Author Archives: Tim Ashby
As another year begins we all find ourselves looking back and looking forward. Or, as we say in Marketing Science – analyzing and predicting.
We would also be the first to point out that predictive modeling relies on the analysis of historic data. Or, in regular-person speak – only by looking back can we look forward.
So, stick with us in our predictions for 2012 because as we focus on the future we will be taking some detours into the past.
Prediction #1
Social media measurement will focus less on ROI and more on brand affinity and purchase decisions.
Legend has it that once upon a time, in a land far, far away websites were once considered an expensive and optional marketing channel. (“This interweb thing is nothing more than a fad.”).
There was also a time when social media was viewed by many as a fad. 2008-2009 saw businesses start to take social media seriously. (“Maybe this Zuckerberg kid is onto something.”).
By 2010 executives wanted to know what their social media marketing spends were doing and social media measurement took off. The listening platform industry was born and several companies offered products that measured how much conversation was there about a brand, what were the key topics of conversation, and what was the sentiment of those conversations.
As social media marketing budgets grew, those same executives wanted to know how their spends were performing. And so 2011 became the year of social media ROI (return on investment). A prior post speaks more to this topic.
As we turn the corner into 2012 we predict that the focus on measuring ROI will lessen as companies stop wondering if social media has value and accept that it is now a fundamental part of the marketing mix. The measurement of social media will evolve into measuring its impact on brand affinity and purchase decisions. Social media will stop being viewed in a silo and seen as part of an end-to-end marketing mix.
Web measurement and web analytics programs can be complex and face many technological, process and cultural challenges. We here at Critical Mass Marketing Science help our clients overcome these problems using a combination of best practice modeling, innovative thinking and the latest technology.
Problem #1 – Unreliable Data
Are the numbers you’re looking at accurate? Do they represent what you think they represent? Is everything being captured that should be? Data collection systems, especially web analytics, are complex and as a result they are also easy to mess up or break.
So what causes web analytics data to be unreliable? There are couple of common reasons, but they both relate to how the coding is done.
Most commonly we see analytics “tagging” that is incomplete or wasn’t customized correctly. This can happen for several reasons, but more importantly, how can we fix it?
Ever since the early days of the Internet, technologies have been developed to track online behavior. Over time many of these have developed into what is commonly referred to as web analytics and now Marketing Sciences.
Some people believe this is a serious invasion of your privacy. Because you sit in the privacy of your home to surf the Web, there’s a belief that your activities should be completely private. In reality though, while you may be surfing in your underwear (ok, maybe that’s just me), people like me can “see” what you’re up to.
Here’s how it works, at a basic level. Most web pages have bits of code that are invisible to the average person. Between this code and cookies that get created and stored on your browser, back end systems are able to track a wide variety of actions that you may take on one or across multiple websites.
In order to explore this further, I’d like to separate out a couple of levels of privacy.
- I.D. level privacy – Credit card information, phone numbers, your address etc. fall into this category. This is the kind of thing that I could commit identity theft with.
- Preferences privacy – Data about stuff you like and perhaps have purchased in the past, but nothing I could stalk you with. You like cookies and organic gardening, but I have no idea “who” you are unless you sign in and tell me specifically.
- Browsing privacy – Data about how you moved around in a website and what you looked at. You downloaded three recipe cards on supertastycookies.com. “You” are totally anonymous.
Legitimate web analytics and advertising tracking operate mostly in level three (browsing) and sometimes in level two (preferences), but never in level one (I.D. level). (To be clear, level one is and should always be off limits to the realms of web analytics and Marketing Science.)
Why? For many reasons, but primarily to make your experience of the web better, easier, faster and more effective. The data can be used to optimize web sites as well as customize the content you might see so it better fits your personal preferences, and in so doing help companies achieve their business goals.
I will also point out that the data is also used to present you with advertising that you’re more likely to be interested in, and therefore more likely to click on hence making advertiser companies more money.
Privacy advocates would promote the idea of a complete cone of silence and that all three levels of privacy would be completely protected. In this scenario, no data would be tracked (and I would be unemployed). Think of this like walking around a very large city with a cloak of invisibility on. You can interact with the world, but no one can see you.
The reality is that when you walk around a big city, people can see you. They may not know who you are or where you live, but you’re not invisible.
If you go into a large department store, you’re being recorded on security cameras. Much like most web tracking systems, the camera doesn’t know who you are, where you live, or what your credit card number is, just that you were there. If someone is paying attention they might also know that you looked at handbags and shoes before you left.
