PDA

View Full Version : AB testing of creatives


AussieWebmaster
12-14-2004, 12:15 AM
I am trying to see if the numbers that are always quoted of such a small number of advertisers using any tracking beyond the landing page are true.... and if people are using the benefits of Google's ability to AB test your creatives.

AB testing of creatives is when you use for example two different ads that are each tracked separately to compare CTR and conversions. Some ads will do great with CTR (I have a bunch in solid double digits) but they do not convert that well. Some ads that convert have a hard time getting any traffic.

Obviously the trick is to find the happy medium that gives you a solid traffic flow that converts at a good number. Why stick to an ad that converts 1 in 3 but only gets you 3 sales when another that converts at 1 in 6 can get you 25 sales. The ability to know that you can pay $10 for a customer and make $40 profit but only make $120 a day while you may pay $20 a customer and make $30 per sale but make $750 a day is a big part of successful AdWords marketing.

I would like to know if others are doing similiar testing or if there are people who would like more details on how to achieve this information.

If you need help all you have to do is ask.

AdWordsRep
12-14-2004, 05:04 PM
If you need help all you have to do is ask.
Superb topic, AussieWebmaster - and I'm hoping that somebody will 'just ask' soon. :)

If not, I will.

AWR ;)

KellySEM
12-16-2004, 12:43 PM
I've been doing conversion testing with Google's A/B serving for 2+ years now. I just add a tracking URL to each adgroup ad (as opposed to running each keyword with a tracking URL) and see which one is most effective from a BTC (book-to-click) perspective. You definitely need to strike a happy medium between strong CTR and good BTC, as both are important to running a successful AdWords campaign. The returns tend to diminish over time, but the campaign's initial performance gains in this area were huge.

I've been complaining to Overture for a couple of years that they need to get A/B testing capabilities and they always claim its just around the corner...I've stopped holding my breath on that...

AussieWebmaster
12-22-2004, 09:08 PM
The diminishing returns are from content in many cases... the regulars see the same ads a lot so to be successful you need to constantly change them.
On the straight search side I have found you can maintain a level within the usual ups and downs caused by events etc.

The Generator
12-24-2004, 01:28 PM
One of the most valuable techniques in Adwords, A/B testing applied consistantly has a good and long term impact on CPC, CTR, and acquisitions.

I've been complaining to Overture for a couple of years that they need to get A/B testing capabilities and they always claim its just around the corner...I've stopped holding my breath on that...

Not to change this Adwords subject, but I know what you mean about Overture. I suppose the only way to AB test on that engine now is to use very similar keywords that get similar traffic, such as "catalytic converters ford cars" and "catalytic converters ford automobiles."

AussieWebmaster
12-25-2004, 11:22 AM
Overture always seems to be getting to things next quarter... if MSN rolls out of beta "next quarter" and decides to do its own ad sales we may see a more pliable and responsive Overture.

Mel66
12-27-2004, 11:47 AM
We've done both A/B creative testing (different ad copy going to same landing page) and landing page testing (same ad copy going to different landing pages) and have had great luck with it. It's amazing how much difference a minute change in wording or page design can make in your conversion. We use different tracking URLs to track results. The ability to test is the single biggest advantage of Adwords over Overture, IMHO. We often test concepts on Adwords and roll them out to Overture.

That said, we've often found that the best-converting ad copy draws the worst CTR. It's frustrating trying to find that happy medium.

Melissa

AussieWebmaster
12-27-2004, 12:19 PM
We've done both A/B creative testing (different ad copy going to same landing page) and landing page testing (same ad copy going to different landing pages) and have had great luck with it. It's amazing how much difference a minute change in wording or page design can make in your conversion. We use different tracking URLs to track results. The ability to test is the single biggest advantage of Adwords over Overture, IMHO. We often test concepts on Adwords and roll them out to Overture.

That said, we've often found that the best-converting ad copy draws the worst CTR. It's frustrating trying to find that happy medium.

Melissa
But that is the Holy Grail of the industry....

The Generator
12-29-2004, 06:22 PM
Indeed, sometimes, but not always, it's necessary to forfeit / delete the creative that has the higher conversion rate. I've found that using a creative with a higher CTR but lower conversion rate sometimes gets a higher net number of acquisitions. Yes, it's a real catch 22.