Facebook Ads Experiment: How We Promote Amazon Sales

This is a real Facebook advertising experiment. As you should know, Facebook is now a popular social network, but many people may question whether it is a good platform for advertisers. This experiment is set up for you if you are interested in whether Facebook Ads Campaign is effective for promoting sales on the internet. This Ad Test may also give you some optimization tips or practical guides about how to advertise effectively on Facebook.

We, Quick Money Saver, just had a website in our network received a Facebook Ads coupon code worth USD$50 for free. Here is what the coupon looked like:

Congratulations. Your Facebook Fan Page XXX has 50 Likes. To help you build new connections, we are giving you USD$50 in free advertising, with no obligations.

Your Facebook Fan Page XXX reached 50 Likes. Grow it with USD$50 in free advertising

Congratulations! Recently, your Facebook Fan Page XXX reached a new milestone, and now it has more than 50 Likes.

You can accelerate your page's growth with Facebook Ads, which allow you to target new customers and build immediate connections for your Fan page XXX.

Be sure to enter your free coupon code when you sign-up new advertising account!

Free Coupon Code: XXXX-XXXX-XXXX-XXXX (16 digits)

Remark: Each free advertising coupon code comes with a validity period of only 2 months, and thus advertisers should use it within that particular time frame. According to Facebook ads manager, it is impossible to delay this coupon expiration date. Remaining balance (if any) will just be wasted and all unused money will become useless beyond this expiry date.

Test 1: Conduct Test For Amazon Sales Promotion

We decided to run an experiment to try promoting our Amazon sales with the USD$50 coupon through this Facebook Ads advertising program. The main purpose of this experiment was to test if Facebook Ads advertising platform should be effective for selling Amazon products.

Our advertising program was called Buy Discount Smart Phones which linked to our Amazon aStore selling hot digital products like iPhone, Galaxy, LG, Nokia etc at discounted prices. You may click through the above link to browse our store for how our testing was done there.

So we set up a simple CPC (Cost Per Click) Ad campaign with an external URL link pointing to our above Amazon aStore (This was NOT a A/B Test). One practical problem was: different tier countries should have different CPC (Cost Per Click), so how to get meaningful test results with this particular tier issue? As an Ad optimization guide, we ran this experiment with 2 separate groups of tier countries. We called them Tier 1 (for countries such as the U.S., U.K. etc) and Tier 2 (for the rest of the countries) groups respectively.

We found the CPC (Cost Per Click) of the Tier 1 country group was quite stable.

For Tier 2 country group, however, at the beginning we set our CPC at over USD$0.15 per click and the ads ran smoothly. At the middle of the test, suddenly we found the actual CPC prices dropped significantly (we do not know why) for all our advertising campaigns in Facebook system. So how to deal with this sudden bid change effectively? We immediately decided to cut our CPC bid to USD$0.09 per click (less than USD$0.1) for this advertising campaign. Neither the CTR (Click Through Rate) nor the number of impressions dropped after our CPC bid cut.

And the test result? Below are the raw data summary of each tier country group during the coupon period (Please note that we have not listed out individual data of each country here, as we do not want to violate any data disclosure condition for Facebook Ads and we feel we should better protect these sensitive data for Facebook):

(1) For Tier 1 Country Group:
Number of impressions: 53,454
Number of valid clicks: 42
CTR (Click Through Rate): 0.079%
CPC (Cost Per Click): USD$0.18
CPM (Cost per Thousand Impressions): USD$0.14
Spendings: USD$7.65

(2) For Tier 2 Country Group:
Number of impressions: 247,639
Number of valid clicks: 198
CTR (Click Through Rate): 0.080%
CPC (Cost Per Click): USD$0.11
CPM (Cost per Thousand Impressions): USD$0.09
Spendings: USD$21.9

(3) Total Summary By Adding These Two Tier Country Groups (Tier 1 + Tier 2):
Total Number of impressions: 301,093
Total Number of valid clicks: 240
Average CTR (Click Through Rate): 0.080%
Average CPC (Cost Per Click): USD$0.12
Average CPM (Cost per Thousand Impressions): USD$0.1
Total Spendings: USD$29.55

Frankly speaking, the above Facebook Ads campaign data did look good for our team. Average CTR (Click Through Rate) was 0.08% and looked fair for most advertisers. Average CPC (Cost Per Click) and average CPM (Cost per Thousand Impressions) were around USD$0.1-$0.12, which were not too expensive as we originally expected.

However, we were disappointed when we came to review sale data of our Amazon aStore, although we were selling hot digital products like iPhone, Galaxy, LG, Nokia etc at discounted prices there.

