How Effective Are Facebook Ads?

Not really encouraging results. My business tested Facebook ads a while back but we concluded that it wasn’t really worth the trouble. Perhaps with all that’s happening around the so called “social search” things will change for the better, but I guess we’ll see.

There is an authorized 3rd party tool called Edge.BI on Facebook that can help advertisers know the real ROI of their ads, taking any back-office / CRM metric (e.g., real sales) into account. This is a leapfrog over standard conversion tracking. Check out Facebook’s official Ads API Vendors page on: http://goo.ly/77ST5

What jumps out to me is an 80% increase in average CPC from 2009. That’s a big increase. Any thoughts on whether CPCs are still increasing or will stabilize?

Jeff, I have pretty much written off Facebook as an effective advertising platform. It’s not just the WebTrends study, but numerous other studies. It’s not pretty. The ad-supported revenue model for social networks is broken and has already reached a critical inflection point. I am not the only one that believes this. You can read about it in my blog. http://bit.ly/hINWCy You are welcome to share this information on your blog or RT. Thanks, Tommy Toy a.k.a. @turk5555

It is always interesting to me when I read how people say they tried or tested some form of media and decide it “doesn’t work.” Really, is it the media that didn’t work or was it the way you used it or the campaign you chose to run? Just running an ad is not a true test. Running the right ad is. Yes, I know that is the tough part. However, I can tell you from experience that Facebook can work and does work if done right. What do I mean by right? Well, it is like anything in marketing, many marketers think if they hang up a shingle or  just throw something out there in the form of an ad or a communication, they should see some ROI. Well, you probably will, but not what you were hoping. It amazes me how many big companies with smart marketers in them, forget many marketing 101 concepts when approaching new media (and old). People still have to want what you are selling. It still needs to be relevant to the market or audience you are targeting. Facebook is no different. Yes, it is a great place to get friends to recommend to friends. Yes, it is a great place to be a billboard on the busiest highway in the superhighway structure. But if they are not in the mood for a burger, they are not pulling over at McDonald’s no matter how big their billboard is. However, that should not stop you from going after all those hungry people that are in the mood for a burger. Facebook has an incredible way to target. You can narrowcast on Facebook and it should be used. No one should just stick up a page on Facebook and expect it to fill up with “likes.” This is not the Field of Dreams. Just like any campaign, it should be multifaceted. If you put up a page, use the Facebook ads to hone in on your customers to get them to come to your Facebook page, or wait, maybe even directly into your store. I just did that over a six month period of time at a very large computer company in Texas. By targeting and giving people what they wanted based on what we knew about them, we were able to increase our relative ROI from $.25 on the dollar to $4 for every dollar we spent. Now, there was a learning curve, and our dollar amounts started low as a test, but then as we honed and achieved better results, we were able to turn great results. In just 104 days we were able to increase our overall results by 12 times. Drive more than 600 million impressions, 200K clicks, and $500K in revenue. If marketers stop being mystified by new media and apply the basics, it is possible to sell on platforms like Facebook.

In some ways echoing, in some ways mirroring, and in other ways extending his observation remarking the self-serving aspect of modern “strategic” management, the prevalence of self-serving initiatives that are rationalized, justified ex post rather than flowed as a logical consequence from an understanding of a systematic model of reality that anyone had; 

I’m not referring to validity here–only that the rationalization, ostensibly to convince oneself that it was an appropriate decision and not an arbitrary one that might or might not have been more appropriate [without a systematic framework, it's difficult to be quantitatively defensible, which grows in importance proportional to the size of the budget or capitalization of the company...]… 

In other words, it very much is a parallel to the humility that has been replaced in the pure and applied sciences with a sort of hubris that ironically blinds us to truths, deferring to an imprudent holding out for humanity to “get it right, we’re so close…” Whether they’re leeches or statements like, “Most people are on Facebook for fun so it makes logical sense that industries such as tabloids and blogs top the list for (CTR) at 0.165%,” both of these assertions, without place in context of a defensible episteme that gives quantitative contextual interpretation to both the premise itself (“Most people are on Facebook for fun,” and whether “top of the list at 0.165% CTR” is compelling or disappointing–remembering my favourite truth that St. Augustine said so well, ‘Right is right even if no one is doing it; wrong is wrong even if everyone is doing it.’ In financial analysis, it’s even more clear THAT we have the same problem (ostensibly why we have housing bubbles, tech bubbles, Greek bubbles, …) but at the time, again, I emphasize that it’s perfectly all right, and in fact a good exercise of critical reasoning, to reject a theory based on an understanding of it equal to a mastery of it–in fact, as you know, most of the theory that I hold to be more representative f reality is what would be called heterodox; but I can’t validly reject what I didn’t really know in the first place–

