Almost every professional sector is guilty of overusing buzz words, phrases and acronyms. But with the exception of banking, digital marketing seems to be the worst of the bunch.
And of all of them, Data-Driven is the one that really grinds my gears the most.
data-driven
adjective
determined by or dependent on the collection or analysis of data.
It seems harmless enough at first and admittedly quite a sensible idea. You collect data then perform actions based upon that data.
So with respect to PPC (there’s one of those pesky acronyms), you can see a huge amount of data for a given keyword, ad copy or other targeting parameters.
Based on the performance of those different aspects you make changes to your campaigns. So for a keyword that performs well you might increase the bid, even if it isn’t one that you would have expected to perform well; the data suggests otherwise.
So what is the problem?
The problem with a data-driven approach is marketeers get tunnel vision and all common sense goes out the window.
Data, for all the information it provides, doesn’t tell the full story.
Just because a keyword appears to be converting for instance, doesn’t mean it is good.
Example:
For those who don’t know, I own a flooring company (oakwoodsoutlet.co.uk) and we’ve gone through our fair share of PPC agencies and freelancers and time and time again we’ve come up against this issue.
The vast majority of our products are mid to high-end and also POA, so there are no prices on the website.
Another important aspect of this particular industry is free samples is the norm.
#1 – Our data-driven friends begin targeting the keyword ‘Cheap wood flooring’.
looking at binary ones’ and zeros’ this search term looks ok. It’s pulling in traffic and converting (I won’t share cost per conversion numbers but it was in line with expectations).
#2 – Our data-driven friends spotted this keyword performing well and increased spend on it and similar phrases.
Again, all the data pointed to it doing good things. More traffic, more conversions.
The problem is that in the real world it wasn’t working.
The conversions were for sample orders which if you remember, were free, or quote requests as the mid to high-end products also weren’t displaying prices.
So we had people searching for cheap flooring ordering samples of expensive flooring for free; costing us money in both acquisition and fulfilling the sample order.
Of course when they received their sample and discovered the prices they wanted nothing to do with us.
If common sense was on-site when the campaigns were created our data-driven friends might have had second thoughts over the campaign.
What should we do instead?
I’m not suggesting you should ignore all your data and just follow your gut; that would be even more stupid.
But, it’s super important to not get tunnel vision when it comes to making decisions based upon data. Use some common sense as well to keep the larger picture in mind so you don’t fall into this common data-driven trap.
My other piece of advice on this front, is keep a close eye on the bottom line.
Check what they are telling you is true
Marketeers love to quote their numbers and how great they are. “CTR is up X% and conversion rate is up Y%.”
But I’m yet to meet marketeers who actually pay close attention to your bottom line and if the campaign is actually worth while.
I’ve personally had to rein in many campaigns over the years that on the surface were looking good. But when I dug deeper, just weren’t profitable.
The truth is pretty much any data can be spun to support a given theory. Throw in some confirmation bias (your marketeer is hardly going to go looking for proof they suck) and you can very easily be told how splendidly your marketing efforts are going when in fact the opposite is true.
Insensibly one begins to twist facts to suit theories, instead of theories to suit facts
Sherlock Holmes
