## Price floors, second price auctions and market dynamics

### September 25th, 2010

One of the things that is often discussed but not often written about are the market mechanics that surround the new RTB enabled exchanges & SSPs. From a design perspective most marketplaces these days have adopted some modified form of a second-price auction. The winner of the ad impression pays the seller not his actual bid, but the second highest bid.

Second price theory works as follows: Imagine that I’m selling a Monet painting. There are people that want to buy it and each has a maximum price he’s willing to pay but of course doesn’t want to pay a penny more than he has to to get the actual painting. If I tell my buyers that they’ll only pay the second highest price then each can safely give me their maximum price because they know they’ll only pay the amount they need to to beat the next highest guy. That sounds nice right? Second price auctions maximize revenue and make everyone’s life easier and create simple and efficient markets.

The problem is, reality doesn’t seem to quite follow the theory when we look at advertising today. Take a look at the below yield curves for two publishers coming in from two different exchanges. Both of these exchanges use a second price auction model.

The way the yield curves are read is pretty simple. On the X-axis you have a CPM bid-price and on the Y-axis you have the % winrate — the probability that you will win an impression from this publisher for this ad-size if you bid this price. On the right we see a relatively logical and predictable curve — you can’t win much below \$0.40, at \$0.50 you win about 10% of the time and above \$2.00 you will win about 70% of the time. The higher the price point, the less demand and hence the higher the winrate.

On the left you see a rather curious pattern, below \$0.90 one wins nothing whereas at \$1.00 you get 80-90% of all impressions. Obviously not an efficient market. In this case, the publisher has set a price floor of about \$0.90 for the inventory.

This is pretty common these days in RTB — publishers are absolutely terrified about cannibalizing their rate card and are hence forcing a “premium” for RTB buyers. This is a particularly interesting case because if you look at the actual win-rates it’s pretty obvious that there is barely any demand above their actual floor… or in other words, it just ain’t worth that much. So why would a publisher set a floor and sacrifice their revenues?

### What the publishers are afraid of

Fundamentally publishers are afraid that advertisers will “game” their auctions and the net result will be lower effective CPMs.

Let’s take a completely theoretical auction on Google AdExchange for a 29 year old male user seeing his third ad on a specific page about ford pickup trucks on CNN.com who has been identified as an in market cell phone shopper. Four buyers are interested in this specific impression…
- Ford buying the keywords “pickup trucks” — values the impression at a \$5.00 eCPM (derived from a CPC price)
- AT&T targeting in market cell phone buyers using third party data at a \$3.00 CPM
- A branded Kraft campaign that is trying to reach 29 year old males at a \$2.00 CPM

In second price theory each would submit this price, Ford would win the impression with it’s \$5.00 bid and pay \$3.00 to match AT&T’s second price.

Here’s the problem… frequency & an abundance of supply. Users see multiple ads. Frequency is also by far the most significant variable for optimizing response to ads. Hence each buyer is only interested in hitting this user a limited number of times with their ads.

For the sake of argument, let’s assume that our 29 year old Male, Joe, is browsing a number of different articles on CNN.com and there are 10 different opportunities to deliver an ad to him. Let’s also assume that our buyers continue to bid on each and every impression. To model frequency let’s assume that after each impression delivered the advertiser will bid half as much for each subsequent impression. Under these assumptions here’s how the bids would pan out over a number of auctions:

Auction Ford AT&T Kraft Price paid
#1 5.00 3.00 \$2.00 \$3.00
#2 \$2.50 \$3.00 \$2.00 \$2.50
#3 \$2.50 \$1.25 \$2.00 \$2.00
#4 \$1.25 \$1.25 \$2.00 \$1.25
#5 \$1.25 \$1.25 \$1.00 \$1.25
#6 \$0.63 \$1.25 \$1.00 \$1.00
#7 \$0.63 \$0.63 \$1.00 \$0.63
#8 \$0.63 \$0.63 \$0.50 \$0.63
#9 \$0.31 \$0.63 \$0.50 \$0.50
#10 \$0.31 \$0.31 \$0.50 \$0.31

What we see is that for each auction the publisher’s revenue is maximized with CPMs starting at \$2.50 but then very quickly dropping down to \$0.63 cents.

