Online Advertising Case Study: Facebook Ads vs. Google AdWords vs. LinkedIn

enerica
Albert Novikov Managing Director, Enerica

Posted on May 6th 2014

Online Advertising Case Study: Facebook Ads vs. Google AdWords vs. LinkedIn

I am not a big fan of Pay-Per-Click advertising. I have always believed in solid content and excellent services. But in right situations and if done well, PPC ads can be quite effective - especially on such platforms as Google AdWords, Facebook Ads and LinkedIn Ads. I have used Google AdWords since 2009, mostly to attract extra traffic to my business website. Then I decided to expand my advertising horizons and explore Facebook and LinkedIn. Facebook and LinkedIn platforms are not as mature as Google, but in the world of online advertising they have made some meaningful inroads.  As a business owner, my goal is to get best return for my investment. I set out to compare these three platforms side by side with a conservative budget of $10 dollars a day for each platform. I used almost identical ads for all three platforms with key objective to improve brand recognition in the geographic areas of our interest – Ottawa, Canada; Rochester, Minnesota; and Fargo, North Dakota.

Here is the scoop.

Google

Google has become very expensive in comparison to what it used to be just a few years ago. Keyword competition is fierce and most common keywords associated with my industry – web design - are very expensive. Cost per Click (CPC) for popular keywords (e.g., build a website) ranges from $3 to $8 dollars. Keywords with high competition are out of reach for me right off the bat. I had to focus on keywords with medium to low competition where CPC ranged from $0.3 to $2 dollars.

I had to become very creative with my keyword strategies and use combinations of words – phrases, making assumptions about what my users might look for. Since it is impossible to get it right just with a few phrases, I tried to cover as many relevant combinations of keywords as possible by creating close to 100 phrases. My strategy was to see which combinations performed better so I could weed out the weak ones. The key was to focus on specific geographic areas identified above. In the past, I have set my geo areas too wide (all Canada or all US) and as a result got lots of bogus traffic with low conversion rates/high bounce rates. Yes, with wide geo areas, I was getting thousands of impressions and it was exciting, but results were very poor. The new strategy with narrow geo focus worked better, albeit less exciting.

On Google, you also have the option to either display ads only on Google's own search network or through display partners. Google's display partners are a loose group of web entities willing to place your ads on their website. First I tried to run my ads only on Google’s own search network. The results were pretty poor. My traffic was non-existent. Then I included display-partners and my traffic soared in a matter of hours. I amassed 25 clicks at $0.48 per in a day. It is interesting that most traffic from the partners’ network came at night. It is very difficult to assess the quality of this traffic, but my gut instinct tells me it is not that great. Based on Google Analytics, this traffic looks very transitory with a high bounce rate (91.67%) and low level of engagement with the content.

Google Adwords Stats

Possibly, my ad focus was not as precise as it should have been, but creating several ad combinations helped me to identify services that performed better.

All in all, Google advertising is expensive and complex, but very robust and with a huge reach. It works if you have a unique and popular product or service. The interface feels professional, with excellent reporting tools, great usability and many various options.

Facebook

My Facebook Ads experience was the most satisfying. With a targeted audience of 30k and a $10 daily budget, I amassed 13,117 impressions, 17 clicks and 9 likes. My cost per click was pretty comparable to Google's and cost per 1k impressions was even better. Moreover, I have increased my fan base ready for my future promotions. All users seemed to be legit on close examination, however some of them were “habitual likers” with up to 2000 liked pages - suspicious.

Facebook Ads Stats

The Facebook Ads interface is quite easy and robust. It took me 20 minutes to setup my ads. It took less than an hour to get them approved – both Facebook and Google admin teams are quite efficient – Kudos!

What I liked about Facebook right away was the demographic and geo focus I could create for my ads. Due to Facebook having tons of demographic information on each user, I was able to focus on specific geo areas, interests, age groups, marital status, gender education, etc. From the close to 600k users in my area, I was able to narrow my focus to just 30k. This paid off right away. Since my focus was on small business owners, I started getting that type of traffic very quickly. I could see that users liking my page are legit as I could check their public profiles right away. My goal was to draw traffic to my business Facebook page and create as many likes as possible.

The Like methodology has enormous potential. People liking your page, automatically become your fans, and unless they choose not to receive your notifications, they become available for your future ads and promotions. One thing about Facebook – it is a place of play and leisure as opposed to LinkedIn or even Google. The secret is not to become too serious in your ads and keep them simple.

