Adwords - Split Testing Ads

[Displaycentreuk]

I have read on number of excellent threads on this subject here on the forum.

But what is the best way to do this? Should I keep pausing the 'losing' ad and creating a new version or is it better to just rewrite the 'losing' ad?

I guess the same question applied to Keywords and Ad groups that have been tested and paused. Should I keep or delete.

What is the best of practice out there?

Thanks

Chris

[Scott-CopyandDesign]

Run two ads at the same time and remove the option to give the better converting ad more impressions. That way, both ads will be shown to 50% of impressions each, and you can see how they perform.

[SteveGibson]

I have read on number of excellent threads on this subject here on the forum.

But what is the best way to do this? Should I keep pausing the 'losing' ad and creating a new version or is it better to just rewrite the 'losing' ad?

They should equate to the same thing.

If you edit an ad, google deletes the old ad and creates a new ad.

There may be some glitch in the algo that says a re-written ad will be approved sooner, but if that's the case, I don't know about it.

Steve

[AaronB]

So you would have 2 adds running with exactly the same keywords, just different text?

Then you tweek it a little?


What about the keywords? when you are happy with the text would you create 2 exactly the same text adds, then just spit test with different keywords?


Aaron.

[Displaycentreuk]

Thank you for the responses.

Not sure that I made my question sufficiently clear.

At the moment my AdGroups have long list of Ads (currently paused) that I have tested and were losers. The advantage is that I can see what I have tried in the past but the disadvantage is that I have to scroll through long lists of paused ads to get to the active ones.

So, I can pause a losing ad and create a new one. Or I can edit the losing ad and so my list contains only active ads.

Which is best?

Thanks

Chris

[SteveGibson]

At the moment my AdGroups have long list of Ads (currently paused) that I have tested and were losers. The advantage is that I can see what I have tried in the past but the disadvantage is that I have to scroll through long lists of paused ads to get to the active ones.

You can delete them. Then, when you want to see your ad history, you just set the display to show deleted ads. When you want to hide them, set it to hide deleted ads.

The menu for this is on the right hand side (just above the ads). Let me know if you can't find it.

Steve

[Displaycentreuk]

Thanks for your help Steve

[SteveGibson]

Thanks for your help Steve

You're welcome. Glad I could help.

Cheers,

Steve

[admagic]

They should equate to the same thing.

If you edit an ad, google deletes the old ad and creates a new ad.

There may be some glitch in the algo that says a re-written ad will be approved sooner, but if that's the case, I don't know about it.

Steve

Steve - I can confirm without doubt - that EDITING ads is a dangerous game, the approval delay can be a serious issue, and can take ads off line for hours even days, worse still there is no rhyme or reason.. My experience is always create a new one to run along side which shows immediately, pause the older one. I suspect you like - me - create new not tinker, not least to keep a record of what you tested.

[zigojacko]

Use Adwords Editor for managing your campaigns, it's much easier to filter through your ads whether they be paused, deleted or active. :)

[SteveGibson]

Steve - I can confirm without doubt - that EDITING ads is a dangerous game, the approval delay can be a serious issue, and can take ads off line for hours even days, worse still there is no rhyme or reason.. My experience is always create a new one to run along side which shows immediately, pause the older one. I suspect you like - me - create new not tinker, not least to keep a record of what you tested.

I'll almost always split-test and, when I do, I create new ads rather than edit old ones.

(i.e. I delete the loser, then create a new test ad)

However, you still get the delay. (can be minutes or, in some markets, as long as a week)

Steve