Now here’s where theory and practice start to separate. In the above scenario, Ford pays an average of \$1.72 CPM to show this user four ads. This is quite a bit higher than the average CPM and Ford decides to try a new bidding strategy to try to reduce his cost. Rather than always putting out his maximum value to the ad exchange he holds back a little bit and decides not to bid until the 6th impression.

Here’s what happens:

Auction Ford AT&T Kraft Price Paid
#1 no bid \$3.00 \$2.00 \$2.00
#2 no bid \$1.50 \$2.00 \$1.50
#3 no bid \$1.25 \$1.00 \$1.00
#4 no bid \$0.63 \$0.50 \$0.63
#5 no bid \$0.33 \$0.50 \$0.33
#6 \$3.00 \$0.33 \$0.25 \$0.33
#7 \$1.50 \$0.33 \$0.25 \$0.33
#8 \$0.75 \$0.33 \$0.25 \$0.33
#9 \$0.38 \$0.33 \$0.25 \$0.33
#10 \$0.19 \$0.33 \$0.25 \$0.25

What you see in the above is that Ford now buys four impressions, slightly further down in the users session but for an average CPM of \$0.33… 81% cheaper than were he just to submit a bid on each and every impression.

Of course this is a hypothetical situation, but it does show a point — if demand is limited then for buyers a very simple bidding strategy can have a large impact on cost and greatly increase ROI.

Let’s now imagine that the publisher realizes the advertisers are doing this and sets an artificially high floor price to try to protect his margins — \$1.50. Instead of accepting a paying ad he will show a simple house ad instead.

Here’s now what the auction looks like:

Auction Ford AT&T Kraft Price Paid
#1 no bid \$3.00 \$2.00 \$2.00
#2 no bid \$1.50 \$2.00 \$1.50
#3 no bid \$1.25 \$1.00 psa
#4 no bid \$0.63 \$0.50 psa
#5 no bid \$0.33 \$0.50 psa
#6 \$3.00 \$0.33 \$0.25 \$1.50
#7 \$1.50 \$0.33 \$0.25 \$1.50
#8 \$0.75 \$0.33 \$0.25 psa
#9 \$0.38 \$0.33 \$0.25 psa
#10 \$0.19 \$0.33 \$0.25 psa

The publisher has certainly succeeded in driving up the average Ford CPM — back up to \$1.50 from the earlier \$0.33. CPMs are down to \$0.65 but the overall average is actually *down* from \$0.71 CPM.

Here we see how the artificially high price actually ends up driving overall revenue down by limiting the number of impressions sold significantly.

Of course those of you paying attention would point out that a lower floor could potentially increase revenue over the no floor situation!

### So what gives?

Direct marketers have known the above for years. This is why very few pure response driven buyers pay rate-card for the ESPN home page. What scares publishers is the idea that branded buyers could start doing the same thing. The knee-jerk reaction in this case is to set arbitrarily high floor prices on marketplace inventory to try to protect the channel conflict.

Floor prices themselves aren’t necessarily that bad. In fact, there are good reasons for setting them. First, there are brand buyers that are paying rate card for guaranteed inventory — there is no reason to expose that same inventory to those buyers on a marketplace for a much lower rate. In other cases, a publisher might just be better off displaying internal house ads rather than showing a crappy blinky offer that annoys visitors at a low CPM.

For example, imagine you’re ESPN and you have a new “Videos” section where you just started running pre-roll ads \$30 CPM. If ESPN were to show the crappy blinky offer (not that they have the demand problem), they’d make \$0.10 for a thousand impressions and probably risk losing a small percentage of their audience in the process. On the other hand, a benign house ad announcing the new “Videos” section of the site would both increase site-traffic and visitor loyalty, but actually generate revenue. If they get 5 clicks per thousand impressions on the house ads they’ll actually be able to net out \$0.15 in advertising revenue from the 5 pre-roll impressions served on the video site (and even more if users watch more than one video).

### Going back to market mechanics

Let’s go back to Market mechanics for a second. Today what we need to avoid are knee-jerk set crazy high floor prices — a floor price that is too high will simply result in lower RPMs for the publisher. Publishers must understand that there is a significant pool of demand, specifically the ROI & response driven side, that simply won’t buy the inventory for rate-card prices.