Successful Facebook Ad

I knew that, based on psychological research, people are drawn to images of human faces showing extreme emotions such as amazement, disbelief, anger and such. It also should be an attractive face to establish subconscious credibility. I used male and female faces conveying the sense of disbelief as the ad cover accompanied by actionable text. For example, one of the most effective ads I had was “Whaaaat? Your business does not have a website! We can help (image of man’s face in disbelief). This ad immediately drew attention and several likes in a matter of a day. It averaged to about one like per $3 dollars that was still quite expensive if you do not take into account future benefits of a fan base.

All in all, Facebook Ads proved to be very effective in building brand recognition. Even if Facebook users do not click on your ad or like your business, they still subconsciously register your brand name through side vision. Research indicates that subconscious recognition works. Even if a user just skimmed over your ad without consciously engaging with it, they may react to your brand more favorably when they see it in some other context (e.g., passing by your store). They may feel familiar with your brand.

This is another reason to keep messaging in your ads positive and informative. It is hard to put a price tag or ROI on such subconscious recognition, but advertisers have been capitalizing on subconscious recognition for years. You have probably seen sitcom heroes brandishing their iPhones or speeding away in their VWs  in a bid to imprint on your subconscious. This works better in Google and Facebook where your ads are more visible on landing. LinkedIn ads, for example, are more hidden.

LinkedIn

LinkedIn is a perfect place for your ad if you want to reach strictly professional audiences. It is probably not for you if you are trying to promote your fruit stand, but it definitely was for us since we are in the business of serving other businesses and building web solutions.

The LinkedIn ad interface is not as intuitive and simple to use as Google or Facebook. It feels like a supplementary afterthought. It took the LinkedIn admin team longer to approve ads (sometimes up to 12 hours, when with Facebook and Google it was no longer than two). It was difficult to navigate, and took me a while to figure out how to use it. After I did though, it was relatively easy to setup and get going.

My LinkedIn target audience based on my demographic was 10,866. LinkedIn offers two bidding modes – Cost Per Click (CPC) mode and the impressions mode (per 1k). Each option cost about $2 per measurement unit. I started in the impressions mode and had more than 5000 impressions in three hours, burning through my daily $10 budget at the rate of $2 per 1k impressions. I got 2 clicks out of that. It is hard to tell if my ads were engaging enough, or ad placement was too obscure to be noticed.

LinkedIn Ads Stats - Impressions Mode

Then I switched to the Cost Per Click (CPC) mode for the 2.5 remaining days. Using exactly the same ads I had only 1,244 impressions and zero clicks! Nothing spent but nothing gained either. My only hope was that the ads might have the subconscious effect and build recognition of my brand.

LinkedIn Ads Stats - CPC Mode

This poor click conversion might have something to do with ad placement. My ads were presented below the fold of the page on the right hand side both in the CPC and impressions modes--not exactly prime real estate.

I also was quite puzzled by the results. My interpretation was that LinkedIn profited from showing my ads as many times as possible in the impressions mode and gave up on my ads in the CPC mode. LinkedIn reporting is primitive and only allows downloading a CSV report.

Conclusion

In conclusion, for my budget and focus, Facebook was probably the best investment. For the same amount of money I got much better results, but most importantly, I got lasting fans. People liked my business page and interacted with it. They further went onto my website.

Google would probably be a better option if I had a unique and popular product that could generate unique and affordable key phrases. LinkedIn, so far, has been a disappointment both on the admin side and in traffic conversion; albeit it may give better focus for certain business sectors or if you want to reach some very specific professional strata like executives in a certain sector.

There's a lot of strategy involved in advertising, which could take years to learn. But the biggest bang for my ad dollar was definitely Facebook.

enerica

Albert Novikov

Managing Director, Enerica

I am the founder and creative director of Enerica, a full service media agency helping small and medium businesses succeed. I have been working with various communications outfits for more than 10 years – in government, business and non-profits. My interest is in strategic communications, content management, accessibility, usability, brand marketing and publishing.

