10 Common Adwords Mistakes - Part 2

This is the second part of 10 Common Adwords Mistakes.

See Part 1
6. Not split testing ads

One of the brilliant things about Adwords is our ability to get hard data on traffic from different keywords, including click-thru-rates and customer conversions. So why wouldn’t we try different ads on the same keywords to see which ones give the best results?

A lot of people don’t do this, so I guess they simply don’t understand the importance of the ad copy in achieving click-thrus and conversions. You could write a book on designing and writing good ads, but the very first and most important step is to have two ads in each ad group and to compare their relative performance. These differences could be minor differences of spelling, punctuation or capitalization. Or they could be major differences in the actual offer being made.

Two sets of ads with minor differences that could produce significantly different conversion rates.

Two sets of ads - but with minor differences that could produce significantly different conversion rates.

So what’s split-testing?

To split-test two ads we simply let them run in competition with each other. Before we can do this we must set the campaign Ad serving setting to Rotate in Advanced Options. The default setting is Optimise. What this means is that Google will show the ads with the higher click thru rates more often. This might seem like common sense, but if you are trying to work out which ad is the best, it is quicker to run them both in rotation for a while until we get a winner. Then we drop the poorer performing ad, and create another one in competition with it. Possibly a clone, but with a minor modification. Just like natural selection.

How do we know which is the better ad?

We need to get enough clicks on each ad so that we can apply a test to see whether the differences in clicks are statistically significant. Sounds technical but it’s actually very easy.

All we need is the CTR (click-thru-rate) and number of clicks on two ads to compare them.

Take this data and use splittester.com to see if one ad is really better than the other.

7. Not having an appropriate landing page

It’s one thing getting traffic to our site, but another to get good conversions. These conversions could be actual sale, or sales leads though to our sales department, or sign-ups to a mailing list. Whatever they are, we will only get enough of them if our website is doing it’s job.

One common mistake is for people to dump all their traffic at their home page. This is equivalent to saying to the user “Here you are – go search.” But they thought they were doing at when they put their keywords into the search engine. Now they are into gratification, and you’d better provide it fast, or they’ll press the back button on the browser and you’ll lose them.


Ads should link to landing pages that closely match the ad content.

If the website has only one or two product or service offerings, then landing pages should be set up for each one. The landing pages should reflect as closely as possible the implicit offer in the ad.

It is also desirable to test different landing pages. This can be simple experimentation with placement of product images or it could be more complex multivariate testing. Google supports this with its Website Optimiser tool.

The big message here is that traffic is useless if your site won’t convert. When it comes to improving advertising ROI, start with the website. Get it right, and then start buying more traffic.

8. Not optimising bid prices.

Number one is the best position for my ad isn’t it?

Well maybe not.

To answer thus question we need to look at ROI – return on investment again. The real issue is that it can be disproportionately expensive to be number one. If we were to graph up the costs per click versus position

It would look something like this:

As an ad drops down the page, click costs drop disproportionately.

As an ad drops down the page, click costs drop disproportionately.

On top of this we find that although the rate of click – thrus will drop away as we drop down from the number one position, it does not do this in proportion to the bid price. Further, the conversion rate per click-thru actually is higher as we drop down the rankings.

This is probably because people who click on an ad further down the list have actually read the ad, rather than just click on the first thing they saw.

The upshot of all this is that it does not pay to be number one, unless you are in a market where you have lots of margin and the cost of Adwords is largely irrelevant to you.

Often a sweet spot exists around position 4 or 5. Here you are on the front page – this is highly desirable – but still reasonably close to the top. You’ll probably being paying about half as much for your clicks, and have traffic that will convert well.

9. Failure to set up separate campaigns for separate geographic areas.

If you are selling the same products and/or services into, say, Australia and the US, set up separate campaigns. There are many reasons for this including:

The US market is a lot bigger, and clicks will be differently priced.

Your competiton will be different, and you may use different ad copy.

Spelling may have minor differences.

You may even have different domain names. Americans are more likely to feel comfortable with a ‘.com’ domain than a ‘.au’.

They operate in different time zones, and you may choose to use day-parting i.e. bidding differently on keywords at different times of the day.

10. Using campaign budget to determine your total spend.

What Google suggests you do if you have a limited budget (and who doesn’t) is to control your total spend by setting a daily budget for each of your campaigns.

Sounds reasonable – but not the best strategy.

A better way to control your spend is to actually reduce your keyword bid prices .

This will mean your ad will drop down the page. You will get fewer clicks. But these things will also happen:

  • Your ads will run all day, not just until the campaign budget is exceeded.
  • Your traffic is more likely to convert, as people will have actually read your ad, rather than just ‘donkey clicking’ on the first link on the page.
  • Conversion costs will drop.

It is important to understand that that there is a non-linear relationship between an ads rank on the search engine results page and the click-thru cost. (Explained in 8. above)

The only time when you need a number one ranking for your ad is when:

  • There is little competition
  • You have a high value product in a market you want to dominate, and the click-thru cost is of lower importance.
Be Sociable, Share!

Comments are closed.