I came across a great post I wanted to share from SaaS expert Jeff Kaplan (@thinkstrategies), “In the Cloud, Big Data Scientists Need Not Apply” (E-Commerce Times). In his take on big data, Jeff asserts:
“The hype has also too often made the path to success appear overly complicated and needlessly costly when a combination of new cloud innovations and higher end-user skills could provide a more economical remedy…”
Jeff’s point jives with what I hear from users who run business and marketing analytics on PivotLink. He notes that while data volumes are growing, it’s not really “new news.” What is new, he says, is the evolution of SaaS solutions to help business users get their arms around this data.
Coming from the business intelligence (BI) space like I do, it’s clear Jeff isn’t talking about your father’s BI:
“smart business executives are recognizing that capitalizing on Big Data doesn’t demand big investments in complex BI systems and high-priced data scientists. Instead, it means giving their employees access to cloud-based analytic tools that can satisfy their growing appetite for actionable information…”
It’s an encouraging direction for marketers who don’t think about “big data” problems, they think “I’m overwhelmed with data and want one view of my customer.” It’s definitely a hot topic in PivotLink’s booth #2659 at Retail’s Big Show (#nrf13) this week.
I recommend checking out Jeff’s article and weighing in on the board. It’s a good read for anyone following #analytics and #bigdata, including marketers – the “amateur data scientists” Jeff’s referring to – who need to cut through the noise.
A guy who spends a lot of time talking to business users about these challenges is PivotLink’s CEO Bruce Armstrong. I bounced Jeff’s article off of him for another perspective.
In a reply to Jeff on the article’s discussion board, Bruce said, “I’d add that general purpose BI – even in the cloud – doesn’t solve the problem any better than on-premise BI does. The reason is that unless you come to the table with a BI *application*, tools are still just that – tools: complex and expensive to buy, use, and maintain. Self-service applications in the cloud are the opposite: easy and affordable to buy, use, and maintain. So, what is the difference between an application and a tool? Apps do things. Tools build things. If you want to build something, buy a tool. If you want to do something, buy an app – in the cloud.”
In addition, Bruce notes, “Big data analytic applications are best built by experts in a particular domain for the benefit of everyday users – and then delivered self-service via the cloud. The way to build a big data analytic application is to pick a vertical market, like Retail, and organizational function, like Marketing, and offer a pre-built application complete with data model, data integration, data analysis, and application workflow.”
As an example of what PivotLink sees, he added, “building a new customer segment to quickly capitalize on the increasing chatter on social networks about a green North Face jacket that was worn by Macklemore at his last concert, but only sending it to existing customers who have purchased North Face products in the last year and have opened the last two emails sent to them, doesn’t require a team of data scientists to create custom scripts to extract data out of Facebook, Twitter, Pinterest, Instagram, Responsys, Demandware, Omniture, and Epicor, load the data into a Teradata machine, run custom SAS Institute scripts and SQL queries, and deliver a PDF report via email – taking 4 weeks at a significant cost to deliver.”
“It’s just three clicks using PivotLink,” Bruce said.
