GIGO: Garbage In Garbage Out

Since our industry loves acronyms and we have all of this conversation around data, this seems fitting for the new acronym of the month(s).

What exactly is GIGO, it’s a term coined in the technology space for a computers ability to process even the most corrupt piece of information and spew forth mountains of useless information, as a result of the “input” being suspect.

Why does this matter to you or I in the digital media space? It matters because presently we are constantly looking for the new piece of technology that we can “bolt on” or use as an aggregation point to corral the massive amounts of data that is created minute by minute in our space.   We want to make sense of it all and use it to earn more revenue.  Unfortunately our race for more and better often involves using bad data sets (garbage in) and producing output that is incomplete (garbage out) and placing it into excel.

This process is akin to taking a person from China, one from Malaysia, throw in a UK lad, and for good measure, let’s take a German and add them to the mix.  Now in their own native language, let’s get them to share notes and talk about their day.  While this is happening, let’s take you or I, and have us translate and compile the conversation.   Don’t see much success with this happening.  We don’t speak the language and each time we butcher the effort, we’ve corrupted the message.

Can this problem be solved?  Sure, every problem has some options. 

One option is to take a multi-lingual person and drop them in the middle.  They may not fully understand the Chinese dialect but they get the pronunciation patterns of the UK lad and they speak German fluently.  So they are getting “close enough”.  In our industry it’s the same as taking one vendor, asking them to integrate their product with another, possibly dump in another vendor for good measure and then use something at the end that will bundle up and report on everything.   That’s the common solution we seem to take.  Just doesn’t seem like it’s going to end well though does it?  And it doesn’t countless times. 

The other option is to hunt high and low and find that person that understands all the languages, understands the vernacular, is an incredible translator and is adept and pulling this conversation together and summarizing for easy digest.

Most take the first option.  Why do we keep going down that though?  If we know it doesn’t work, why are so many continually trying this path? 

I would say that we all have this dream that an easy solution is the right solution. 

The reality is that a majority of those involved in our industry have purchased technology as it came available.  Each time there were limitations and deficiencies and to rectify those meant starting over or adding on?  Without any technology vendor choosing to pause and rethink things, options were limited.

When I talk with most who are knowledgeable of data and how it can be leveraged for gain, all the while reducing this concept of GIGO, they typically point to two options.

One is get technical very quick and bring in expertise that has worked with big data sets and has experience with aggregating data sets using common identifiers.   Creating the common data language for all systems to merge into, presents an easier path for actual mining an analysis of the information.  These people are typically working at organizations like Amazon, Google or the NSA.  They study math and computer science in college and they have an understanding of cool things like Hadoop and MapReduce.

The other approach is to do the work in one swift move replacing your core technology systems to a more comprehensive platform.  A solution that was built with today’s data environment in mind.  One that speaks the same language across components so data is available in one place for easier retrieval and analysis.   This method limits your exposure in the long run as integration points are reduced, data is in a common language for easy analysis, and the solution is actually architected to handle today’s data load.

So now you have these two paths, what can you expect to get out of a system that is actually managing data in the proper way?  Well, for starters, you will greatly reduce the GIGO syndrome that so many companies suffer from.  More importantly, you’ll speed up the process of identifying performance in relation to KPI’s for your business.  Looking across multiple data points, sales management will be able to identify unsold inventory, revenue associated with the inventory, historic purchasers, price points and sold rates in relation to rate card.  For operations people, the ability to quickly identify campaigns that are under or over delivering, campaigns without creatives, and more efficiently respond to requests from the sales teams will surface naturally as information is readily accessible. 

Looking toward the future, having data available and formatted in the proper way, allows for easier visualization and perhaps predictive modeling.  All nice things as we scale our industry and make decisions on direct selling or programmatic buying as beneficial for a publishers business.

Chris HanburgerSince our industry loves acronyms and we have all of this conversation around data, this seems fitting for the new acronym of the month(s).

What exactly is GIGO, it’s a term coined in the technology space for a computers ability to process even the most corrupt piece of information and spew forth mountains of useless information, as a result of the “input” being suspect. Why does this matter to you or I in the digital media space…