16 replies to this topic

[Craig B]

Hello,

I understand that Pay-Per-Click advertising can be tracked from the entry point of a website to a conversion page using a variety of techniques such as:

1. Appending a string at the end of the URL
http://www.somewebsite.com?refer=2344 in the link of the paid listing and then pass that refer along in a GET or a POST until a customer (hopefully) reaches a checkout or something defined as a success.

2. Using a 1x1 pixel gif and appending a tracking to number to it using javascript

What I would like to know is if there is a similar way of tracking a site SEO campaign. For example, could I track that someone clicked in from Google to a website and then at some point made a conversion? (possibly using a refer?)

The reason I ask is because our client would like to track the success of a targeted natural search campaign versus a pay-per-click campaign using conversions as the success criteria.

Any information on this would be most helpful and appreciated.

thanks,
Craig.

Edited by crazed_canuck, 09 October 2003 - 06:18 PM.

[Matt B]

Absolutely you can.

It's very easy, usually done by tagging visitors that reach the destination page and working backward to the referrer. NO invisible gif's required.

Natural SEO had the ability to track conversions and sales long before PPC came into the picture, PPC just brought in the demand for "easier" tools. :learn:

I've also been able to show the difference in conversion rates and customer acquisition costs between natural SEO and PPC marketing. With very surprising results. One customer had a customer acquisition cost from natural SEO of $0.73, while it jumped to $2.30 for PPC. The main difference was the cost per customer. Natural SEO cost per customer is $0.12, compared to over $1.50 PPC.

I'm sure there are many variations and differences by how you run a PPC campaign, but after a number of months the difference was very clear to us and the client. However, I won't promise the same to another client in another industry - it can be very different depending on demand v. competition.

Edited by Matt B, 09 October 2003 - 03:54 PM.

[meta]

Craig,

The best tracking methods involve the use of cookies. There are many methods for this, but the key is that you drop the cookie to the visitor the first time a file is requested from your site, and then each time that an additional file is requested you will be able to identify the visitor's cookie. Within your site, this is no problem, but it is more challenging when you need to incorporate information about something that goes on outside your site, such as entering by clicking on an ad. No site may access any cookie but its own, so to get information involving visits to sites other than your own requires some special effort.

You've mentioned tracking codes, these codes associate an entry with a particular campaign. The use of a tiny image, sometimes called "page dotting" is a way of having a visitor to a third party site download a file from your site, allowing you to give them a cookie. Here, the visitor is normally unaware that a file has been downloaded from your site. Even without a tracking code or a page dot, your logs may contain the referring page, though not all do.

So you have options, but however you go about it, the analysis can be a challenge. Eyeballing logs only gets one so far, so you're going to need software to help if you are serious. There are a number of software packages for log analysis out there. They differ widely in capability and price, so shop carefully, and ask your own questions.

[Vertster]

I just ran a report like this for a client yesterday. I suggest you look into using Clicktracks for this. If you have not heard much about it, take a look at:

http://www.clicktracks.com/

You can read more about it in this thread as well.

http://www.highranki...php?showtopic=5

It is a very powerful tool that will let you easily seperate who came from natural search, and what keywords they used.

The only problem I see is that in many MANY cases, the numbers people are looking at using this software are way too small to be statistically significant and make decisions on.

[meta]

The only problem I see is that in many MANY cases, the numbers people are looking at using this software are way too small to be statistically significant and make decisions on.

Vertster,

Could you elaborate on this a little?

[Craig B]

Thanks for the information. Very informative.

As far as the size of the data set, the website has substantial traffic so there is enough to make decisions on. We also use ClickTracks and other R&A (research and analytics) tools so it isn't the software but more the process side of things that I am investigating.

As far as the other posts, I have some questions:

It's very easy, usually done by tagging visitors that reach the destination page and working backward to the referrer. NO invisible gif's required.

I understand the tagging at the end of the process, but how do you work your way back if haven't passed something in the URL or in a hidden field or in a cookie along the way?

I've also been able to show the difference in conversion rates and customer acquisition costs between natural SEO and PPC marketing. With very surprising results. One customer had a customer acquisition cost from natural SEO of $0.73, while it jumped to $2.30 for PPC. The main difference was the cost per customer. Natural SEO cost per customer is $0.12, compared to over $1.50 PPC.

Awesome! Is there available software that automates these calculations? or how did you figure this out if you did it by hand.

The best tracking methods involve the use of cookies.

I understand the use of cookies to track a user through a site but I don't understand how you can corelate users that came from a search engine to a cookie. ie. can I write to a cookie 'google-user_2342342' on a cookie if they are referred from a search engine?

It is a very powerful tool that will let you easily seperate who came from natural search, and what keywords they used.

Great, but does it tell me whether these clicks from search engines converted into a sale?

Thanks again for all the great posts and information.

-Craig

[Jill]

The latest version of WebTrends also can track conversions of "natural" search visitors.

Jill

[Craig B]

Cool. I think they are using Omniture (http://www.omniture.com/ but I would have to check to make sure. Basically, I am doing some digging around for my own understanding.

[Think Web]

I understand the tagging at the end of the process, but how do you work your way back if haven't passed something in the URL or in a hidden field or in a cookie along the way?

In ClickTracks, you'd tag the visitors that made it to the target page(s) or conversion points. From there you would go to the Search Report section and view the search phrases that were used of those tagged visitors.

ClickTracks does a fine job at matching the passed query strings (keywords/phrases) from the search engines. No cookies or hidden anything are needed.

If you already have ClickTracks, I suggest learning it inside and out. :learn:

[prophecy]

Hi,

Ecommstats can do this for you easily without having to change anything. It can do most things that people are talking about on here actually. I'm a llittle biased, but check it out for yourself.

[Jill]

Welcome, Prophecy!

Jill

[Scottie]

Welcome Prophecy!

I won't do it here (I've sworn off giving people reviews of their site when they didn't ask for them) but once you get to be an active member (10 posts) you may want to post for a website review.

You've got some usability issues that could be easily corrected...

[prophecy]

You've got some usability issues that could be easily corrected...

This would be great to know. 10 posts hey?

[Scottie]

You're halfway there...

[AussieWebmaster]

The tracking is a definite must. Especially for PPC advertising..you have to know what is closing for you and where you are wasting your money. ROI tracking is a necessary element to make the effort worthwhile.
It has also helped me find fraud even the originating engine was not aware of.
I use KeywordMax but there are any number of programs that all seem to work as well as each other.
They are not 100% yet as the third party search result providers cannot always be effectively tracked through the PPC engine where you are paying.
But the numbers are good enough to show you what areas to concentrate on and what ones need work.