Gerbz Logo

Open For Business

Gerbz LLC is a transparent lifestyle business incubator. We're sharing what we learn as we validate our process for evaluating and transforming ideas into successful businesses.

Our Mission

Experimenting On The Google Adwords Display Network

Google has been sending me emails with a promocode for a free $150 of Adwords spending when I spend $50. Seemed like the perfect opportunity to dip my toe into online advertising and begin driving prospects to my site.

To be honest I’m more interested in stats regarding my landing pages and conversions than I am in optimizing my Adwords spend at this point. Hopefully once I know my landing page’s conversion rate and average cost per acquisition I can pay someone to do this for me or rely on affiliates. That being said, the ultimate goal for every marketer using Adwords is to acquire the most customers for the cheapest price – so that’s what I attempted to do.

Campaign 1: Blast

Campaign Settings

For my first test, I decided to handpick 7 “placements” which I believe best serve my target customer. I ran it as a CPM campaign with a max CPM of $2.50 and told Google to optimize for impressions and serve as quickly as reasonably possible. My thought here was that since I had targeted so narrowly my ads would convert well and it would be cheaper to pay for impressions vs clicks.

Creative

I designed 2 different 300×250 ad units and a 728×90 unit. For each variation I tested 3 different calls to action: Download For Free, Free PDF Download, and Read For Free. Here’s a sample of what they look like.

6PS Cover - Download For Free 300x250
6PS Book - Free PDF Download 300x250
generic read for free 728x90

Results

Date Campaign Imp. Clicks CTR CPC CPM Cost EB Goal DB Goal CPA
10/23/13 Blast 30,272 12 0.04% 4.48 1.78 53.79 3 – 25% 2 – 66% 17.93

Only 5 of the placements generated any impressions at all and the lions share was from 3 of them. One of those three was a forum which after generating 7k impressions, costing $7.84 and not generating a single click I paused to let the other ones run. Of the remaining 2 placements one had 3 times the impressions and a 0.05% click through rate and the other had a 0.18% click through rate.

From an Adwords point of view the results were a major flop, however my landing pages and funnel reacted incredibly well. Of the 12 visitors 3 requested the eBook (25%) and of those 3, 2 of them actually read the damn thing and clicked through to the product page.

So essentially I payed $17.93 per person do distribute a FREE eBook. These results are not looking promising but clearly its not enough data to reach any conclusions.

Campaign 2: Blast 2

Campaign Settings

This time around I decided to let Google do the targeting for me. I used the Keyword Planner to create a list of 580 relevant keywords and I excluded a bunch of “topics”. For example, I excluded “Sports”, “Beauty & Fitness”, Children’s Literature”, etc.

I set the bidding strategy to focus on clicks and let Adwords set my bid with a maximum of $.10 thinking that there must be pages on the internet somewhere that need some cheap inventory. After getting no impressions I upped it to $0.50. At $0.50 it took 2 days to get 220 impressions so I raised it to $1 and then the impressions started rolling in.

Creative

In the first campaign I saw higher click through rates of my “cover” ads than I did in the ads with the book thumbnail. This time around I wanted to get a definitive answer as to which call to action text worked best, so I only used 1 300×250 ad design with 3 versions each with the different call to action text.

Results

Date Campaign Imp. Clicks CTR CPC CPM Cost EB Goal DB Goal CPA
10/23/13 Blast 30,272 12 0.04% 4.48 1.78 53.79 3 – 25% 2 – 66% 17.93
10/24/13 Blast 2 30,332 30 0.10% 0.53 0.52 15.86 0 – 0% 0 – 0% ???

This time around I got a 250% boot in my click through rate. My CPM and CPC plummeted but my CPA was a big fat goose egg. Not a single conversion. I ended the campaign when the number of impressions reached the same as the first campaign.

In Google Analytics I dug down into Acquisition > Adwords > Placements and selected “Placement URL” as the primary placement to see the following:

Adwords Placement Urls

Notice anything interesting? Besides the one user who apparently spent 5 seconds clicking every link on my website the Avg. Visit Duration on the landing pages is 00:00:00. I can’t help but think these were all accidental clicks. I also went to some of the URL’s Google placed the ads on and they were horribly irrelevant to my customer. There is no way I’m letting Google target using keywords in future tests.

Campaign 3: Blast 3

Campaign Settings

Armed with a few insights and some experience under my belt I sought to build a better campaign. This time I went back to the 2 targeted placements that worked the first time around and added a third. I set the bidding strategy to CPC, focus on clicks w/ manual bidding at $1. After 24 hours of no impressions I upped it to $2. I also excluded some topics as I had before.

Creative

The big difference this time around was that I custom designed special ad units for each of my placements. I’m not sure if this is even allowed, but I figured why the hell not try. Basically this means custom background colors, custom text colors and fonts and call to action buttons that look like the buttons the site uses. Here are some examples:

Stockcharts ad white - 300x250
Benzinga white - 300x250
Finviz white - 728x90

Results

Date Campaign Imp. Clicks CTR CPC CPM Cost EB Goal DB Goal CPA
10/23/13 Blast 30,272 12 0.04% 4.48 1.78 53.79 3 – 25% 2 – 66% 17.93
10/24/13 Blast 2 30,332 30 0.10% 0.53 0.52 15.86 0 – 0% 0 – 0% ???
10/26/13 Blast 3 14,564 31 0.21% 1.84 3.93 56.96 3 – 10% 2 – 66% 18.98

This time my click through rates soared. The highest ad unit had a CTR of 1.56% but that was with low impressions. Of the units with more impressions had a CTR of 0.25% which I think is pretty good considering.

The strangest thing that I can’t seem to wrap my head around is why the conversion rate dropped to 10%. It could be because I’m running Google Analytics Experiments and the lower converting pages were being displayed during this test.

Conclusions

Campaign Settings

Obviously I’ve barely skimmed the surface of what the GDN is capable of. The best combination I’ve found involves targeted placements with CPC bidding and custom creative. As you can imagine, that’s also the most time consuming, work intensive option.

Creative

Custom creative is clearly king (and I’m not sure if it’s even allowed). As far as call to actions are concerned, Read For Free was clearly king.

Call To Action Blast 1 Blast 2 Blast 3 Average
Read For Free 0.07% 0.13% 0.25% 0.15%
Free PDF Download 0.03% 0.06% N/A 0.045%
Download For Free 0.04% 0.03% N/A 0.035%

Results

Overall, Adwords does not prove to be an effective medium for driving quick, cheap, targeted traffic. Not only is it A TON of work, but it’s horribly confusing and difficult to use. At an average of $18.50 for each person who received my FREE ebook, Adwords is either too expensive, or too competitive for the type of testing I’m doing. That being said, it’s certainly a powerful channel and with the right optimization can be a key ingredient to my marketing mix. But until I know what my CPA is, the time and energy requirement to run effective campaigns on the GDP is too large. Next I’ll test some other channels to compare.