A New PPV Keyword/URL Testing Strategy for Low Budget Testing

Hello,

This is my first post here. I enjoy statistics and I've found interest in marketing thanks to your website Cheers.

This is a strategy I'm testing currently for finding high performing keywords and URLs on the LeadImpact network on a low budget. I'm in school currently and don't have a lot of money to invest so this is how I'm using my schooling to squeeze keyword information cheaply.

CK

My Strategy for URL/Keyword Testing on Low Testing Budgets on PPV - LeadImpact

I've been widdling away at 248 URLs and keywords split into three groups run at 4 day intervals for about two weeks now and it's come down to 10 URLs, no keywords.

I use this strategy in order to squeeze each keyword for information with the least amount of spend. Some URLs and keywords will hit you hard and won't convert at all, monitor your reports each night. You never know until you test them I've learned, you can loose lots of money quick. Some keywords or URLs will have a lot of conversions: but, a lot of impressions driving up your PPC cost of your CPV.

I kept track of each 4 day interval group in a seperate excel spread sheet exported from LeadImpacts reporting system. I then aggregated the data sets and pulled them into Microsoft Access so I could sort each row by the lowest PPC and the highest impressions and conversions: you can sort data rows by running queries using boonlean it's a database program, just import your aggregate excel sheet.

I then save my filted keyword list I found using boonlean into a new table and export this table to Excel again by exporting the table from Access and then I brake down the same footer figures as in the attached image.

Why do I use Trend and Deviations? You'll understand your parameters from the population at a glance, read the footer figure rows left to right.

Keep in mind my population does not include keywords that may have had high spend but no conversions, you cannot compute Standard Deviation and Trend with empty cells, they are the only outliers removed.*

Definitions & Other Data Collection Information

Standard Deviation - a way of measuring the center of your data or representing the most percise average.


Other Figures

• Their was a total of $98.00 spent on this testing for my niche.
• We had some super high traffic and conversion keywords that destroyed our EPC to PPC/CPV ratio.
• Their were about 5,123 impressions.
• I rotated 10 different offers.
• I tested four different varitations of a landing page, the one I thought sucked destroyed the others after tracking conversions with LinkTrackr.
• Low Impression high Conversion ratio keywords are kept because some keywords performed horribly with high impressions and no conversions.

*I did not remove outliers and have not discounted any extreme changes from any data sets unless the conversions row read 0. All final data in the attached image are parameters of the population of our keyword groups that were rotated.



I'm testing out new offers now from multiple networks now that my landing page and URLs are in-line for my niche.

I'd love to hear input on others testing strategies.