Banner Ad: Please Fix Your Pacing Algorithm
Decorative Header Image

Tag Archive for campaign optimization

Alternative to the CPM

Online advertising transactions are all CPM based. You might think my wild assertion is out of line. You might think you’re buying ads on a CPC or a CPA basis. But when a publisher is looking to sell ad inventory, they’re thinking about the CPM. “How many dollars can I get for every thousand ad views?” And when that CPA deal or that CPC deal comes in the publisher’s doing the math to convert that number into a CPM.

Blowing up the CPMFor a CPA deal they’re estimating how many acquisitions they can send to the buyer for every thousand ad views. For CPC, how many clicks per thousand ad views. They’re boiling it down to a CPM because that’s how they can compare the deals. It works like this all the way up and down the funnel.

The CPM has been around for a long time. With the advent of the RTB auction model, the CPM is very dynamic. Each impression up for auction is individually valued based on countless bits of information about the user, the page, the size, the date, the historical performance and a variety of other variables. Even though a separate auction is run for each impression, the bid prices are still in the form of a CPM. It’s in our blood. It is the end result of normalizing the value of an ad impression so that it can be compared to its peers.

Some Problems

I want to point out a couple of problems with the CPM. First and foremost, it’s a single number. This aspect causes a couple of secondary problems that put the buyers at risk. One of the big ones is that there’s no guarantee that the ad will actually show on the page. Steps have been taken to address this by several companies. The result of this problem is a topic near and dear to my heart: discrepancy. Read more

One bid per DSP per impression – why?

Why historically (and currently) only one single bid was allowed for each DSP per impression? Why hide demand from the exchange and create opportunities for the DSPs to arbitrage? – I know this is changing now with the possibility of multiple bids per DSP (openRTB v2) but why ad exchanges let this happen at the beginning?
This question was asked on quora, below is my answer.

One bid per impression, why?Short Answer

A multiple bid response was discussed at the very first OpenRTB meeting.  It was not seen as a favorable feature by the demand side, at first.  They preferred submitting one bid.  Supply side partners were not in a position to force the issue, nor had the necessary research been done to support the idea.

Early Days

From the supply side’s perspective, as with many transaction systems, early efforts in RTB were focused on connecting the pipes.  RTB represented a new source of demand and the pressure was applied to getting plugged in to as many DSPs as possible. Read more

What are the most important KPI’s to monitor when launching a proprietary demand side bidder?

What technical, operational and campaign performance [bidder] KPI’s should be considered when ramping up, and are there any industry benchmarks?
This question was asked on quora, below is my answer.

Some bidder KPIs can be monitored with tools like graphite (not angry birds).From a technical perspective you’ll want to measure how many different types of inventory you support: mobile, app, web, video, facebook ads etc… You’ll want to track how many SSPs you’re integrated with and how many impressions are available to you. You should look into creating a feature matrix and decide which advertising features you and your customers find most important.

On the operations side you’ll want to make sure your bidding system is responding to bid requests quickly. The round-trip time for a bid response, from an SSPs perspective, should be no more than 100ms – and even that is pushing it these days. Your internal bidding algorithm should probably make a decision in less than 30 or 40ms. This allows about 60ms for network latency between the bidder and the SSP. Some SSPs have DSP latency monitoring available. This type of monitoring will give you insight into what the SSP is seeing. Read more

Do DSPs provide advertisers with more data…

… than when purchasing through an ad exchange or even using an ad server? I am tasked with helping to expand my company’s online marketing, the options we are looking at are essentially 1) using an ad server to manage ‘private’ media buys 2) using an ad exchange like OpenX Market or 3) using a DSP. One major factor in this decision is the amount of data we can collect in order to optimise media buys.
This question was asked on Quora, below is my answer.

In the current marketplace a DSP is going to be able to give you more insight into your buys across many exchanges and SSPs. They are built from the ground up to cater to the needs of buyers.

Data

You’ll want to find a DSP that can work with many SSP and exchange’s private marketplace technology stacks. You may even find yourself using an SSP’s interface to place orders with particular publishers. Those deals are likely to be executed through the SSP/DSP technologies, so you’ll still need your DSP to act as the buying agent once the deal is done.

Since data collection is one of your primary needs, you might consider using a 3rd party ad server if the DSP you’re working with doesn’t have one that suits you. You can use an ad server as a service or set aside some hardware and install one on your own.

Why does the average RTB win price jump up significantly at midnight EST?

…and do others see this jump at midnight in their own timezone?
This question was asked on Quora.com, below is my answer.

I dug into this win price problem several months ago after noticing the same jump in spend at that hour. Rubicon is on Pacific time so we refer to this as the “9 O’Clock Bump” effect.

Richter's Dodo didn't fix his algorithm, it's causing a win price jump

Dr. Richter pointing at a Dodo bird. “Adapt or perish”

After asking several DSPs about the problem we determined that it was, indeed, campaign budgets resetting combined with less-than-optimal pacing algorithms and in some cases lack thereof.

We’re in the process of finishing up some documentation on our pacing algorithm that does a pretty good job pacing to the needs of the campaign while considering the fairly predictable traffic pattern throughout the day. We’ll be putting this information out in the next couple weeks. Hopefully it will inspire some folks in the market to upgrade their systems and resolve some of this win price inefficiency. I’ll update post with a link to the document once we release it.

UPDATE: The document is finally out the door. You can read it here.