Big Data or Big Holes?
Big Data and the accompanying Big Hype have puzzled me for quite a while, especially in the context of business analytics.
Big Data is supposed to be this Big Concept: we have lots of data, and even more data is coming at us faster than ever. The future of humanity hangs in the balance if we do not master it, understand it, and leverage it. In a nutshell, Big Data is a Big Deal. Is it really?
Big, more, and moving fast does not mean relevant, complete, sufficient, or necessary to analyze and help improve business performance. The avalanche analogy comes to mind.
A lot more granular data can indeed help in analyzing the economy, finance, climate, and other scientific research; but it has little to do with the analytical problems faced by most businesses around the world unless they are themselves in the business of science.
The rest of businesses (which is probably 99% worldwide by number) do not suffer from the Big Data challenge. I know of no business (outside of research) that is limited by the sheer amount of its own internal data. Computers today are powerful enough to handle most of the business data thrown at them.
But I know a lot of businesses that are very limited because the data they collect is either:
-incorrect
-incomplete
-inaccessible
-irrelevant
-not timely, or
-disjointed in silos of marketing, sales, production, customer support, and finance
Or some businesses just do not collect the data that could be relevant because it is:
-too expensive to collect, or
-too hard to analyze
The most common business analysis and data problem I’ve observed over the years is the tendency to generate separate Data Towers of Babel by marketing, sales, customer service, production, and finance teams—each in a distinctive language with no reliable translator on hand.
A great example is the Digital Marketing Tower of Babel. It is made up of 4 different incompatible data file formats for email, website ranking, paid online advertising, and social media marketing. In other words, we cannot even easily understand where our Internet leads are coming from. The problem is not Big Data or the lack thereof….
For the vast majority of businesses worldwide, the problem is somewhere else. It’s the lack of technical and logical compatibility and the lack of translators between various business data files. It’s data with Big Holes. As a result, it is hard to analyze what works inside our businesses and how to allocate operating capital among various marketing, sales, and service alternatives.
Thus, guessing substitutes for infeasible analysis. Energy is wasted and money drained down the data hole.
This is the problem that humongous armies of Accenture, IBM, and other system integrator consultants have been trying to solve for a long time for very large businesses for very large sums of money. They make gigantic Data Towers of Babel talk to each other for Fortune 1000 companies.
For smaller businesses, there was no hope until very recently. Collecting relevant data and then ‘stitching it’ across departments for the elimination of Data Towers of Babel was too expensive. Also, the technology to analyze data has been too costly and too hard to use for most.
No longer. With the advent of cloud-based software, cloud technology vendors decided to cooperate. The prices of software packages went down a lot as cloud vendors spread the development cost among many more users worldwide.
The cloud software vendors decided to speak the same language among themselves for the first time in the history of information technologies. The new language is called API, which stands for Application Programming Interface. It translates to a simple but powerful concept of various software packages ‘talking’ to each other in one common data exchange language.
For the record, API is not that new. What’s new is the serious embrace of the concept by cloud software vendors worldwide. They’ve realized that no software vendor is powerful enough to impose its own language on everyone else. Until recently, the most powerful vendors have been jockeying for a position of dominance. No longer. They threw in the digital towel and started cooperating with each other and the smaller players using a common language of data exchange.
So now, one can stitch the ‘best of breed’ cloud software packages for marketing, sales tracking, dispatching systems, call centers, finance, and many others needed to run a business. This can be done at a very reasonable investment.
However, such integration is not very helpful without the tools to analyze all functions across the business. Fortunately for smaller businesses, analytical software packages also became very inexpensive. They are now much easier to use. And to share the analytical findings among all who need and care to know.
In conclusion, for most businesses, don’t worry about Big Data. Worry about patching Big Holes—left over of from numerous Data Towers of Babel still towering over your business.
#martech #analytics #digital #integration
This is so insightful! For (extremely) big businesses, they have so much money to spend on marketing, and until their stocks decrease for long enough, they don't spend the time or money to look into whether or not that marketing money is fruitful. For small businesses, their problem is usually the lack of understanding - or realizing - the connection between all of these different outlets, from digital marketing to outbound sales calls, and being able to connect and track the results of each campaign. Thanks for the read, Professor!
Very interesting article Greg, but it's not just the data that's a problem, it's the software and ownership of that which is tied also to silos of power and legacy systems, and thereby licencing. These are not things that are easily overcome, and are made more challenging when a move to cloud is added to the mix. Smaller newer businesses are sure to fair better as they are more agile to start with.
I agree completely. Particularly with operational data about manufacturing plants or water and sewage works, which is what I deal with. That sort of data is almost always incomplete and needs specialised judgement to determine if it can be used or not.