Big data good, big (shared) data better?
Are there two types of Big Data? New Big Data and the old one? And which is more equal than the other?
The coining of the term big data to describe the voluminous quantities of structured and, crucially, unstructured data that “traditional” (often relational) database analysis would fail to process has been something of a media rollercoaster.
Amidst woeful warnings of the “inherent danger and complexity of big data”, vendors have sought to offer deeper and ever more insightful data analytics (much of it cloud-based) to alleviate us from our information-based afflictions.
But is beating big data really just about new ways to exert additional cloud-centric processing power and storage intelligence?
Perhaps we have a lesson to learn from George Orwell’s Animal Farm i.e. “All big data analysis approaches are equal, but some are more equal than others if they empower the datacentre with shared big data intelligence from best practices.”
Increasingly, businesses don't just want Software-as-a-Service to save time, money and administrative hassle. They demand best practice and data-driven guidance integrated into the product experience. This is the view of Adrian McDermott who leads Zendesk’s product management, engineering and operations teams.
McDermott points to Google AdWords which not only made it easier to buy advertising “by the drink” with its pay-per-click mode, it also educated customers on what words to buy and for how much to pay for them with tools like AdWords Optimizer. There is a “shared collective intelligence” message here and we can take it forward to the world of cloud CRM.
“Salesforce relocated Siebel into the cloud and turned a 12-month complex implementation into a 12-week one, but it still requires customisation,” asserts McDermott who says that his company believes in "convention over configuration" as a big data strategy.
Zendesk uses shared data across 20,000+ customers and their 15,000,000+ monthly support tickets to expose best practices and benchmarks before then integrating this intelligence into its product.
“We recently launched a point and click configuration integration with salesforce that allows customer support managers to tie salesforce automation data to customer interactions and conversations within the support. Zendesk is replacing the Force.dot.com developer with the empowered customer service manager."
Big data has been variously described as a shoebox of allsorts or that drawer in your house that becomes home to a mish mash of miscellaneous odds and sods that really ought to be filed and sorted into their proper home somewhere. This assertion towards being able to gain insight into big data through shared intelligence rests upon the assumption that we can always apply the behaviour of these 15,000,000+ monthly support tickets (or whatever base we use) to new data as it occurs.
But if new data is new big data that doesn’t sit in any kind of comfortable parallel with the shared big data we are trying to use as a template for intelligence, then surely the argument breaks down and Zendesk’s model only serves to obfuscate and confuse. Doesn’t it?
If it does work then does this mean that Salesforce itself and similar firms such as RightNow Technologies are being classed in some downgraded form as first-generation SaaS customer tools? Can collective intelligence from best practice be used to drive convention over configuration for more efficient cloud-based CRM in this way? Will shared big data ultimately champion over big data as once knew it?
The answer is that it surely could. Zendesk itself is a partner with Salesforce, but the product here is Zendesk for Salesforce – and not the other way around. Keep reading for now though as we’re not quite in line with Russian Revolution in terms of Animal Farm disruption yet.