I mean I *can* but that phenomenon is what I call the 5 year old an the Ferrari (remember, the chick that wnted an ice cream, and her dad gave her the keys to his Ferrari, cuz he was too occupado with his buds to get her ice cream.. and somehow she makes it driving across the street and back… so she now thinks that she totally can drive–safely…so even if the theory weren’t applicable, if one didn’t arrive at the rejection of it as a syllogistic consequence of theory compared to empirical observation, most importantly within a valid episteme (being able to change relevant contexts is totally self-serving cuz then you can state exceptions when they really don’t apply most of the time.. or ignore relevant exceptions on the same bases) I mean, for example, whether you agree that Markowitz’s “revolutionary” innovation and extension of the capital asset pricing model: Modern Portfolio Theory–is a meaningful model of the interactions between risky securities in a systematic way, How quantitatively “good” a model it is (for example, correlation coefficient in regression; validity, or r-squared value of a statistical model) you can see easily made irrelevant if you then point out that the model is rigorous for a TWO-asset portfolio, largely contingent on the concept of COVARIANCE between the two–so if you can in any way undermine the utility of a two-security portfolio–for example, if you show that even if it is not all that different with more than two, “nobody has a two-asset portfolio,” or “an index fund would never have only two assets..” then you completely undermine what would otherwise be a useful model.So therein lies what I’m saying essentially–the key isn’t whether or not a model ought to be rejected wholly or not; ostensibly they would not have made it through a process that passed peer review if it did not have at least a significant significance–but it lies, again, in the critical thinking requisite to understanding the models in the first place–and then the higher order synthesis to adapt it to what you *do* believe to be more/most relevant.(We’re essentially both saying the same thing now–what you just said, “If you think about it from a social network perspective…” which is exactly what I was addressing–that unlike in science and theory construction, in practice, we either no longer have the skills/ability or willingness to identify the most relevant epistemological lens through which we set out to investigate something, and then continue to use the same one for analysis and interpretation–and that’s why we can now so easily say on the one hand what we think we want (which people still confuse with “result of what I’m willing to do”–they’re not the same things, of course–for all the talk about being goal-oriented, not being able to see the difference between an outcome that is the result of what we’re willing to do and setting out to “git er done” in order to reach a goal that we know aligns with what we need/want….)Consider the case of “stuffing channels.” As a major (but not as major as Top Purchase, for example) software vendor, you buy up a gazillion copies of the new Windows 15,372 in anticipatinoe of what your marketing research tells you (I’m explicitly leaving out analytics from here, since it speaks to the validity of an individual purchase propensitiy, not a “typical individual that fits this segment” and that I wanna leave out)–so whether your numbers on the sales for each day of a campaign are bang-on or not doesn’t even matter when you consider that there is a defensible number of units that you can have ordered, not yet paid in full, and that you have shipped, sold but not shipped, have deposits in receipt for but neither shipped nor even received yet…. The acceptability of whether these expenses and revenues can, ought, should be included in your weekly, monthly, quarterly revenues–and in fact, even having transitioned to IFRS and new revenue recognition rules less than two months ago in Canada (the US still dragging its clubs on this–and we’re already amongst the last to deprecate GAAP, really in favour of an international standard–not that a given rule for revenue recognition is appropriate for all countries–the Euroe, we knew this to be a fatal flaw due to diferential cultures, values–even within Canada, there is pressure on our own dollar due to local inflation that we artificially are able to mitigate)

 Absence or presence of perfect information is almost beside the point in many cases, because if you can defensibly call into question the *applicability* of a model with known validity in a situation, then nobody cares if your estimates are, to the penny, exact, for sales each hour, minute, year, day, quarter–and not investing that effort, time, into fully elucidating what construct is responsible for what we believe we see, and consequently, what we believe to be the most appropriate modulator of these outcomes, when it comes later to C.Y.A., Jeffery A. Thomas, must necessarily either turn into a pink slip or ex post rationalization pointing out some critical factor that “could not have been anticipated.” http://robincheung.info/samples/walgreens.pdf isn’t a very pretty report but it illustrates what I meant by not only sensitivity analysis (what if my estimates are 5%, 10% high, low, sideways?) but in fact, the very reason that I’d require of any similar valuation at leat a parallel evaluation of at least three–hopefully more these days–valuation models themselves, their intrinsic behaviours in certain contexts, what spread of valuations they would come up with, and which factors in a given situation being most relevant would cause me to favour/value which valuation more strongly than others for which reasons, no situation being a pure exemplar of any of them [so I’d expect still always to see multiple valuations, still see a quantitative comparison of critical factors, and still see which valuations you’d favour under what contingencies, then–for example, a valuation method that weights more recent earnings heavily, I’d still want to see, accompanied with an evaluation of the likelihood that the situation could change to make it more relevant than the one currently believed to be most representative, and why…)The problem is we’re not even doing that anymore–so what we have a responsibility to due diligence to do, we’re excusing not really because we’ve done the evaluation of it and decided it wasn’t really relevant–otherwise it would be in the discussion–but because we don’t want to do it, and instead rather spend the time on something else risky without due diligence, too

Thanks Robert…so very helpful