Today, I want to share with you Experiment 2. But before I do, let’s talk about a couple of your concerns here:
Why is a high Click-Thru-Rate (CTR) & Zero Conversions Considered a Success?
Something I left out yesterday was that I was testing my keywords with a brand new landing page that had never been run before on any PPC experiment.
I would normally NOT test more than one variable at a time, but the double-digit-CTR was an unexpected bonus result. Because, at the time I put up the new page, I was trying to improve conversion rates, not click-thru.
When I started, I was convinced that the new page would perform better than any other before it. After all, it had all the elements of sound marketing and was extremely Google friendly as was seen by my very high Quality Score.
Unfortunately, this was a bad assumption.
It converted nothing. So after I made my post yesterday, I changed the target URL to the old landing page that used to convert at 12%. When I did, I finally started to get some conversions!
Either way, though, the high click-thru-rate was a good result. Because you can’t get conversions if you’re not getting enough clicks. By increasing click-thru, I increased the pool of visitors to my site that I could potentially convert with a better landing page. I can then spend the rest of my time working on finding that winning, better-converting, landing page.
Scrapping the new- and returning to the old-landing page was the first step.
Should You Ignore My Previous Advice About Having Only One Keyword Per Adgroup?
The advice is based on an important adwords principle that will never change. Always, always, always keep your adgroup keywords around a single very tight idea. Your keywords and ad should be a perfect match for one another. You get ultimate control with a single keyword.
Adding hundreds of keywords into the same adgroup was just a temporary research phase that helped me avoid customising new ads for all my keywords without knowing which ones are worth my time.
It cost about $30 to find out.
But if i continued to run this as is, I would lose that $30 every single day.
Experiment 2: The Control…
This morning, after a full day’s worth of data, I studied my keywords closely.
I sorted by CTR and considered breaking the ads with high CTR out on their own.
Bad idea. Some of those keywords only had a few impressions with very high CTR. With the passing of time and more impressions, those keywords may sink to the bottom.
Then, I considered the ad position. I looked at keywords that had lowest rank and considered increasing their bids. But what would that do?
It would get more clicks…but not conversions.
That’s an experiment best left for another day.
Several ideas later, I decided what to do: I sorted by conversions and found I had THREE top performing keywords. These were the “diamond in the rough”. Their combined CTR was 9.28%. Not quite a double-digit-CTR but still very high. They had several hundred impressions each from yesterday’s experiment. And each had converted at least once in the last 24-hrs.
I paused all other keywords. They were just costing money and would distract me if I left them on. I’ll come back to them later.
Meanwhile, I let the three sexy keywords keep running.
Today’s experiment is called the “Control”.
I will run them for 24-hrs without changing anything. I want to see if they continue to perform well after having swapped back to the original landing page. It’s probably a bad time to run such an experiment because it’s Saturday and that’s rarely a typical day.
On the other hand, I just need some preliminary data to tell me it’s worth working on the landing page. I’ll run more “statistically sound” experiments later on the landing page.
Stay tuned, let’s see how this pans out.