Guide for testing

[BonesawMcGraw]

Hi all,

Like most newbs, I thought my ads would rock and started using paid traffic while adjusting my ad on the fly. I realize that a more structured plan should be used going forward.

I'm having a hard time finding any examples of how to best lay out your campaign before testing. In other words, is there a steadfast way to "test" your campaign ads?

What are the variables of the ads I should be testing and is there a good pictoral guide that shows all of the different things we should be testing for? Maybe I'm not explaining it the best but if there's a solid guide out there, I should be able to figure things out.

 

[Graybeard]

You need a flow chart -- make one I guess ...
It's an if then or else conditional pattern by design ...
List your possible attributes.
The ad networks are pretty lame basically the user data is the webserver access log data.
Some ad networks will have GEO city or postal code tracking -- the more sophisticated ones.
Then some claim they have keyword *interest* segmentation
Facebook has what data people volunteer

There really is not a generic standard per se

I have my own databases that are in depth by GEO fencing within US (only) census tracts and larger areas 2KM radius about. It's all aggregate info though. income, housing values and occupancy, crime rates other government published data ... I can segment a lot by the zip code ...

 

[azgold]

What are the variables of the ads I should be testing

I don't know what type of ads you're pushing - display, search, etc. but generally speaking....

- Headlines
- Images
- CTA button text, colour, size, location
- Colours
- Text
- Angle

That's probably enough to keep you busy for a bit.

Start with 3-5 ads per offer, keyword, angle, whatever. Take the best performing ones and make only one change at a time, or you won't know what works or doesn't. Keep optimizing your ads and your targeting. Ideally, you should be seeing your numbers climb once you find what converts best.

 

[awesomesauce]

The only thing that really matters at the end of the day is how much money you make with your ads (ie. conversions). You can have a 100% CTR ad that makes you nothing and you could have a .000001% CTR that makes you thousands of dollars.

Like Azgold mentioned there are lots of things you can test that may or may not have any effect on how much money you make.

 

[ralphmao]

Hi all,

Like most newbs, I thought my ads would rock and started using paid traffic while adjusting my ad on the fly. I realize that a more structured plan should be used going forward.

I'm having a hard time finding any examples of how to best lay out your campaign before testing. In other words, is there a steadfast way to "test" your campaign ads?

What are the variables of the ads I should be testing and is there a good pictoral guide that shows all of the different things we should be testing for? Maybe I'm not explaining it the best but if there's a solid guide out there, I should be able to figure things out.

What type of offers are you running??