When we first set up RevaHealth.com our business model was very similar to Google AdWords. Clinics would effectively bid for position in our featured listings section according the treatments that they delivered and the market in which they operated. The model was very elegant for a number of reasons.
Firstly and most importantly it eliminated the need for us to price, and pricing is difficult. In fact pricing is incredibly difficult. This is because in order for the price you set to maximise your revenue it needs to reflect the value that the customer places on the product or service, and value is subjective. That means that the right price for one person is the wrong price for the next. (Take a read of Eoghan McCabe’s related post on the Contrast blog, You Are What You Charge.)
In our case we do business across 90 countries and about 1,000 treatments. Pricing across these 90,000 combinations would be impossible. So the second reason we liked the auction model was that in a liquid market auctions discover the value for you. They don’t discover the value to the highest bidder but they do uncover the value to the second highest bidder (unfairly referred to as the loser). Pricing never uncovers this.
Thirdly, the auction model eliminated the risk of over pricing and excluding potential customers. This is always the worry of normal pricing, however with auctions it is the customer that sets the price, making over-pricing impossible. The customer can still feel that the price they’re paying is expensive, but they come up with the price and they choose to pay it.
So Why Didn’t Auctions Work For Us?
OK, so in a way the auction model kind of worked for us, but as hard as we tried we simply couldn’t get it to scale. There were four problems:
- Auctions require a liquid market. Put simply an auction with one bidder achieves nothing. When we started RevaHealth.com we were effectively dealing in a number of niche markets in the medical tourism field, and there wasn’t the required demand or liquidity in each market to ignite its own auction.
- Auctions require the bidders to be actively involved in the auction. Our customers simply didn’t want to be that heavily involved, they would rather have paid a fixed fee without the complexity of bidding. In fact, the only time our customers did change their bids was when we picked up the phone and told them they should.
- Auctions required educated bidders. Our customers at the time weren’t sophisticated marketers. They didn’t keep accurate and diligent marketing records and work out their ROIs. They spent their nights poring over medical records not marketing data. As a result they weren’t really sure what the value of our product was and therefore were uncertain about what they should bid.
- Finally the most important factor for why the auction model wasn’t the runaway success that we had envisioned was that our customers weren’t used to doing business that way.
What We Changed
We needed to find a way to sell to the people who didn’t “get” the auction model for one reason or another, i.e. the majority of our potential customers! We introduced featured listings with fixed annual pricing at four different price points. This was an immediate success and required a substantially shorter sales cycle as the product is already well understood by the customer base.
There were a lot of disadvantages to doing this, particularly for our customers. In an auction model, the risk they took on was linked directly to the value they received, i.e. they received leads or enquiries at a cost that they determined themselves. By paying a fixed annual fee irrespective of the number of enquiries generated, the link between the risk and the value was broken. However, the fact that the annual fee product was easy to understand and sell outweighed all of the disadvantages, for us at least. We still maintain the auction model as a premium feature on top of our featured listings for the customers that can see the value in it.
Could the auction model be applicable to your market? Have you tried it already? Let us know what the key determining factors in setting your prices are in the comments below.
Does the auction model drive the fix price listing? In that does it provide data on what the value of a listing is.
In that the auction model provided us with a base line that allowed us to model the pricing so that we didnt’ lose money. However the auction model created 10s of thousands of price points so the pricing model could only model it in the crudest form.
personally i’m quite happy using an auction model, as long as it works!
i’ve been bidding for position in SERPs on sites using the ‘auction model’ since the days of pre-Google search engines such as Go — for me the most important factors are 1) the bidding process should work smoothly, and 2) the bidding process should be stable
the first few months we were with Reva, the scripts were so slow it was near enough impossible to do ‘test’ bids against competitors — just to make simple changes to our bids on a dozen treatments would take the best part of a morning, most of the time being spent waiting for pages to refresh or windows to open and close — methodically checking out how our bids fared against competitors wasn’t an option
then followed a period where the speed of response improved, but there were numerous minor problems with changes in the method used to implement bids — sometimes it was necessary to phone Reva to find out how to change our bids, as the links and buttons were not merely counter-intuitive but actually misleading (eg they were no longer active, or hadn’t yet been implemented)
currently the bid ‘position’ shown after ‘saving’ a new bid seems to stay the same regardless of how much one bids (whether €1 or €999), which means one is bidding blind — of course one could run tests by doing searches, but in the past the the theoretical position shown on the bidding page has not always been reflected in the actual position in the search results — and now, according to the positions shown on the bidding page, it’s possible to have a higher country position for a bid than for a town within that country, which is hard to make sense of
during the periods of relative stability it’s been possible to monitor competititors’ bids, and prices would start getting higher — but then there’d be more changes and more problems, and prices would drop back to base levels again — it’s as though the software never gets out of beta before a new version comes along
of course, i’ve no way of knowing whether other customers have had similar problems — perhaps you’re right in thinking the size of the market and the nature of the customer is such that an auction model is not appropriate for Reva — but, to be blunt, it sounds to me a little like an attempt to shift the blame on to the customer for failures in product design
however, to end on a more positive note — the postings on Twitter, and here on the blog, have been an important factor in reestablishing Reva’s credibility after some disheartening experiences bogged down in our account management pages
Thanks for your comment. Your feedback is much appreciated.
Owen Darbishire explains why auctions are a great idea if at all possible in the ‘Building a Business’ series of lecture from Oxford University. His lecture on negotiating can be found at …
http://www.sciencelive.org/component/option,com_mediadb/task,view/idstr,OX-tag_2009_02_09_172102_812_manstud_bab2008_audio/Itemid,96
Simply put, in a room full of people, you can achieve the maximum price for a finite resource. You could think of this as, “you get to sell the dumbest person in the room”. Where resources are not finite then aiming for a median price works better for the reasons above. Google et al can create enough liquidity to make a ‘market’ work.
Foreign companies list on US stock exchanges for the same reason – there aren’t enough buyers and sellers in their local market.
Creating a market is only for the super-big and by their nature, they are only going to appear when there is complete supremacy in a niche. So I guess the question then becomes – which markets have enough liquidity to make an auction work? Adwords, bing, ebay et all make their money be charging for access to their market. The next best thing to being a market is figuring out a way to win in that market. I guess there is a lot of work that could be done in integrating tightly with these guys. In an ideal world, Google would allow advertisers to buy ‘enhanced’ profiles on Revahealth while creating ads.
Reverse auctions seem easier to establish. Voices.com, elance.com, odesk.com have all achieved this. Many small businesses don’t know what a sales pipeline is, much less conversion rates and average dollar value. They can however understand the concept of bidding on an real piece of work. At a guess, I’d say this is a much better strategy to pursue, especially when still in a growth phase.
Informative article and nice tips. i recommend!