Understanding Big Data
By: Sitecore MVP Randy Woods
A collection of our thoughts on Big Data: what it is, how you can use, and what it can do for your organization.
What is Big Data?
Wikipedia explains Big Data as “an all-encompassing term for any collection of data sets so large and complex that it becomes difficult to process using on-hand data management tools or traditional data processing applications.” Great, so now we know what it is, but what can it do? In this post, we take a look at the phenomenon from many sides.
First off, let’s look at where to begin and what to avoid. Big Data: Is the juice worth the squeeze?
Big Data: An origin story
Everything has its own origin story, including the evasive big fish that is Big Data.
And just as that fish is likely expanding at an alarming rate as we speak, so too is Big Data. As the data grows, so too do our data mining capabilities for business. Between social networks, review networks, government data sources, blogs, media outlets, website interactions and cutting edge new players making waves like wearable tech and mobile apps, the expansion of Big Data is limitless.
For a deeper look into Big Data origins read: Where did all this Big Data come from?
Dip your toes in the water of Big Data
Now that we know what Big Data is and where it comes from, it’s time to go for a swim, but even the most basic swimmer knows to test the waters. We need to dip our toes in and get a feel for what’s out there.
The first three considerations to analyze Big Data sound basic, but are often overlooked.
First, you need to know what you want your visitors to do. What are your objectives, target actions and goals? Second, you must track your conversion rate. This will provide you with incredible insight into the effectiveness of your website. And last but not least, you must understand website abandonment- understanding why visitors are leaving your website will allow you to see where your website may be lacking in content.
For more information, take a look: Three considerations to help you start small with Big Data
A little fish in a big pond: Extracting quality from quantity
At this point, we generally see one of two reactions:
- Those who become incredibly overwhelmed by the extensive amount of data and deem it next to impossible to make actionable
- Or alternatively, those who realize that there IS a way to make Big Data more accessible and that they might actually (gasp!) enjoy the process
We encourage you to become part of the second group.
The first step in heading in that direction is to understand that, like most any other technology solution being developed today, your work with Big Data must be approached with a user-centric point of view. Ensuring that you are delivering valuable content for those who need it in an intuitive way is key.
To read the rest of our insights into finding quality from quantity, check out: Extracting quality from quantity: Preparing for Big Data analysis
The hook, line and sinker: Big Data ethics
99% of all the world’s data has been collected since 2012. And with that data have come amazing achievements in technology, medicine, entertainment and beyond, each with their own set of challenges associated with the collection, analysis and presentation of Big Data insights. Chief among these challenges is how to deal with the ethical concerns of using this sensitive data.
Is it ethical to gather and use information about your consumers to create customized, targeted advertisements? Or is gathering and using that information a violation of personal privacy? In other wise, when does customization become creepy or go too far?
Companies must understand that their Big Data policies are representations of corporate values, and while its use may not necessarily have legal implications, they must consider the ethics and perception of targeting consumers in this way.
Not sold? Read our post to see examples: Understanding Big Data ethics
Big Data: X marks the spot
The next step in your Big Data analysis requires mapping out your data journey.
You must thoughtfully develop a protocol to follow when gathering, analyzing and actioning on Big Data. We have borrowed a protocol from the natural sciences to provide you with a step by step methodology that will establish a standardized procedure to ensure simple replication.
This 10 step protocol will help you avoid easy mistakes, be efficient, clear and concise, and ensure valuable insights are produced from analyzing your Big Data.
To have a further look into these 10 steps, take a look at: The 10-step Big Data protocol
How to make Big Data work for you
Wile E. Coyote wants to catch the Road Runner. Tom wants to catch Jerry and you want to see the results of your Big Data efforts.
For marketers, those results drive decision making across various channels, particularly in the ecommerce space. Remedying the issue of cart abandonment is crucial for companies relying on online sales to drive their business, and Big Data can help. Wondering how? These two steps are a good start but be sure to take a read of the attached post for a more detailed explanation of how to use Big Data and the Sitecore DMS to rid you of your cart abandonment problem.
Step one: Go find yourself a data scientist. We’re kidding, but you will need someone who has the skills of one, which is an easier task than you’d think. If you’re working with the Sitecore DMS, you’re already well on your way! The system already has many of the tools needed to customize.
Step two: Analyze the conversion flow. This requires combing data from multiple sources, including the DMS, your CRM, or any finance tools you have included as part of your system.
Check out the following article to see how using Big Data can eliminate cart abandonment:Eliminating cart abandonment with Big Data and Sitecore DMS
Critical data for critical learning: Higher education and Big Data
Big Data provides significant insight into an organization’s target audiences and what information they find valuable. When you provide your audience with the information they are searching for, you become a trusted and/or leading source in your field.
To prove this, we demonstrate how Big Data can provide essential insight for higher education institutions and as a result, increase revenue.
Post-secondary education institutions are already collecting data to monitor retention, but have typically failed at creating a comprehensive picture of analysis for the rest of the information they’ve gathered. Just what could they be doing with this extra information? Easily evaluate the relationships and correlations between variables and students connected to the institution to create effective marketing persona or increase alumni donations through a thorough analysis of what motivates alumni to donate.
Intrigued? Read: Bringing Big Data to higher education
Big Data: Adjusting the all or nothing metric
Let’s face it. The ability to use conversion metrics to assess the value of visitors is very black or white…but what if you want metallic grey or neon green?
We’ve got your back!
Unfortunately, the biggest challenge with engagement models is building the model in the first place. There is no “one size fits all” approach for every marketer to use for their campaigns. Your organization will need you to design a model and then apply it to your content.
The Sitecore DMS (and other marketing-aware CMS products) provides marketers with an alternative. Sitecore can score the value of a visit based on the content viewed and the actions taken. This provides an engagement value, which allows your organization to assess the value of ANY given piece of content or action on your website.
Learn how it’s done with Making Big Data work: Visiting scoring and engagement value
The “not provided” apocalypse guide
Any digital marketer will tell you that traditional website analytics relied heavily on the search results data provided by Google to determine the relevant keywords related to a given site, shedding light on consumer motivations and actions. We were avid users, integrating organic search keywords with the Sitecore DMS to customize our own website experience.
nonlinear has been using search term techniques for personalization and we can use this data for personalized marketing. Products like Sitecore let us target content based on sophisticated user personas.
For more information read: Coping with the “not provided” apocalypse
Disinfecting dirty data
Missing customer records, multiple representations, existing outside reasonable ranges and multiple entries in the CRM are just a few examples of the most common quality issues in data systems.
Failing to clean your data gives you a high probability of false findings and skewed data. The “insights” end up being more smoke and mirrors than anything else and you’ll have wasted your time, energy and revenue. If you’re dealing will relatively small data sets and a limited number of sources, it may be possible for a developer to use scripts or ETL tools to create a uniform view of the data.
Take a look at: 5 tips for cleaning dirty data
