How to best do split testing of ads in Google Adwords

Hi experts, just googling around and watching some youtube videos on this, I'm getting some conflicting theories - some say have only TWO ads in your campaign, run those for awhile, drop the loser and add in a whole new ad, then repeat. While another camp says it's okay to have 5, 6, or more ads at a time, run all those, keep the best one. Add 5 or 6 more new ones, repeat that. Ultimately pick the best one out of all that, then optimize that one ad.   I'm pretty much a newbie at this and it's making my head spin...
   My product is a new software program that It'll be selling from my website (I'm a programmer) and I'd like to start promoting it with Adwords and would like to find the nest way to do "split testing" of my Adwords ads to find the best one. Can anyone offer me some help please?

Thanks!
    Shawn
shawn857Asked:
Who is Participating?

[Product update] Infrastructure Analysis Tool is now available with Business Accounts.Learn More

x
I wear a lot of hats...

"The solutions and answers provided on Experts Exchange have been extremely helpful to me over the last few years. I wear a lot of hats - Developer, Database Administrator, Help Desk, etc., so I know a lot of things but not a lot about one thing. Experts Exchange gives me answers from people who do know a lot about one thing, in a easy to use platform." -Todd S.

Brian ClausenCommented:
Hi Shawn,

Your approach will be impacted by the amount of traffic that you have, as you want to get to a statistically significant data finding before drawing a conclusion.

If you have a small amount of traffic, the approach of only testing two variables will allow you to more rapidly determine which of the two approaches is optimal.  

If you have a large amount of traffic, you can test more variables and potentially move through your decision tree more rapidly. If you test many options, make sure that you identify the differences between each approach so that you can isolate the factors that are leading to improved performance.

In either approach, it is critical that you identify the key differences between each ad so that you understand the factors that you are testing.

Thanks,

Brian
Lucas BishopClick TrackerCommented:
Shawn,

There's a slew of data points you can interpret through Adwords tests/experiments. I recommend you have your Google Analytics (GA) account connected to your Adwords account so you'll have access to on-page behavioral metrics. You can add these columns into most reports inside of Adwords as seen here:
google analytics in adwords
Since your question is centered on "ads" I'll break down some common areas of focus that you can test with an ad:

Click Through Rate (CTR%) - This hinges on the ad copy which consists of your headline, desc line 1, desc line 2 and display url. Since there are 4 components that play into this, You may only want to test one component at a time (ex. headline-a vs. headline-b.), as mentioned above by Brian.

GA Metrics (time on page, bounce rate, etc.) - While these numbers are heavily influenced by the landing page, the ad-copy is tied into the landing page performance. For example, if someone clicks on an ad for a "red ballpoint pen" but the landing page is just a generic "pens" category, you'll likely increase your bounce rate. The ad copy should sync up with the landing page effectively. Hence, if you're testing multiple "Headlines" in your ad copy, you may want to test landing pages that match up with your headline.

Conversion rate - Depending on what you measure as a conversion (sale, download, email list subscription, etc.) the ad copy can heavily influence your success.  For example, you may find the price conscious person downloads and signs up more often than the person concerned with your product features (ie. "30 day free trial" vs. "make your computer faster in 1 minute").

Being that you're new to the system, I'd suggest starting with small simple tests. As you become more comfortable with the system, then you can ramp up and have a variety of tests running simultaneously. With the former, you're less likely to spoil your test results accidentally.
shawn857Author Commented:
Thanks for the feedback guys. I really don't know how much traffic this will pull in... but based on my advertising budget at this juncture (ie. not a lot), it probably won't be terribly huge. So I guess it won't be enough clicks to merit testing 5 or 6 ads simultaneously... I'd better stick to two or so. But here's another question: should my 2 ads be *completely* different - analyze the metrics, drop the weak one, and make a new 2nd one? Or should the 2 ads be basically identical with only one thing different ...ie. just change one thing at a thing, compare, pick the best, then test another thing ?

