Many malls around the world have coin-operated rides to keep the kids happy. You know, the pretend fire trucks, or spaceships, that take 50c to a dollar and then gyrate around for a couple of minutes. Invariably, they have lots of knobs and buttons inside their cockpits that do nothing. Okay, they might make a sound when pressed, or tweaked, but that’s about it. They help sell the fantasy. They distract the kids from noticing that they’re just gyrating around in place - something that could be simulated just as easily (and more cheaply) by sitting on their parent’s knee.
That’s Facebook and Google marketing.
In the cost-per-click paradigm that has taken over the online advertising world, these two companies - and others like them - have gotten really good at getting people to feed coins into their machine to buy clicks from people that really like clicking on things (and aren’t necessarily the audience you want to attract), while giving you (the site owner) lots of buttons and knobs to play with that are supposedly improving your targeting and increasing engagement.
Every time I advertise via Facebook and Google, I am disappointed with the results. Then, I must ask myself, why do I keep doing it? The honest answer is - I want traffic, and they give it in droves. It’s nothing more than a hit of crack cocaine - you get a high when all the visitors hammer your server, but the morning after you hate yourself. Then, like an addict, you go back for more.
This time, though, I’ve decided enough is enough. What’s different this time is that I’ve woken up to the massive shill game that these corporations are playing. In order to explain, let me take you on a little voyage of mine…
Over the past three months, I ran identical campaigns on Twitter, Facebook and Google Adwords. The campaigns were text only. Some had a definite call to action, whereas others were full of buzzwords that should resonate with the target audience, especially when searching on similar keywords. The point is, the text and quantity of ads in each campaign were practically identical.
All three services allowed some degree of targeting (these are the knobs and buttons from the mall analogy). Facebook had the most, followed by Google, followed by Twitter.
In terms of objectives, I had two: first, to attract quality traffic to the site (in terms of low bounce rate, lots of pages, and time spent reading) - traffic that I hoped would become repeat visitors - and, second, to get people to register on the site.
What were the results?
Shockingly, Google adwords was the worst. I say “shockingly” because they not only have all the data about my website and its performance (they run Google Analytics, after all), but the ads, themselves, were not being served outside of Google’s own search results (i.e. not on any third-party networks), while the copy of the ads, also, were in line with what people who came organically to the site had searched for in the first place.
Let’s reflect on this a moment. The quality of traffic that Google Adwords brought to this site had the highest bounce rate, the lowest pages per sessions, the lowest time spent per session (by a factor of 1400%) and was four times worse on every metric than Google’s own organic search traffic - not to mention worse than the other two paid search services mentioned above.
I find this astounding.
Now, I would expect there to be some difference between people who came via a search that clicked on a website link listing, versus a text ad - but not by such a factor of difference. In both cases, someone is searching for something on Google, and relevant links and ads are put in front of their face. Why should those clicking on the ad be such a magnitude of crap compared to those who clicked on a link?
Google benefits from the largest Internet Database in the world. They catalogue every site, measure those sites’ performance, track their visitors wherever they go and whatever they do… you’d expect them to have the very best targeting data known to man, when it comes to ad optimisation.
Ah, but is ad optimisation really their goal? Or, is it payment optimisation? Is it to get advertisers to spend as much as possible in as short a time as possible?
We’ll come back to this question in a moment.
Facebook fared a bit better. Bounce rates were similar, but visitor session data was considerably better than Google’s. Even so, it wasn’t much to get excited about. The stats were not that great overall and, quite frankly, I expected a lot better. Why? Because with Facebook’s walled garden, customer analysis, and targeting tools, you could serve ads to, "cheeto-loving, chihuahua owners in Toledo who have a second mortgage", if you wanted. Or, maybe that’s all smoke and mirrors? Maybe, those are like the knobs in the coin-operated rides, again; they fuel your imagination and make you believe you are serving more targeted campaigns, when, in fact, they’re just serving up the same old slop.
In fact, I simultaneously ran a campaign to gather ‘likes’ for our FB page, along with the website clicks campaign. At one point, I had so many people from Mexico liking us (85% of the total) that I had to cut them off. Why? Because they were burning through my budget and didn’t do anything once they liked the page. They didn’t comment, they didn’t share… nada. I also realised that, although, I had mandated a University degree, most of the people I got did not have one; and, if they did, their profile would say things like they attended the, "University of Life".