In the end it comes down to information and controls. Publishers need to understand the market mechanics and the yield that they can derive from their inventory. At the moment I’m not aware of any major marketplace providers that supply this type of information.

I have a lot of thoughts on the tools and controls a pub should have and also on how one might change market mechanics away from a true 2nd price auction to efficiently deal with this — but I think this post is getting long enough… I’ll save that for the next one (which hopefully won’t be 5 months in the making!)

• http://www.sociocast.com Albert Azout

Mike, as always, great post!! You should post more often

Do you think that when we look at a publisher’s future RPM mix, we will start to see higher yield on non-guaranteed inventory due to audience provider attaching high-demand audience segments to these impressions? (Of course, assuming the publisher’s traffic generates or overlaps with quality audiences).

In your examples, it is the audience behavior that increases the bid prices. Should the performance of these segments be high, will we eventually see a better alignment of both channels?

• Matthieu

Hi Mike,
Nice article!
There is a small mistake in your auctions : auction 3, AT&T is \$1.50 and not \$1.25
I’m waiting for your next article on market mecanics! The next tools could be inspired by airplane companies yield optimization

• Mike

Oops! You’re right I screwed up my math, 3/2 != \$1.25!! I don’t think it detracts from the end point, will fix the #s tonight.

-Mike

• http://www.cogmap.com/blog/ Brent

Err, this is the small hill problem. You are maximizing revenue in the context of this auction, although this auction probably represents a tiny amount of their revenue. If their rate card imploded, that would be worse, probably.

• Mike

RE: Brant — Can you elaborate? If that’s the case shouldnt’ sellers not do remnant at all?

RE: Albert — Great question. I think we’re seeing this today already through RTB but there is definitely a struggle as publishers are trying to figure out how to capture *more* of that value. The challenge for the pub is this — if I’m an advertiser looking specifically for ‘Joe’ then am I willing to pay 10x more to get him on cnn.com versus somesportsblog.com ?

• Dino

Mike,

Great post as usual! I have been following your posts for sometime now and I have to say that I was wondering what on earth is going on-given that the last post was 5 months ago!

Anyways,
Do you think that the only option for a publisher putting a floor price is to offer a ‘house ad’ if there is no demand at a certain price level? How about offering ‘Class1/directly sold guaranteed inventory’ if there is no bid from an auction. Of course that requires that careful attention be paid to forecasting inventory available for direct sale but in my view that gives a publisher the best of both worlds?

• Mike

Hi Dino — In an ideal world you do have a guaranteed ad but the problem is in many cases the whole point you’re auctioning off the impression in the first place is because you don’t have a paying class-1 ad!

Many publishers today either drop a house ad, a PSA or also a google adsense tag as they feel that the text-link format is less competitive than letting a network show a display ad.

• Dino

Ok let me explain.

I work for a large(very large..and you probably know who i am talking about..)network and increasingly we are seeing in our campaign reports that the bids often exceed the price the Class1 inventory is sold for. We could of course start charging higher for our premium inventory but that would probably not work out as only certain specific types of impressions show this pattern and modelling this segment for pricing would confuse buyers at best.

We are therefore contemplating putting ‘ranges’ off guaranteed impressions and pitch our guaranteed inventory against the remnant buys to get incremental revenue lift. While this is certainly possible technically, I am wondering if you can share any light on whether anyone is actually doing that?

(sorry i missed your presentation at the recent EU summit as was hoping to catch up with you and get some more clarity on how you work with Ad Nets within the DR domain)

• http://Forbes.com Achir

What I really would like to have is floor price to be passed on to the exchange as part of the ad tag for each impression. That way we can value the impression and if not sold for the set price (calculated based on our knowledge of the user and the page he is on), we can display our something else or pass it along. Of course this would require a technology on the publisher side to evaluate the impression, much like DSP’s do.
Most publishers don’t have the technology, but few who has the technology could benefit.

• http://www.acceleration.biz Clive

I particular like your paragraph about using house ads to potential drive impressions else where on the publishers site. Generally publisher severely under-estimate the value of the house inventory. I’d suggest that publishers usually have a marketing initiatives that they could promote within their house inventory, even if it is just the promotion of other sections of their site.

