, and we were always concerned about the data collection and the integrity of that data, especially when it came down to some of the smaller stores that were being gone into and having that price data taken out of. One of our concerns there was, of course, a smaller store might not have the entire range. If you take my particular instance, you have a larger independent, or Coles or Woolworths, that would sit at somewhere around 22,000 or 23,000 SKUs. In my particular instance, we sit on about 16,500 SKUs. Seeing as nobody knows what is in the prices that they are taking, if I do not range one of those products and they come into my store, what do they actually mark down? Do they take a larger size, a smaller size? Was that veracity ever checked? Again you have that same problem if you look at a shelf label. Even for one who is in the industry, when you go to short descriptions it is very difficult to actually figure out what the short description stands for. Quite often you have to go to an item number or a PLU to get what the actual product is. There is a whole pile of issues in there that make the data collection difficult.

Senator BARNETT—All right. I want to go back to the initial website the ACCC was setting up, and I have read your submission and some of the compelling arguments that you have put. You have got 500 products, 600 supermarkets and 61 regions. Frankly, in your view, was it ever going to work?

Mr Cummings—I still cannot see what the relevance is of a product available in Merredin to a product in Broome—which was the area that you are looking at. It would be like travelling from Greece to London to do your grocery shopping in the afternoon.

Senator BARNETT—So the way it was designed was a fatal flaw?

Mr Cummings—I have no doubt that the consumers who looked at it voted with their fingers and decided to go away from it because it did not deliver them any meaningful information.[15]

1.23        In addition to being significantly more expensive, Retail Facts already had a contract with Woolworths when it took on the GROCERYchoice tender, and Retail Facts' potential conflict of interest regarding the integrity of the data collected was raised.

1.24        The ACCC advised the Committee in its Answers to Questions on Notice that:

The majority of data collectors engaged by Retail Facts to collect price information for the ACCC were not used for the collection of price information for Woolworths. There were a very small number of exceptions to this in remote regional areas.[16]

It went on to say that in these instances, price collections for the ACCC and Woolworths were undertaken in different weeks and were never performed in the same store.

1.25        Associate Professor Frank Zumbo argued that the ACCC should never have been given operational management of the website to begin with.

Prof. Zumbo—The ACCC was the wrong body to give this website to. I believe it was poorly planned in the sense that there was a lot of time pressure being placed on the ACCC to deliver. I think haste is not a good thing, especially when we are told that the ACCC spent $3.64 million on this website, which was completely wasted.[17]

Limitations of GROCERYchoice

1.26        Overseas, grocery price comparison websites appear to operate relatively well.

In the United Kingdom, a privately-funded and owned company operates a grocery comparison website which allows consumers to select items for the country's four leading supermarket chains: Tesco, Asda, Sainsbury's and Ocado. One feature of the website includes the opportunity to for users to place an online order based on their search. [18]

The Italian Government has set up a similar service for grocery price comparison, whereby consumers can send an SMS from their mobile phone to check the average prices of different foods in across the various regions.[19]

And, in July 2009, Ireland's National Consumer Agency announced that it intends to set up a grocery price comparison website with real-time information for consumers.[20]

1.27        While the usage of these websites is small or yet to be determined – it's estimated around only 3 percent of the UK population use[21] – its approach appears to be decidedly more user friendly and useful to consumers in the information it provides.

1.28        The ACCC and CHOICE cited difficulties with the supermarket chains as a reason for the ineffectiveness of GROCERYchoice.

According to CHOICE's submission to the Committee:

Woolworths was resistant to the idea from the beginning. They cited a range of issues from technology constraints to trade practices breaches, all of which CHOICE was willing and able to address. ALDI and FoodWorks were supportive and cooperative and while they cited technology as a challenge, they were willing to find solutions and work with CHOICE. Coles and Franklins were initially cooperative and open to the idea, but became increasingly distant.[22]

1.29        According to the Australian National Retailers Association, which represents companies including Woolworths, Coles and Franklins, providing price data on a real time basis is not realistic and would incur significant compliance costs:

No retailer has a centralised data system which records in real time the prices of grocery items sold across the chain.[23]

1.30        The ACCC similarly stated in its July 2008 report, Report of the ACCC inquiry into the competitiveness of retail prices for standard groceries, that the head offices of Coles and Woolworths:

...set the shelf prices for most of its products in each of its stores. It also sets promotional prices, although not all stores necessarily have the same promotions at any one time. The local store manager can reduce prices below the standard shelf price in a range of circumstances including clearances of discontinued stock and stock approaching its use-by date and as a response to local competition.[24]

1.31        However, supermarket IT systems do enable checkouts to price in real time. Associate Professor Frank Zumbo in his submission to the Committee stated that:

Since scanned pricing information through their checkouts is in real time, the major supermarket chains could provide real time pricing information to the public if they chose to do so.[25]

He advocates supermarkets commit to full pricing transparency:

Given that they have some of the most sophisticated IT systems that enable them to collect all pricing information scanned through their checkouts, it is clear that the major supermarket chains have the technical ability to implement full price transparency through their own websites in relation to all products sold in each of their supermarkets.[26]

1.32        Another issue which added to the apparent irrelevance of the website for consumers is that data appears to have been out of date before it was even released online:

