Yet another article on Big Data, do I hear you groan? ‘fraid so… but while most of the debate so far has been “ooh look, there’s something big and scary coming”, the myths are mounting and few commentators have flagged up just what we should be doing about it.
As with all things new and shiny, Big Data is over-run with jargon and hype. “Map-reduce”, “Hadoop”, “Hive”, “Cloud based storage solutions”; our data world is now awash with a new language that all seems very intimidating and confusing.
When starting, focus your time on working out what data you have at your disposal, whether it requires Big Data infrastructure to handle it and what business challenges it addresses. Being clear on this will then let you start to pick out the parts that make sense for you. You don’t need to deep dive into complex machine learning and predictive modelling. Some data sample filters and defined metrics applied to your data can reap huge value before moving on to more complex techniques.
Where should I focus?
Big Data can also equal big distraction. As with any data analysis, understanding and clarifying your business and marketing goals and keeping these front of mind when working with various data stakeholders and development teams is critical.
Keep reiterating the overall goals as you get into your data. Inevitably you or someone from your team will start to go off track as you dig into all the data and potential opportunities it can unlock. Keep asking if it addresses the core business challenges.
What data should I be using inside and outside my organisation?
Consider what value you can extract from internal data sources and investigate what external data you can use to augment and enrich this. For example, does understanding what is happening in social media as trends enrich your understanding of your customers? Could overlaying your CRM and customer service team data with live Twitter data help you address customer queries? Could you augment weather data with your online product set and dynamically update relevant products on your site and in online ad creative?
Can I rely on gut feel?
As much as this makes some marketers despair, Big Data has little to do with gut feel and intuition, it is about having the statistical aptitude to make real sense of what you are presented with, not applying your own intuition or filter. A recent CEB study noted that almost half of marketers tested could not answer 90% of basic and intermediate statistics questions put to them.
Our own early experience in understanding the output of our data led us to feed on previous experiences and generalisations in order to make sense of our output, in an often-incorrect way. Only after spending a few weeks brushing up on statistics fundamentals were we really able to make true sense of our data. That’s not to say gut-feel and intuition don’t play a role as it certainly can help cut through some of the data “noise”, but be armed with both tools.
How much time can I save with Big Data?
One of the biggest misconceptions about Big Data is that it must be fast too. Not true, Big Data ingestion, processing and access is much slower than you think. As marketers we probably gravitate to assuming that Big Data means lightning-fast responses from our data in pretty visual outputs.
Running several queries against two years of your historic consumer purchase data will take time.
The other factor is cost: you can access big chunks of data in a faster way, but the price of having a bunch of computers continuously up and processing this data will start to lead to mounting costs.
How can I engage others in Big Data?
Keep in mind that as you start to pull in and make sense of your data that the output value can extend far beyond traditional marketing and business intelligence. Perhaps some of the data mined can help with refining existing product features and services or creating brand new ones. Perhaps customer satisfaction benchmarks for staff can be defined using your data output. Consider what you can build into your data strategy that extends the output to various stakeholders and helps them make meaningful sense of it.
On the subject of making sense of the data, at we have found that one of the best ways to understand our data is to visualise it. It is amazing how a human can see worthwhile data points, trends and even technical issues, when the data is presented in a new visual format.
How closely should I guard my findings?
Big Data case studies, architecture and knowledge is surprisingly open and transparent. There is a lot of information out in the open market about the challenges and how to address them. Your natural inclination may be to be closed in your approaches and methodologies. Being open and collaborative both inside and outside your business will allow you to quickly tackle broad technical and logistical challenges and keep your focus on tackling your unique business challenges.
We constantly learn from both indirect and direct competitors in what they have shared to the Big Data community and we also contribute back into this ecosystem. Better processes that are more efficient and are cheaper to execute developed by a collective, mean individual businesses can focus on their unique business challenges. Consider how you can engage with this community to learn and to contribute.
Ray Jenkin is a director of Affectv
