DoubleClick Search

Floodlight is an optional feature in DoubleClick for Advertisers (DFA) that allows advertisers to capture and report on the actions of users who visit their website after viewing or clicking one of the advertiser's ads. For the implementation of Floodlight, Floodlight tags must be placed on the advertiser's webpages.

Floodlight tags enable an advertiser to track, measure, and report on post-click activity -- particularly conversions. The tag is a line of HTML code inserted on the web page where the activity takes place, typically, on an order confirmation page. For example:

IMG SRC="http://ad.doubleclick.net/activity;src=805681; type=travel;cat=hotel;tran=89888888; WIDTH=1 HEIGHT=1 BORDER=0

Note the name value pairs type=travel and cat=hotel. These variables allow you to classify the conversions that occur on your site. With these variables, you can create as many distinct Floodlight tags as you wish. This allows DoubleClick Search to track conversions resulting from search engines to a very granular level. For example, if the confirmation page for a product purchase contains a particular Floodlight tag, DoubleClick Search will know which product or service the customer purchased as a result of a search.

The Problem

As an advertiser, how do I optimize a keyword when unintended conversions occur? How do I accurately account for these conversions?

For example, an advertiser that runs a travel web site offers services for Car Rentals, Hotel Rooms, Airfare Tickets, and Cruise Packages. The advertiser may assume that keywords related to Car Rentals would always result in Car Rental conversions. However, consider the following:

1. Customer searches on the keyword "Car Rental Dallas" in the search engine.
2. Customer clicks on the ad that appears as a result of the search.
3. The landing page for the ad appears. This is the main Car Rental page.
4. The customer reviews this page and is intrigued, but the Cruise Package link at the top of the page catches her eye. Without booking a car rental, the customer clicks the Cruise Package link.
5. The main Cruise Package page appears. The customer navigates to a few more pages and finally makes a purchase - a cruise package. The Cruise Package Confirmation Page appears.

Even though the advertiser intended for the "Car Rental Dallas" keyword to result in only Car Rental conversions, it can actually result in Cruise Package conversions. In fact, Car Rental-related keywords could result in conversions to any of the services that the advertiser offers. The following table lists an example of what the conversions could look like for a two-week period:

Table. Conversions for Car Rental Keywords, January 1 - 15

Without Floodlight tag bidding, DoubleClick Search will not account for the varying profit margins of these distinct conversions. It treats all profit margins equally.

The Solution

With Floodlight tag bidding, DoubleClick Search now accounts for the various profit margins of the different types of conversions that a set of keywords produces.

You can use Floodlight tag bidding to create a Cost Per Acquisition (CPA) target for selected Floodlight tags. Floodlight tag bidding weights the actions according to their CPA values to determine the true CPA of the keyword. This allows the keyword to be evaluated properly on performance. Is it performing accurately against the CPA goal, is it overperforming, or is it underperforming? Once this has been determined, the DoubleClick Search Bid Management system properly adjusts the keyword bidding to achieve its overall goal. The following example illustrates the benefits of using Floodlight tag bidding for CPA targets.

Determining the Cost Per Action for Each Booking

An advertiser that runs a travel web site offers a wide range of services with varied prices. The services include Car Rental, Hotel, Airfare, and Cruise Packages. A Car Rental booking has a $50 average price. On the other hand, the Cruise Package booking has a $5,000 average price. Each booking type has its own confirmation page that contains a unique Floodlight tag.

Because of the wide variety of average prices, there is a different profit margin for each type of booking. For example, the advertiser will make an average profit of $20 per car rental, but they will make an average profit of $1,100 per cruise package booking. The advertiser allows for a certain amount of the direct profit from each purchase to be used for search marketing. For example, the advertiser will put $9 per car rental towards Google or Yahoo!. $9 is the Cost Per Action (CPA) for Car Rental bookings. The following table shows the average price, average profit margin, and CPA for each type of booking:

Table. Cost, Profit, and CPA

Conversions Per Type of Action

As the advertiser evaluates its search program, they start to notice that keywords targeting Car Rentals end up converting on more than just car rentals - that is, the car rental keywords result in a few hotel bookings, a few airline bookings, and a few cruise package bookings. The following table shows the conversions that each set of keywords produces, and the related total profit:

Table. Conversions and Profit

If we look at the Car Rental keywords, only $600 of the $6,420 total profit for these keywords can be attributed to car rentals. So we can see that the keywords for a particular type of offering often result in other types of conversions.