The ACCC's version of the GROCERYchoice website only provided a very limited monthly "snapshot" that was out of date as soon as it is put on the website. As supermarket shoppers will typically shop on at least a weekly basis, they want the most up to date information possible about the cheapest local supermarket and products in their local area.[27]


In the ACCC's version of the GROCERYchoice website, the products included in each month's survey were rotated, which meant that consumers had no ability whatsoever to compare prices month to month.[28]

1.33        Associate Professor Frank Zumbo explained in his submission that:

The ACCC's version of GROCERYchoice contained out of date pricing data and extremely generalised information that failed to give consumers any meaningful data that consumers could seek to rely on to help reduce their grocery bill. The key failure of the ACCC's version of the GROCERYchoice website was that it failed to assist consumers to find the cheapest individual local supermarket or to find the cheapest individual products they may be looking to buy during their next supermarket visit.[29]

1.34        Further, the 61 regions as set out by the ACCC's initial version of GROCERYchoice meant that supermarkets hundreds of kilometres away, in some instances, were being compared to each other.

The state of Tasmania was broken up into three regions, for example:

Senator BARNETT—You can understand their concerns, particularly in Tasmania where you had three regions and you were comparing supermarkets in St Helens, Scottsdale and Launceston, for example, which are up to 100 kilometres apart. In terms of the benefit or merit of that to consumers it is, frankly, an absurdity.[30]

1.35        The ACCC advised the Senate Committee that:

Regions for the GROCERYchoice website were determined using information from the Australian Bureau of Statistics. Each region represented an aggregation of a number of Local Government Areas.[31]

1.36        However, as previously stated, shoppers tend not to venture further than 5 kilometres to buy their groceries and as such this data became irrelevant.

Furthermore, the website did not detail which supermarkets had been surveyed, but rather kept the details area-generic.

1.37        In comparison, supermarkets employ an established practice that was not set to cease under GROCERYchoice and which provide consumers with greater, more-localised information:

Woolworths, Coles, Aldi, Franklins and the thousands of independent grocery retailers already spend many millions of dollars a year through print, radio and television advertisements to tell customers which items on "special" represented the best value directly in the customer's area.[32]

1.38        Woolworths advised the Committee in its response to Questions on Notice that it alone publishes 8 million brochures each week about its store specials and prices.[33]


1.39        While the intent behind the GROCERYchoice website was genuine, the practicality of it in terms of its operations and its usefulness to consumers on a day-to-day basis seems to have been an issue since its establishment.

1.40        While these may have been identified and clarified had more time been allocated towards the planning of the website rather than the rush in which it seems to have been conceived, it appears that any grocery price comparison website will not provide consumers with any additional useful knowledge unless it is real-time, suburb specific, advertises the daily specials and lists all products in store.

1.41        The decision to close GROCERYchoice appears to have been rushed and without full consultation by the newly-appointed Minister to the portfolio, Dr Craig Emerson, with the website's new providers, CHOICE.

While Minister Emerson clearly identified that the website in its current form, without real-time, localised information, was not able to provide consumers with relevant information, the Minister could have facilitated additional dialogue between all supermarket chains and CHOICE regarding possible improvements to GROCERYchoice before making his decision.

1.42        It was raised during the Committee hearings that discussions have occurred to look at establishing a similar website to GROCERYchoice; however, to date it remains unclear whether this will eventuate.

1.43        According to new OECD price data, food prices in Australia have increased 41.3 percent since the start of 2000, so there's no question more needs to be done to address these price hikes which appear to be driven largely as a result of a market duopoly. Accordingly, a price comparison website based on real-time information can only be part of the solution.

1.44        A grocery price comparison website was considered initially because Coles and Woolworths have such a dominant market share and it is this dominance which raises competition issues.

1.45        If the government wants to address high grocery prices, it needs to improve competition in the groceries sector overall, by requiring supermarkets to provide full price transparency to enable and empower consumers with pricing information before they shop; by enabling greater entry to the market by independents and small retailers; and by addressing geographic price discrimination, predatory pricing and other anti-competitive practices.


Recommendation 1

That the government improves competition in the groceries sector by requiring supermarkets to provide full price transparency to enable and empower consumers with pricing information before they shop, enabling greater entry to the market by independents and small retailers; and by addressing geographic price discrimination, predatory pricing and other anti-competitive practices.

Recommendation 2

That the system of dealing with tenders by the ACCC be improved and more transparent given the curious and unsatisfactory explanation given for why Informed Sources was not awarded this tender on the basis of cost and its prior work with the ACCC.

Recommendation 3

That prior to any government-run or government funded price comparison website being established in the future, significant time be allocated towards planning, modelling and consultation so to ensure effectiveness, relevance and requirements of such a website.

Recommendation 4

That companies providing bids for government projects identify any potential conflict of interest and that they be required to provide detailed information on how confidentiality and integrity of the project will be adhered to. Further, that an ongoing audit of their work be carried out at random intervals throughout the project, regardless of whether an incident has first arisen to cause suspicion.

Recommendation 5

That the Trade Practices Amendment (Guaranteed Lowest Prices – Blacktown Amendment) Bill 2009 is enacted, to deal effectively with the anti-competitive practice of geographic price discrimination.



Independent Senator for South Australia

Navigation: Previous Page | Contents | Next Page