There can’t be any­one left work­ing in dig­i­tal mar­ket­ing who doesn’t accept the impor­tance of track­ing and mea­sure­ment. But are we in dan­ger of over com­pen­sat­ing with unnec­es­sary com­plex­i­ty? And are there exter­nal forces con­spir­ing to make the job of mea­sure­ment even hard­er than it needs to be?

Bob Hoff­man, author of the pop­u­lar Ad Con­trar­i­an blog, has just pub­lished a man­i­festo of sorts, demand­ing a return to Rad­i­cal Sim­plic­i­ty. Accord­ing to Hoff­man, it is intend­ed to be “a pre­scrip­tion for rid­ding our busi­ness of the imma­te­r­i­al, waste­ful and frus­trat­ing prac­tices, peo­ple, and behav­iors that are con­fus­ing the shit out of us and under­min­ing the effec­tive­ness of what we do. My belief is that it stems from a down­ward spi­ral of oppres­sive, unnec­es­sary com­plex­i­ty that has infect­ed our cul­ture, our think­ing, our sys­tems, and our orga­ni­za­tions.”

He lists a num­ber of com­plaints, but two in par­tic­u­lar stood out:

• Media prac­tices that are so alarm­ing­ly arcane, no one knows where their adver­tis­ing is run­ning, how much they are pay­ing, or even if it is run­ning.
• Data, mea­sure­ments, and process­es that have names but no val­ue.

There has been a grow­ing drum­beat of con­cern on the first point.

Despite the mas­sive growth and inter­est in pro­gram­mat­ic ad buy­ing (accord­ing to new research by the Inter­net Adver­tis­ing Bureau), almost half of UK online dis­play ads were bought via pro­gram­mat­ic last year, amount­ing to near­ly £1 bil­lion in spend­ing), there is equal con­cern that much of this adver­tis­ing spend is being wast­ed through fraud and lack of viewa­bil­i­ty. The recent announce­ment by Apple that iOS9 will have ad block­ing tech­nol­o­gy built-in has put the media busi­ness into a fur­ther fren­zy of anx­i­ety.

Over-Complexity

Are pub­lish­ers already over­com­pen­sat­ing with track­ing and mea­sure­ment tech­nol­o­gy, mak­ing it hard­er on them­selves to see what’s work­ing?

Fred­er­ic Fil­loux ana­lyzed the track­ing tech­nolo­gies used by 20 news sites and found they used near­ly 500 dif­fer­ent track­ers between them. One site, Politi­co, used 100 track­ers alone. Leav­ing aside the impact that these track­ers have on page load and site per­for­mance, the ques­tion has to be: who has the time to ana­lyze and inter­pret the data gen­er­at­ed by 100 dif­fer­ent tools?

As Fil­loux point­ed out:

“This kludge of track­ers reflects more des­per­ate moves than thought­ful strate­gies. Tra­di­tion­al pub­lish­ers tend to stuff their sites with all they can think of: you’ll notice that native media com­pa­nies (Vox, Vice or even Buz­zFeed with only 11 track­ers) are much more selec­tive in their choic­es of track­ing sys­tems than old media (Politi­co might be a fan­tas­tic edi­to­r­i­al pure play­er, but when it comes to ana­lyt­ics it behaves like an old-media.)”

Referrer Spam

So are “native dig­i­tal” orga­ni­za­tions already begin­ning to adopt Rad­i­cal Sim­plic­i­ty (or at least adher­ing to a “less is more” approach when it comes to track­ing and mea­sure­ment)? Maybe.

But even if busi­ness­es get bet­ter at reduc­ing the num­ber and com­plex­i­ty of mea­sure­ment tools, one of the most com­mon­ly used ones, Google Ana­lyt­ics, has its own issues: ghost vis­its and refer­ral spam. The for­mer has been recent­ly doc­u­ment­ed by Car­los Escalera on Moz. This is a tech­nique that spam­mers use by exploit­ing Google’s Mea­sure­ment Pro­to­col. And many sites refer­ral traf­fic data is being poi­soned with refer­ral spam from a vari­ety of uneth­i­cal SEO and Inter­net mar­ket­ing firms.

At worst, one of the most com­mon­ly used mea­sure­ment track­ing tools is deliv­er­ing a high­ly skewed view of one of the most impor­tant inbound chan­nels. At best, even where it is iden­ti­fied, the over­head of remov­ing pre­vi­ous spam refer­ral data and pre­vent­ing future spam refer­rals being count­ed is tedious to say the least (and con­tribut­ing to Hoffman’s point about waste­ful mar­ket­ing activ­i­ty).

Spam & Complexity: What’s To Be Done?

In the short term, it looks like Google Ana­lyt­ics users will have to live with the over­head of clean­ing up their refer­ral data. One can only hope that Google itself will take fur­ther action to keep this false data out of our ana­lyt­ics reports.

On the more gen­er­al issue of over-com­plex­i­ty, it is per­haps worth review­ing just how many tools are being used in the name of track­ing and mea­sure­ment in your orga­ni­za­tion – and tak­ing steps to ensure that a min­i­mum viable lev­el of tools is being used to under­stand what’s work­ing and what isn’t.

What do you think? Are orga­ni­za­tions throw­ing more track­ing tech­nol­o­gy at the prob­lem of mea­sure­ment than is jus­ti­fied by the insight gained? Is Rad­i­cal Sim­plic­i­ty the answer?