Kevin Lee spoke first about Paid Search Performance Evaluation and Tracking. Kevin had his first child, a girl, a couple of weeks ago.

Ofen the data being used to make bidding decision is based data and if you have bad data you will make the wrong bidding decision.  If your data is not accurate you will make fundamental wrong business and financial decision.  Your foundation should be good - and people assume it is, but often it’s not and if you go back and look at your data you’ll see a lot of missed activity.

Tracking based on just surface information

• revenue
• new vs old customer
• lag time (when they visit vs when they convert)
• referrer
• traffic type (search/contextual)

Even if you are managing to a CPA or CPO, additional data is valuabe (above).  It depends on how your paid - sometimes you don’t need the extra data.

The world does not exist purely on the web and Google/ComScore studies show the importance of offline conversions.

25% of searchers purchased an item directly related to their query while 37% completed the purchase online while 63% completed a purchase offline following their search activity

What is you net gain from Paid?

Is there cannibalization away from organic links when you run paid?

Which pages convert better? (maybe you don’t want the searchers clicking on organic links at all)

Are you tracking the best mix of creative when you take the extra real estate?

The page you design to convert a visitor might not be the page you’d create to do well in search engines - because you have considerations for Search that might interfear with the best possible conversion page.

Tracking must include more than just the last touchpoint - most analytics don’t look at the touchpoints all along the line (the funnel).  You should attribute value along the funnel.   The last thing we did might not be the only thing of value (buy something or sign up something).  Search Marketing is thinking about the process of what people went though while they are buying something and attributing the source of it.  In order to get the right level of weight to all the touchpoints.

Cookie Issues:

Do you know the impact of cookies on your decisions?

what is the right cookie length?

Does your audience purge cookies?

(if you have a long buying cycle - there are more of those you will not be able to track).

Referrer data can be important for several reasons

• engine
• actual keyword string
• splitting out AdGroups
• Monitoring for Fraud
• understand conversion by engine with may influence position decisions

Even if your referrals are missing a referrer - that can be sign of fraud.  There are many things that you can look closely at for fraud.  Didit.com puts up a conversion index every month (i’ve seen it) and we look at how each engine is converting (this is an audience issue). For example, did you know that this month AOL is the second highest converting paid search engine - you can take action on that data.

Good Tracking is a Foundation

You may have data overload but make sure you collect the data

Marketers must us the technology at their disposal

You need the data to win.