According to Amazon data, we got only 16 clicks at our Amazon aStore and had no sale (hence zero conversion) at all during that period!

Since Amazon aStore report itself does not provide any data about impressions or CTR (Click-Through Rate), we can only list down the available raw data from Amazon report as follow:

Referral from aStore: 5 clicks
Purchase through aStore: 6 clicks
Remote shopping cart: 5 clicks
Total number of clicks = 5 + 6 +5 = 16 clicks.

Assuming number of Amazon aStore impressions was 240 (equal to total valid clicks from Facebook Ads), then we estimate that our Amazon aStore CTR (Click-Through Rate) was 16 / 240 = 6.67%, which was not bad at all.

In any case, since there were no items ordered or items shipped (zero Amazon conversion) during our experiment, we can only conclude that Facebook Ads platform is not so effective for promoting Amazon sales on the internet. It appears that although Facebook users are interested in our digital products (with a sound 6.67% CTR), they still trend not to buy them. Perhaps Facebook users are still expecting social interactions rather than direct purchases on Facebook. After all, Facebook is still a personal social media network. Referrals from Facebook Ad campaign do not seem to be able to trigger direct sales effectively.

So our Facebook advertising optimization tips: As a smart money saver, we can now only suggest to avoid promoting Amazon sales through Facebook Ads advertising program.

Test 2: Conduct Test For Facebook Fan Page Promotion

On the other hand, since the coupon was given to us for helping our Facebook Fan Page to build more connections, we also ran a separate small experiment to see if advertising on Facebook itself can really get more new 'Likes' for our Facebook Fan Page.

So we set up a simple CPC (Cost Per Click) Ad campaign for our Facebook Fan Page (as an internal URL link) with an option to Like the Fan Page itself. Here are some raw data for this small experiment:

Number of impressions: 16,706
Number of valid clicks: 9
CTR (Click Through Rate): 0.054%
CPC (Cost Per Click): USD$0.1
CPM (Cost per Thousand Impressions): USD$0.05
Spendings: USD$0.9
Inbound marketing: 1
Number of new 'Likes': 1

Please note that, by definition from Facebook, this number of new 'Likes' is the number of users who like our Facebook Fan Page within 24 hours of viewing our Ads or our Sponsored Story, or who like our Fan Page within 28 days after clicking on our Ads or our Sponsored Story.

So the test result was simple, USD$0.9 for one Facebook new 'Likes'. Does it sound good to you?

Test 3: Conduct Test For Sponsored Story Promotion

Since sponsored stories are quite a new feature on Facebook, we would like also to test if sponsored stories are effective for our advertising campaign. To help our Facebook Fan Page to build more connections, we decided to run another separate small experiment to see if advertising on Facebook sponsored stories can really get more new 'Likes' for our Facebook Fan Page.

So we set up a simple CPC (Cost Per Click) campaign which was a 'Page Like' sponsored story with an option to Like the Fan Page itself, and then we got the following raw test data for this small experiment:

Number of impressions: 279
Number of valid clicks: 0
CTR (Click Through Rate): 0%
CPC (Cost Per Click): USD$0
CPM (Cost per Thousand Impressions): USD$0
Social Coverage: 58
Spendings: USD$0
Inbound marketing: 1
Number of new 'Likes': 1

Please note that the number of impressions for our sponsored story was limited and was only 279, as it was just story from News Feed and, unlike standard Facebook Ad campaign, our sponsored story could only be shown to our confirmed friends. However, since our sponsored story gave visitors a clear call to action, we still could get one Facebook new 'Likes' from our sponsored story (same as the above Test 2). More importantly, we did not need to pay for any advertising fee for this new 'Likes' (zero cost!) as the visitor just directly liked our Facebook Fan Page without clicking on our sponsored story. What a free lunch! That was great particularly for us (as a Quick Money Saver).

That is the end of our whole Facebook Ads experiment. We do hope it can really give you some useful optimization tips and practical guides for advertising with Facebook.

More Bonus Advertising Optimization Tips: one good thing is Facebook Ads will not cost more than you budgeted. If you set your whole budget as USD$10 and you have already spent USD$9.9, you can still bid a CPC ad for USD$0.15. Once you win the bid (by USD$0.15), you will only spend USD$0.1 for that bid (because your whole budget cannot exceed USD$10). This bidding system seems to be considerate for most advertisers.

Another good tips is that it is not difficult to set up and get ads approved on Facebook. It appears Facebook does not have complicated advertising guidelines as compared to Google AdWords strict rules. We did not experience any Ad campaign rejection by Facebook compliance team when we advertised our items during our whole experiment.