Three Common Myths in Analytics

Happy Holidays to you and yours! With the end of the year in sight, I thought it would be fun to address a few of the common myths floating around in our industry and provide some of my thoughts around them.  
 
I think the biggest struggle for clients these days is the fact that there are new technologies popping up every day and each one of them has a promise to deliver on something better than anyone else. It's a lot of noise, and fortunately firms like Stratigent exist to help sort through it all. Despite that, here are a few of the most common myths that need to be addressed and dispelled:
 
  1. Tag Management Systems remove the need for developers 
  2. Your data is accurate
  3. Personalization requires the ability to identify an individual
     
Tag Management Systems (TMS)
We've all seen the marketing about how TMS makes tagging "easy" and how "anyone can do it." Truth be told, TMS do make things a lot simpler and makes the lives of the business users so much easier. However, the "easy" implementations are really only factual if you are using an out-of-the-box (OOB) implementation and thus using the pre-built app the vendors have supplied for you.
 
Unfortunately, the OOB approach will probably only work for your simplest tags, such as ad pixels. If you are looking to do anything custom or valuable for your business you will still need to know how to write code and develop the rules within the TMS interface. Fortunately, you can do it all centralized within one place, but you still have to do it.
 
I should also mention that not having data available in the correct way on the pages themselves (i.e. designing a strategic data layer) will also make your development efforts more complex as you build out your TMS implementation.
 
Data
Your data will never be completely accurate given the nature of the digital world. Even if you spend all the money in the world on an implementation, there will still be issues in the data. But, why does it matter? 
 
Being a perfectionist is great when you're designing airplane engines, bridges, etc., but it's not going to help you with your analytics program. Too often, I see clients so hung up in the nuances within bits of the dataset that ultimately wouldn't impact the decision or valuable insights that can be drawn from the full dataset.
 
I completely understand that there are "mission critical" aspects of your deployment that must be implemented properly, but the major failing point we see is that everything gets put in as "mission critical" or clients want to "report on everything." The digital universe is evolving seemingly every day -- with that comes a lot of complications across your implementation. So, take some time heading into the coming year to identify what's most important and focus your efforts on that.
 
Personalization
Arguably a bigger buzzword than big data right now, personalization can take many shapes and sizes. I've heard quite a few people comment that personalization is only effective if you know who the person is (login, loyalty ID, email address, etc.) and can thus access a variety of bits about that person using other datasets.  While that would be the perfect world situation, there is quite a bit of value to be gained by simply using whatever you might have about that person in the moment. 
 
For example, we have many customers using third-party data from sources, such as Quantcast, to help provide demographic information about anonymous first-time visitors. Basically, you create a handshake between these third-party vendors and your strategic data layer to build an anonymous profile about the visitor with your first-party cookie as the identifier. When this happens over time, you can hopefully tie this to a unique identifier, such as a login, but you can rely on the cookie for a "good enough" implementation.
 
Behavior is also quite powerful, as well as any actions a cookie'd visitor might have completed over the last X days. You could even just build some simple contextual adjustments based on the referring campaign, site, etc. All of this can provide you with simple ways to target customers and attempt to move them along in the funnel while linking in with your marketing automation platform to take action. 
 
An important thing to remember is that if you’re leveraging a TMS, a first-party cookie is not a bad thing. A balanced approach to building use cases for the business will yield a more consistent, well-governed, personalization program for your business and ensure that you're able to show your bosses both activity and results.
 
We would be more than happy to sit down with you and talk about your 2016 strategy and how to overcome a few of the myths I've dispelled above. I hope you have a phenomenal 2016 and look forward to connecting with all of you in the coming year!
 
Regards,
Bill
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
By Bill Bruno
About the Author:

Bill Bruno is the CEO - North America, Ebiquity.

Contact Us Now