PPC Case Study #1

  Written by Richard Bravo on Mar 19, 2009 1:03 am

Read More: Marketing  |  Pay per Click  |  Sponsoring and Recruiting

• 1
   VOTES

  Up
• Tweet

Hi everyone,

First... sorry for the delay getting this group started.
Life through me a curve ball recently and, well, you know, you just have to take it as it comes right?

So instead of leaving the discussion stranded I've decided to share some data from a recent short run campaign about a week or so ago; just to get things started at least.

Ready? Here goes then....

~=~=~=~=~=~=~=~=~

PROJECT GOAL:
Make some sales in a short run campaign on a product that had less than a 72 hour window of availability remaining. (new product / limited time launch)

NETWORK:
Google AdWords

CAMPAIGN SETTINGS:
76 Keywords
9 Keyword Themes
76 AdGroups
2 Ad Variations per Keyword Theme
Ads Optimized to Show Better Performer
18 Display URL Variations
9 Landing Pages
Content Network Off
$150.00 Daily Budget
Language: English
Countries: US, Canada, UK, Australia, New Zealand,

~=~=~=~=~=~=~=~=~

OK, so the way I ran this campaign was a combination of things that I picked up from various courses and personal experience.

When I had heard about this product launch it was almost over so I had a very narrow window of opportunity to capitalize on the project.

• I started by immediately running through the squeeze page and sales page to get familiar with the product; a $497.00 home study course.
   
• Next I used the Gogle keyword tool and used the URL search tool to get my initial list of keywords. (choosing to show ALL data columns in the results) From that list I pulled out roughly 7-10 words & phrases that were the most relevant to the course material at a root level. I then took those 10 keywords and put them back into Google using the phrase search to generate an expanded keyword list based on the root list.
   
• I repeated this process, gradually adding keywords. I probably repeated this 10-12 times; occassionally bouncing to the free keyword tool at SEOBook to explore a new root word for traffic potential, adding it to my original root list in Google and generating a new list. When all was said and done I had roughly 76 keywords.
   
• I export that list as a .csv file. I took that data and copy & pasted it into a custom Excel spreadsheet I use to sort my list, add tracking IDs, build my ads, etc. I know there's software out there that may be able to do all of this but I like Excel and I've got my process down pretty good now that I can build a campaign like this in a few hours from start to finish.
   
• Fisrt thing I do is assign a "keyword theme" to every keyword. I start by sorting the list alphabetically and then scan it for patterns in the words. It's usually pretty easy to see patterns in the list once you have it sorted this way. Each theme is based on one dominant phrase or keyword.

  i.e. I might find a cluster of phrases where the keyword "MLM Training" is the the dominant phrase, or "theme" in that cluster. Or I might see a pattern where "Online MLM Training" is the theme. At first it's pretty easy to see the patterns in the majority of the list, but the lines get blurry as you start to get down to just a few abstract remaining words in the list; I just give them the closest theme I can from something I already used that makes the most sense. Other times I just might make a theme that only has 1-2 words in it.
   

• So in this case I wound up with 9 themes that I used to organize my keyword list into "relevant" or similar clusters. I then wrote 2 ad variation for each theme cluster using the dominant keyword in the headline and ad copy; using the residual words from the cluster in the ad copy as well. Each ad variation had a different Display URL, which was a variation of the theme keyword. In all I had 18 display URLs; 2 for each of the 9 themes.
   
• I then created 9 theme optimized landing pages using a simple iFrame redirect to cloak my affiliate link. I used about 350-400 words of keyword focused content in the optimized frame text, including the title tag, description and body text. I did not use the keywords tag. All 9 pages were uploaded to a single sub directory on one of my domains. This domain is not optimized for the product or keywords.
   
• I put each keyword in a separate AdGroup for this campaign using broad, phrase and exact match in each; so each add group had a total of 3 variations of a single keyword.
   
