Predictive Analytics: The Perfect Use Case for Cloud Computing
cloud help predict the future? Okay, that’s a loaded question, but there are certainly voices out there making the case that cloud computing provides the processing and big data support needed for predictive analytics. Predictive analytics — matching current datasets against historical patterns to determine the probability of an event occurring in the future — requires a lot of compute power and draws on a lot of data. In other words, a perfect use case for cloud.
James Taylor, automated decision management proponent and author/co-author of two books on the topic, says cloud computing is elevating the art and science of predictive analytics to a whole new level. No longer do such efforts need to be be constrained by companies’ current server and storage capacity — with online, sharable resources, the sky’s the limit. Based on a survey Taylor conducted at the end of last year among 200 business intelligence professionals, 43% have already developed predictive analytics solutions within their companies,and 82% have predictive analytics in their plans going forward. “Separately, predictive analytics and cloud solutions are changing the way organizations do business,” he states. “Together, they open up a wealth of opportunities.”
Other industry research confirms the growing level of interest in moving business intelligence to the cloud. Another survey of 1,364 IT managers by Gartner, for example, finds that almost a third (27%) already use or plan to use cloud/SaaS options to augment their BI capabilities for specific lines of business or subject areas in the next 12 months.
While predictive analytics has a range of applications, from fraud detection to production system management, Taylor’s survey identifies the “sweet spot” for cloud-based predictive analytics as the effective acquisition, management and retention of customers. The top two areas for predictive analytics projects ar marketing/customer acquisition (with 61% implementing or having specific plans to implement) and customer retention (50%). Other areas that scored well included the broader category of customer management (48%), sales and cross-sell/up-sell (46%). A similar question focused on cloud adoption showed campaign management (60%), and CRM scoring highest (59%).
Time to value, pervasiveness, agility, scalability and data access are the key pros of cloud-based predictive analytics. Taylor also points out.
Organizations don’t want to wait more than a few months before seeing a positive outcome, he explains. “Using cloud-based predictive analytic solutions has a much faster time to value than alternatives and this represents a critical advantage for the approach. This rapid time to value is also generating interest in cloud-based solutions for testing, experimentation and development even where deployment will ultimately not be cloud-based.” And since cloud-based solutions focus on simple, standardized interfaces, “this minimizes integration effort and cost, reducing an organization’s dependency on IT resources to adopt these solutions. It also allows tight behavioral integration without tight systems integration.”
Access to big data is another key advantage cloud computing offers. “Many new big data sources are only available in the cloud,” Taylor says. Plus, cloud means organizations will be less restricted by data transmission rates. “All this increases the value of moving analytic modeling to the cloud where it can be near these new sources of data.”
Examples cited by Taylor include a fitness company that sells directly to consumers that was able to create a cloud-based analytic solution to build buyer and inquiry models to target catalogs for one of its brands, resulting in a 25% higher response rate and 44% higher return on investment compared with prior campaigns. In the case of a credit card company, a cloud-based predictive analytics approach was instrumental in targeting and shutting down “bust-out fraud — in which individuals suddenly max out their credit lines and then disappear – “three to five days earlier than existing industry models, saving the card issuer more than $40 million per year in losses.”
Why Cloud Providers Need to Change Their Approach to Cloud Services
While cloud services aren’t exactly a new idea of service delivery for Canadian organizations, there is still much to be desired.
Right now we are seeing a strong offering of IaaS services, the odd SaaS (as in the case of big Blue and their Microsoft partnership) and managed (hosted) network services such as WAN optimization, virtual PBX and security.
The great thing about this is that we are seeing some traction in cloud providers in Canada to offer these types of services, but we’re not quite there yet. The reason is that decision makers in the provider space are still trying to figure out the business case for cloud.
The biggest disconnect I am seeing is the failure to link all these disparate solutions together. Right now most services are offered a la carte, from different corporate divisions (Security, UC, Network) and there is no alignment. I would love to see a provider who is not afraid to start from scratch and realign services to vertical, but don’t expect it from the larger, slower moving providers.
Here is where the little providers have a chance. By realigning services to vertical markets, you can build service portfolios that speak to these groups individually. For example, if you decide that education is a good market, you can build services that address the unique needs of education (BYOD, content filtering, security/privacy) and offer a solution that provides all the components for them to outsource their IT services. After all, what sense does it make to recreate the same thing in-house when you are under-funded, under-staffed and have better things to focus on? As a provider, if you are able to say “We understand your market. We know these privacy and compliance issues are critical, integration between systems (standardization) is a must and that you struggle to keep up with students thwarting your security controls.” Why wouldn’t an education organization at least hear you out? It not only saves them money, but saves them tons of hassle and headaches and they can offload everything knowing it makes them compliant to whatever controls they need to be compliant with.
Replicate this across other verticals using the same story and your value proposition has gone from “We sell everything to make you more productive/secure/innovative” to “here is what we do to secure educational organizations like yourself.” Who doesn’t want someone else to deal with vendors and figuring out the best solutions, selling internally to make sure everyone is onboard to fund the project.
By relying on cloud providers to give them the right technology mix while taking advantage of an OPEX vs CAPEX situation means these organizations can actually focus on their core business instead of wasting time figuring out what everyone else is doing to stay ahead of technology changes.