Setting Up a Bid Strategy

While setting up their search programs (without using Floodlight tag bidding), the advertiser determines that they will target a CPA goal of $200. This is the approximate average of the allowable advertising spend per booking. They hope that this will allow them to roughly hit the advertising budget for each type of booking.

However, as the program continues to run, the variability of the product profit margins and the unreliability of using keywords to determine the CPA for a type of booking prevent the advertiser from hitting the individual goals for each booking. They decide to forego the flat CPA goal and use Floodlight tag bidding to create a multi-dimensional goal that will reflect the desired CPA for each booking:

• Advertiser sets a CPA goal of $9 for instances when Car Rental keywords drive a Car Rental conversion.
• Advertiser sets a CPA goal of $20 for instances when Car Rental keywords drive a Hotel Room conversion.
• Advertiser sets a CPA goal of $30 for instances when Car Rental keywords drive an Airfare conversion.
• Advertiser sets a CPA goal of $700 for instances when Car Rental keywords drive a Cruise Package conversion.

After the advertiser sets these goals, DoubleClick Search applies the appropriate goals to the corresponding incoming conversions:

• Every time a Car Rental keyword drives a Car Rental conversion, DoubleClick Search evaluates the Car Rental against the $9 CPA goal.
• Every time a Car Rental keyword drives a Hotel conversion, DoubleClick Search evaluates the Hotel booking against the $20 CPA goal.
• Every time a Car Rental keyword drives an Air reservation conversion, DoubleClick Search evaluates the Air reservation against the $30 CPA goal.
• Every time a Car Rental keyword drives a Cruise Package conversion, DoubleClick Search evaluates the Cruise Package sale against the $700 CPA goal.

You can also use Floodlight tag bidding to create an Effective Revenue Share (ERS) ROI target for selected Floodlight tags. Floodlight tag bidding weights the actions according to their ERS values to determine the actual ERS of the keyword. This allows the keyword to be evaluated properly on performance. Once this has been determined, the DoubleClick Search Bid Management system properly adjusts the keyword bidding to achieve its overall goal. The following example illustrates the benefits of using Floodlight tag bidding for ERS targets.

Determining the Effective Revenue Share for Each Product Category

A department store runs a web site that offers a wide range of products with varied prices. Their product categories include Apparel, Hardware, and Electronics. The average Apparel purchase is $50, while the average Electronics purchase is $500.

Because of the wide variety of average prices, there is a different profit margin for each product category. For example, the store will make an average profit of $20 per Apparel purchase (or 40% of the price), but they will make an average profit of $100 per Electronics purchase (or 20% of the price). The store allows for a certain percentage of the average price to be used for search marketing. For example, the store will put 24% per Apparel purchase towards search marketing, because they know they make a profit of 40% per Apparel purchase. 24% is the Effective Revenue Share (ERS) for Apparel. The following table shows the average price, average profit margin, and ERS for each product category:

Table. Cost, Profit, and ERS

Setting Up a Bid Strategy

While setting up their search programs (without using Floodlight tag bidding), the store determines that they will target an ERS goal of 20%. This is the average of the allowable advertising spend per category. They hope that this will allow them to roughly hit the advertising budget for each category.

However, as the program continues to run, the variability of the product profit margins, and the unreliability of using keywords to determine the ERS for a category, are preventing the advertiser from hitting the individual goals for each category. They decide to forego the flat ERS goal and use Floodlight tag bidding to create a multi-dimensional goal that will reflect the desired ERS for each category:

• Advertiser sets an ERS goal of 24% for instances when Apparel keywords drive an Apparel purchase.
• Advertiser sets an ERS goal of 22% for instances when Apparel keywords drive a Hardware purchase.
• Advertiser sets an ERS goal of 14% for instances when Apparel keywords drive an Electronics purchase.

After the advertiser sets these goals, DoubleClick Search applies the appropriate goals to the corresponding conversions:

• Every time an Apparel keyword drives an Apparel sale, DoubleClick Search evaluates the sale against the 24% ERS goal.
• Every time an Apparel keyword drives a Hardware sale, DoubleClick Search evaluates the sale against the 22% ERS goal.
• Every time an Apparel keyword drives an Electronics sale, DoubleClick Search evaluates the sale against the 14% ERS goal.