• For my starting Max CPC bids I simply dumped my list with all 3 match types into the Google traffic estimator tool, left everything blank except the countries, which I used the 5 listed above, and let it rip. I leave the bids and daily budget blank to get the estimated bid for the 1-3 position right off the bat without having to guess. You can quickly create all 3 match variations in your list by going to SEOBook , scrolling down to Bulk Keyword Editing Tools and selecting the Keyword Wrapper tool; follow the directions.
   
• Based on those initial figures I dropped the broad match bids by 20% across the board, and dropped the phrase and exact by 10% across the board. I was aiming for position 3-5 placement for my ad to begin with to see what kind of actual numbers I would get starting out.
   
• I take everything in my spreadsheet and paste it into Google AdWords Editor to upload to my account.
   
• Once everything is uploaded I usually go outside for some fresh air and wait about 20-30 minutes for the campaign to publish completely. After about 45 minutes to an hour I download an update in AdWords Editor choosing to get estimated first page bid requirement and check to see how my initial placement is going.
   
• The first thing I do is sort by quality score in descending order; highest at the top of the list. I go down the list to see which ads have low quality scores then cross reference online in the actual account. This can be a little tedious but I go one by one in each adgroup to see why these have a low quality score. I immediately paused anything that had really low quality scores or excessively high first page bids. This cut my list to about half, which is what I wanted, so I could find out which keywords Google would let me put into play immediately. I needed to know this as fast as possible. Considering the time crunch of the launch I just didn't have time for a bunch of optimization.
   
• All in all I wound up with 8 out of the 76 keywords that generating all of the traffic in the 2.5 days that I ran the campaign. I made some gradual tweaks to the ad copy throughout the life of the campaign. I used AdWords Editor for all of it.
   
• My only real edit was a reference to how much time was left on the products availability in the ad copy. So When it got down to 24 hours before the launch was over I changed the ad copy to reflect "Less Than 24 Hours Left...", then "Less Than 12 Hours Left...", then "Less Than 6 Hours Left...", "Less Than 3 Hours Left...", "Less Than 2 Hours Left..." "Less Than 1 Hours Left..." at which point I stopped making any adjustments and waited for an announcement that the launch was over and then paused the campaign.

~=~=~=~=~=~=~=~=~
Here's the Final Campaign Data:

Campaign Start: 3/6/2009
Campaign End: 3/8/2009

CAMPAIGN STATS:
8 Producing Keywords
Clicks: 1,1,1,9,3,2,12,1
Impr: 2,4,11,116,52,63,393,378
CTR%: 50,25,9.09,7.76,5.77,3.17,3.05,0.26
Avg CPC $: .05,.05,.61,70,.23,1.85,.52,.81
Cost $: .05,.05,.61,6.31,0.69,3.70,6.28,0.81
Avg Pos: 1.5,5.3,3.7,2.6,3.5,7,5.6,2.5

CAMPAIGN TOTALS:
Clicks: 30
Impr: 1294
CTR: 2.32%
Avg CPC: $0.62
Cost: $18.50
Avg Pos: 4.1

PROFIT ANALYSIS:
Total Sales From PPC: 2
Commission Per Sale: $225
Gross Commission: $450
Cost Of PPC Campaign: $18.50
Net Profits: $431.50
Conversion Rate: 7%
Cost Per Conversion: $9.25
ROI: 2440%

~=~=~=~=~=~=~=~=~

Let's see...... I think that's it. I also made other sales from this launch using "free" marketing methods; so all in all it was a pretty successful campaign in my book for 2.5 days of work.

It's been a long day and I didn't get to sleep until 5am last night, or this morning, or however you say that, so I hope it made sense to you.

Anyway.... let's open this for questions, comments, feedback, criticism.... whatever's clever.

Cheers!
~Richard

PS: I'm going to try to start doing my case studies on video for the next round.... if the group stays open.

Login or register to post comments  |   Views: 207  |   Comments: 5