the lectern banner

By Steve Kelman

Blog archive

How to make reverse auctions even better

Most of my life is the life of a regular professor (teaching, research and stuff), but there are a few companies -- not many -- in the federal marketplace with which I work. One is FedBid, the reverse auction provider for government, on whose advisory board I sit -- because their mission is to save the government money, and it's a mission which I truly support (especially in this budget environment).

I recently had two informal sessions at FedBid's headquarters in Vienna, Va. One was with a group of junior FedBid employees who interact at the front lines either with the government buyers who use FedBid to do reverse auctions or with the vendors who bid in these auctions. The other session was with the senior managers at the company.

The conversations showed me just how frequently vendors considering bidding on a government requirement point out problems with the product specs that have been posted online for the auction. One of the services a market-maker such as FedBid provides the government is help in getting out information about poor-quality requirements before it is too late -- before an inappropriate product has been delivered at an inappropriate price.
With the senior managers, I discussed two ideas for ways to increase the value reverse auctions provide the government. One idea is to implement an eBay/Amazon-style system to allow government customers to give quick and easy past performance feedback on purchases made through their system. I discussed this idea in a blog post a while ago, spurred by having seen that taxis in Beijing, China, were introducing an instant customer feedback system where customers could press a like or dislike button on leaving a taxi. Perhaps customers could simply vote thumbs up/thumbs down on their overall satisfaction, or maybe do one vote for delivery timeliness and other for product quality. The idea would be to keep it simple so there are lots of votes that can establish patterns.

FedBid already has a "performance alert" system that allows customers to send in information, which is posted with the vendor's information online, about problems they've had. I argued that this was too punishment-oriented, and that a system should allow good vendors to be rewarded as well. Also, with a very tiny number of performance alerts, these can be influenced by idioysyncratic factors of buyer personality or some special features of an individual transaction that led the buyer to post an alert. In this sense, a broader system that encourages wider participation after more transactions would give a fairer picture, based on many responses. There remains the problem of getting people to vote, since it would have to be after the product (or service) was received. Perhaps FedBid could hold a monthly lottery for a $10 Starbucks gift card (comfortably under the government's gift limit) for customers who cast a vote after a transaction, which could increase the number of responses and make the ratings more useful.
Another idea we discussed was using reverse auctions for the price/labor rate part of more complex services procurements (such as IT service contracts). Reverse auctions are most suitable to buying commodities or simple services. But one could imagine using it as part of the buying process for more complex services. Here, bidders would submit a proposal, with many elements, as normal. However, for the price part of the proposal (either labor rates by labor categories and/or fixed price contract line items), there could be a reverse auction where each bidder's lowest price by labor category got established.

Those lowest bids would then be plugged into the rest of the proposal, as part of proposal evaluation. Especially in a world where the government often is too short-handed to negotiate about price, and where list price labor rates (say on GSA schedules) are not discounted, even for large jobs, as often as they should be, reverse auctions could be a powerful tool to get the government a better deal.
There was sympathy in the group of managers for both ideas. FedBid is thinking about each. FedBid friends, as a member of your advisory board, I hereby cast my vote for moving forward with both these approaches. 

Posted by Steve Kelman on Apr 08, 2011 at 12:09 PM

Operation Jump Start

FCW is proud to support this Feb. 9 event to help the soldiers of Operation Iraqi Freedom and Operation Enduring Freedom transition to civilian careers.


Reader comments

Mon, Dec 26, 2011 Eric Sullivan West Palm Beach

Steve, in terms of FedBid's mission to save the Government money, would you please comment on the following? First, FedBid typically calculates savings based on an Independent Government Estimate (IGE). It is widely the case in the federal acquisition community that IGE’s for commodities (the primary substance of reverse-auctions) and commercial items in general are typically based on a single commercial quote to provide the items required. Thus, IGEs for these items are neither independent nor do they originate within the Government. In short, that doesn't make for a good baseline or benchmark for estimating savings. Second, and much more importantly, calculating savings off of an IGE does not equate to real savings. To calculate the real, realized savings of reverse-auctions, one would need to compare the savings from reverse-auctions to what would have been saved by the Government in the absence of reverse auctions. In short, there should be a valid benchmark. Calculating savings off of an IGE and indicating that the use of reverse-auctions saved a certain amount as if that entire amount is new savings that would not have otherwise been realized ignores the fact that standard procurement methods other than reverse-auctions typically already save a significant amount off of the IGE due to the benefit of competition requirements. To demonstrate real savings, the outcomes of reverse-auctions need to be compared to a benchmark showing procurement outcomes and savings in the absence of reverse-auctions. Third, there seems to be no accounting for the potentially costly impact of the fees charged by FedBid. FedBid imposes a fee up to 3% on sellers when the winning bid is below the IGE. Any smart business is going to account for this fee in their bids and pass along the cost to the Government, thus resulting in instances where the Government pays up to 3% more than they otherwise would have due to this fee. Thus, these fees will offset any savings and the question of whether any real savings are being realized cannot be answered unless a benchmark to procurements in the absence of FedBid is used. Making matters worse, companies are now accounting for an additional fee up to 3% in any quotes they provide to the Government which are subsequently used as an IGE. The result: up to 3% higher prices less any savings from reverse auctioning. There are instances where reverse-auctions have indeed cost the Government more than they would have without using a reverse-auction because of this and these occurrences are not being accounted for in the calculations of savings from FedBid.

Sat, Apr 9, 2011 Steve Kelman

Lu, thanks so much for your comment. The elements cited in the OFPP statement are, of course, important for past performance ratings for big contracts, but apparently the OFPP guidance, at least on the four elements, extends even to buys under the simplified acquisition threshold. Part 13 of the FAR specifically authorizes simplified past performance evaluations for simplified acquisitions. Dan Gordon is interested in vitalizing the government's use of past performance. If this memo is standing in the way of a simple rating system for small buys, it should be changed. Dan Gordon, ball in your court!

Fri, Apr 8, 2011 Luther Tupponce

Writing on behalf of FedBid, we completely agree that a simple commercial-type rating system would be a useful addition to the online marketplace. Unfortunately, current federal policy makes it difficult to implement. For example, the May 2000 OFPP memo regarding past performance best practices (cited in the recent Jan 21, 2011 memo on improviing past performance assessment) states that all rating systems should track 4 basic assessment elements and use 5 basic ratings. It also stresses the importance of narrative elements. Absent some change in policy, there is little chance a simple commercial-type rating system will be deemed permissable. For a more comprehensive discussion (from way back in '06), check out this link:

Please post your comments here. Comments are moderated, so they may not appear immediately after submitting. We will not post comments that we consider abusive or off-topic.

Please type the letters/numbers you see above