See Full Profile >

Comments

Ramesh Ranjan
Posted on May 6th 2014 at 11:26PM

Interesting to see that your experience with LinkedIn ads was disappointing. I'm curious as to what they'll be doing in the next year to improve their interface and experience for advertisers!

enerica
Posted on May 7th 2014 at 12:43AM

Thank you Ramesh, yes, I don't think we will be using LinkedIn for advertising anymore, but we are preparing for a Facebook ads campaign soon and will certainly share our experience here...

erikamendoza
Posted on May 7th 2014 at 3:13AM

Great insight! This is s great article for ROI analysis of Facebook Ads

CodeImprov
Posted on May 7th 2014 at 6:24PM

Thanks for putting this together. I was curious about LinkedIn and had my assumptions, it's just not as full fledged as it could be at this point.

I think they will alter things and make it easier to advertise within the next 6 months, but I agree with your insights and will focus on Facebook for now.

terrakinder
Posted on May 8th 2014 at 10:03PM

This is a very limited study. It's important to remember to speak to your audience per channel. I'd like to know more; were you using one ad across all platforms, were you using a/b ads, were they all posted the same time of day, etc... tk

James Meyer
Posted on May 8th 2014 at 10:52PM

Interesting but my question is on Facebook it sounds like your goal was a "Like" whereas on the two other platforms you were taking the view to a sales page?


Did any of this prospecting result in a sale?  Or atleast put the probability of a sale above 50%?

enerica
Posted on May 8th 2014 at 11:21PM

Our take on the "likes" and fans is that each of them could be a long-tail direct or indirect sale. It may take time for such a lead to convert, but we take a chance. We think the ad industry is too much driven by immediate ROI and disregards long term benefits of social engagement. 

ukscoob
Posted on May 9th 2014 at 12:45AM

Thanks for the comparison.

I actually found that LinkedIn Ads was getting us several clicks, more than I had expected and as they are more targeted than say Google, it is people that we want to be clicking. I to am offering services to businesses and to be honest I think our funnel fell down on where we were sending those clickers. This is something we're looking to address and run a new campaign. The ads have an image, title, description and link and seemed to work. CPC is the only sensible option if you want clicks in my opinion. If you want to spend money on maybe imprinting your brand/business name on peoples minds with impressions then feel free, but that is rather unmeasurable. Like you say, impressions don't mean actual views by people.

One thing you didn't touch on was the LinkedIn Ad Leads. A feature from LinkedIn where you can accept lead interest. Someone can click your advert to go to wherever you have set it, or they can click to register an interest and request more information via the LinkedIn platform. We had a handful of these requests but all that were messaged failed to respond.

In terms of value, Facebook is the winner. It's cheaper and the targeting is fantastic. You can also have the ad take them to your site rather than run a like campaign. Which might be more benefitical given that facebook now restrict what page likers actually see from said pages in their news feeds.

Avtar Ram Singh
Posted on May 9th 2014 at 2:49AM

What would have been the icing on the cake is if you'd seen how many of those visits are tracked by Google Analytics. I've actually been noticing that GA only tracks about 45% of the visits Google AdWords says it sends to me, and 70% of the visits that Facebook says it's sends to me.

richtwood
Posted on May 9th 2014 at 8:37AM

I think this is well written and thanks for sharing it but I think the study itself is a bit flawed.

You aren't using the same metric across the all of the ad networks so the comparison doesn't make sense. You compare traffic increases to page likes. Granted, page likes offer a retained value in a fan base, but they aren't guaranteed site visitor, especially with Facebook’s changes to organic page posts in news feeds etc. 

Similarly, you mention at the start of the post that you were looking to build “brand awareness” for your company. AdWords (with the exclusion of retargeting display ads) is fairly rubbish at brand awareness. For a consumer, in my experience it is much more “fire and forget” unless you are already trading on the back of a recognisable brand name. Clicks are driven more by position and ad copy.

Equally, from the numbers, it looks like you have only run this test over the span of three days. I don’t think that is enough time to reduce the effect of outside factors like time of day, day of the week, weather (sounds stupid but we see a massive jump in PPC traffic when it’s wet and cold outside).

For me, a much better and fairer benchmark would have been to use actual lead generated, business secured, clients won. Something tangible that is relevant to each network and you as a business.

James Meyer
Posted on May 9th 2014 at 10:22AM

Richard, I completely agree with your observation.

Michelle Mastrobattista
Posted on May 9th 2014 at 1:05PM

I would agree with Richard below. Also, the one thing that grabbed me was that you said you used the same ad on all three platforms. Each platform has a very unique culture. The ads need to be written for/tailored to the platform. What appears as a normal ad on Google, may not look natural in the Facebook Newsfeed.

I would however agree with your findings on LinkedIn Ads. I've had a similar experience.

Michelle