By exploiting unsold inventory with house campaigns and using the same due diligence that external advertisers use to place their campaigns, they could understand the real value of the inventory and get much greater returns. Once they know this value, publishers are in a much better position to set floor rates with the remnant suppliers.

• Emily

Non-advertising second price auction question:

EBay runs a second price auction, and so hypothetically when someone is buying something on eBay they should put in the price they are willing to pay and let the auction happen. If they win, they will pay the second price. But in practice, people wait until the last second to enter their bids, even using sniping software to enter incrementally higher bids until they are winning. Why is this?

I could see that people would get carried away with the desire to win, and hence overvaluate at the end when they are losing, but this “sniping” thing seems premeditated and widespread now. Are people trying to avoid giving away info on their valuations, which could cause other people to overevaluate?

• ian

interesting article, but I think it has flawed logic.

their arguments are that floors reduce profitability for the publisher, and shouldn’t be used.

if you look at example tables, the first one makes \$13 for the publisher, the 2nd \$7 (which uses a smarter bidding strategy), and the 3rd (with floors) \$6.50.

the smarter bidding strategy sounds the best, but it doesn’t include the probability that the user won’t get to the 3rd/4th impression (ie the risk associated with the browser not seeing the ad at all), and he might have to end up raising it to get his inventory filled.

the 3rd one sounds bad, but it assumes that the publisher’s house ads can’t make more than 50c over 6 positions (have a lower CPM than 9c), which may not be true.

what it does point to (to me) is that the publisher should be putting his house ads into the RTB as well, which they are doing by putting a floor on there. The floor being the house-ad or best offer the can get not using the RTB.

• http://www.viraladnetwork.net/blog/author/Tim%20Wintle/ Tim Wintle

Emily: the differences between the ebay and ad-server second price auctions are that on ebay the auction happens once (as there’s only one item), and bidders reveal information about themselves by bidding (since others can see their bids).

As a result, there is asymmetric information between the bidders (the person who has the current highest bid has revealed more information than others)

For ad serving you have sealed bids (so nobody knows what the other bidders are bidding). Optimum bidding strategies are different as a consequence.

• http://www.ciblage-comportemental.net/blog/2010/09/25/price-floors-second-price-auctions-and-market-dynamics/ Behavioral Advertising / Publicité Comportementale » Price floors, second price auctions and market dynamics

[...] One of the things that is often discussed but not often written about are the market mechanics that … [...]

• Ilya Kipnis

New reader here (Mr. Scott Menzer advised me to read your blog since I’m interested in working for your firm)…I wonder how fast companies can change the rates on the rate card, because I think with the proper use of statistics and the employment of a binary search or newton-raphson algorithm, can’t (theoretically), a company get to an optimal floor price for which its revenues are maximized?

Furthermore, just from a ten-thousand foot perspective with the second-price RTB model…say bidding just begins and we start at oh, for the sake of simplicity, a penny for a set of impressions that can be appraised for say, \$.30 CPM. If I know this is a second-price model, can’t I make a ridiculous bid such as \$10.00 PM and then lock in some sort of low price? I guess I’m going under the assumption that A) the bidding happens in discrete time B) bid reception is instantaneous and C) all information is public (EG if I bid 30 cents, then all other bidders immediately know I bid 30 cents).

Now I’m probably missing something here. But I’m not sure what it is.

• http://www2.recoset.com/content/2012/02/peeking-into-the-black-box-part-2-algorithm-meets-world/ Peeking Into the Black Box, Part 2: Algorithm Meets World | Recoset Machine Learning and Predictive Analytics

[...] type of auction; it certainly doesn’t raise the bid value to win an unprofitable impression. Mike on Ads has a good post about why it’s a bad idea for publishers to do this, based on some hypothetical behaviour of [...]

• http://www.zip-repair.org/ free zip repair

very useful info, Thanks for sharing….

• http://www.key-logger.ws/ keylogger

Thanks for the great article, found some really good

• sachin ruhela
• http://www.keystrokecapture.ws/ keylogger

Thank your for article, it’s very interesting

• Hardwood Orang

I found Awesome company who serve Flooring
they offer best service and provide all types of flooring http://efloorsusa.com