Thanks
    Shawn
Big Business Goals? Which KPIs Will Help You

The most successful MSPs rely on metrics – known as key performance indicators (KPIs) – for making informed decisions that help their businesses thrive, rather than just survive. This eBook provides an overview of the most important KPIs used by top MSPs.

Brian ClausenCommented:
Hi Shawn,

Your second question could be answered either way, depending on where you are in your program.

If you feel like you are at square one and don't know a lot about what will work, I would test two (or maybe even 3) totally different approaches and then begin optimizing around the one that works best.

If you believe you know the general direction you should take, then I would start with a single approach and then identify 1 to 3 factors that can be different between the two ads that you test. From there, you can narrow down the proper approach for each of these factors.

Given that the program is brand new, testing out two to three totally different directions and then optimizing from there sounds like it may be the appropriate approach.

Thanks

Brian

Experts Exchange Solution brought to you by

Your issues matter to us.

Facing a tech roadblock? Get the help and guidance you need from experienced professionals who care. Ask your question anytime, anywhere, with no hassle.

Start your 7-day free trial
Lucas BishopClick TrackerCommented:
I would follow Brian's recommendation above. I'd start with two vastly different directions so you can narrow down what resonates with your audience faster.

To help get a head start, I would analyze competitors Adwords programs. Hopefully they've already done some work in optimizing their ads/pages/offers, so you can leverage what they are doing and improve on it out of the gate.
shawn857Author Commented:
Good advice guys, thank you. I've made 2 pretty different ads and I'll run those simultaneously and evaluate. But after one emerges as a winner, should I immediately work on just optimizing that particular ad... or come up with a whole new different ad to try against the winner? If the latter, then how many new ads do I keep creating/comparing before I finally stop and just concentrate on optimizing the winner?

Thanks!
    Shawn

P.S: Lucas, I successfully linked my Google Analytics account to my Adwords account, yet the entry for "Google Analytics" does not appear for me, as shown in your "ga-columns.png" screenshot. Would you have any idea why? Thanks.
Brian ClausenCommented:
Hi Shawn,

I would probably start by mapping out a decision tree on a piece of paper that represents your hypotheses about what may or may not work, and about what factors you want to test.

Instead of doing random testing on different approaches, you want to develop thoughts on what might work, and then use your tests to prove/disprove your theory.

Then, you should move systematically down your decision tree based on the results of your test (although the decision tree should be dynamic, where you introduce new hypotheses as you make progress and gather data).

So, I would start by challenging yourself about what the two options you are starting with are really testing. After your first test, you can then optimize around the winner or test against a third alternative, depending on what hypothesis you want to test next.

Thanks

Brian
Lucas BishopClick TrackerCommented:
Lucas, I successfully linked my Google Analytics account to my Adwords account, yet the entry for "Google Analytics" does not appear for me, as shown in your "ga-columns.png" screenshot. Would you have any idea why?

You'll need to click on "Columns > Modify Columns" (from within campaigns/keywords/ads/etc tabs):
adwords modify columns
This is where you'll get to the screen I posted previously.
shawn857Author Commented:
OK thank you Brian.

Yes Lucas, that is what I had done.. but it's not there though - please see my attached screenshot.

Thanks
    Shawn
Adwords.JPG
Lucas BishopClick TrackerCommented:
There are two places to link Adwords/Analytics together. I suspect you've updated the settings inside of Analytics so that it can pull your Adwords data in. However, you also need to update your Adwords account to pull in Analytics data. You can do this via:

"Gear Icon" (Top Right in Adwords)
Linked Accounts
Google Analytics
shawn857Author Commented:
Ah that was it Lucas, got it now!

Thank you both guys... I'm on the right track now!

Cheers
   Shawn
It's more than this solution.Get answers and train to solve all your tech problems - anytime, anywhere.Try it for free Edge Out The Competitionfor your dream job with proven skills and certifications.Get started today Stand Outas the employee with proven skills.Start learning today for free Move Your Career Forwardwith certification training in the latest technologies.Start your trial today
Internet Marketing

From novice to tech pro — start learning today.