That was when I started to get the feeling that all Facebook was doing was targeting people that liked to click on things - not the kind of people I was asking for in my campaign. In other words, there’s a big difference between people that click on something with the intention of going further (if it meets their expectations), versus people who just like clicking. And, in our cost-per-click world, those people are worth more to them than they are to us.
Twitter was significantly better than the other two combined. Their targeting tools were pretty rudimentary. Nevertheless, they seemed to do a better job of enticing the kind of visitor that I actually wanted. Ironically, I had to wait a long time to see results, because they are either more selective in who they serve your ads to, or they have less traffic overall (and, thus, opportunity to) than the other two. When you flip the switch on a Google Adwords campaign, it’s like a firehose of visitors. Twitter, however, is more like a trickle, but their trickle had more value.
All told, though, neither of these three sent traffic that converted to registrations. Here’s what I got (in terms of conversions to registration):
Google Adwords - 0.7%
Facebook - 0.5%
Twitter - 1.5%
Compare this to Google Organic search - 4%
For the sake of this discussion, let me illustrate what happened when I was able to get two writer-focused sites to list our short story competition in their competition listings (which is similar to the types of ad campaign I ran on the other three). Here are the registration conversion rates for them:
Site 1 - 20%
Site 2 - 27%
In fact, not only were the conversions better, but those sites sent better traffic overall. The bounce rates were the lowest (6% vs 24% for Google Adwords & 20% for FB), the page views per session and time per visit were also higher. And these didn’t cost a penny!
So, what did I learn?
I learnt that I want traffic to come from other trusted sites, rather than from paid advertising. Not only is the quality higher, it’s higher by an enormous margin. This is despite the fact that those sites offered no targeting tools other than the theme of their website.
I also learnt that Google and Facebook, predominantly, are more interested in money than in my satisfaction. Sure, they provide lots of whiz-bang analytics, plenty of tools to slice and dice targets and segment the population - but, at the end of the day, these are just hollow knobs and empty buttons. Google and Facebook are just drug dealers. They give us a huge hit of traffic and then wait for us to come back, desperate for another fix.
Now, some of you may have a more positive experience than me. That’s fine. I’m sure that there are circumstances when it works. Perhaps, a large corporation used to paying big bills for television advertising find Google a much better and more cost-effective proposition. Nevertheless, I still think they’re running a shill game. I say this, because all their bells and whistles resulted in zero improvement to the metrics, when compared to their basic service.
If you need more convincing, consider that Facebook has recently reduced your brand's exposure to its own followers down to 2% (if you have less than 500,000), unless your post is shared by those few that it's shown to, or you pay to boost it in a campaign. Facebook claims this is to reduce spam, but we all know what it's really about. They're just greedy. After paying them to get all those likes in the first place, they won't even honour their deal and post your update on your followers' Timelines, unless you pay them more for the privilege. As far as I'm concerned, that's a travesty that undermines their whole ecosystem.
Twitter, which uses a more black box approach, actually did a better job. In fact, I’m experimenting now with Outbrain, which is totally a black box, yet delivers results equivalent to Twitter (so far).
Yes, Outbrain uses a content-driven approach to promotion, rather than advertising, but the point is that they use a black box. They give me no knobs or buttons whatsoever, yet deliver similar, if not better, results. I'll let you know how that goes, when I get more data in.
Here’s what it comes down to. Google and Facebook hire an army of PHD’s in mathematics to design algorithms to scrutinise the world. These are the people who should be designing the black box and running my campaign. I’m not an Internet marketing guru, nor am I likely ever to be. If I’m paying them good money, which I am, then they should be the ones to optimise my campaign in such a way as to maximise my return in terms of what I want. They know how to do that - I don’t. Don’t give me knobs and buttons and tell me to have a blast, when all you’re doing is maximising your revenue at my expense.
That’s why the black box approach is the most honest. It says, “We’ll analyse your content and we’ll place it in such a way as to get the highest metrics for you that we can - for the same cpc as anybody else”.
Beware of fancy knobs and buttons.