How will businesses compete in the race to gain competitive advantage from the ever-increasing volume of internal and external data to which they now have access? One option is to build sophisticated in-house analytics toolkits, which could potentially generate a wide array of insights from such information. But an increasingly useful alternative will be to gradually buy those tools – and the underlying processing capability and data experts – when they are needed.
This is an approach to consider particularly because different types of business problems require different analytics techniques, talent and fluctuating levels of support for project execution. Also many of the most valuable insights awaiting discovery are not currently imaginable. As with many areas of innovation, the potential for analytics is huge, but new tools and capabilities are constantly being invented and developed.
All this makes analytics as a service an attractive solution for many firms, as they struggle to exploit the potential within data they already possess or can now access. Structured as a holistic solution hosted on a cloud platform that can be shared across the entire business, AaaS is delivered on a subscription basis by suppliers. This can help to deploy advanced analytics capabilities – technology and talent – throughout the company much more rapidly than the business could do itself.
The benefit AaaS delivers is agility and scalability. This is valuable, given that modern analytics is now effectively replacing the management reporting processes designed for an age of limited data that changed slowly. By contrast, executives now want to work with dynamic data that reflects real-time reality.
AaaS can not only deal with that dynamic data but also adapt quickly to solve new problems as they emerge. The service may begin with one outcome in mind, but it can evolve to meet new challenges with more data and experience with the business and its customers. And as it is scaled across the firm, bringing in more functions, more of those newly found challenges will also be confronted.
New Clarity Ahead
The ultimate aim for AaaS should be to deliver insights on which the business can act. What is the outcome the business desires or the problem it needs to solve? What are the best analytics tools to reach that outcome? And what data will be needed for this? Answer those questions, and the service can deliver insights for specific individuals in the company to act on to achieve the desired outcome.
There are also two types of output created by AaaS: output to humans, such as dashboards and alerts, and output to machines, such as recommendations served up directly to a website when a customer visits it. Due to the breadth of potential AaaS applications offer, the actionable outcomes may be very diverse.
One example is demand forecasting, where AaaS enables retailers to determine what merchandise to stock in each of their stores based on electronic POS data, as well as external market data. One U.S. consumer goods firm is using a service like this to optimize product placement (i.e., which mix of products will work best and which outlets will yield the highest returns).
Additionally, AaaS can help telecommunication companies identify which customers are likely to move to a competitor at the end of their contracts and segment the database to see which customers are most desirable for the company to retain.
Risk-based analytics is yet another possibility. For example, banks are using AaaS to identify which customer segments are likely to default on credit card or mortgage repayments. Insurers are also using such tools to better identify fraud risks.
Choosing an Approach
It is possible, of course, that businesses could develop these analytics solutions for themselves. And some are embedding such solutions within their functional systems and processes as they develop and buy new products and services. But whether businesses are starting from scratch with analytics or building on the tools they already possess, AaaS occupies an increasingly important position in the mix of potential solutions.
Certainly it will be difficult for individual companies to recruit in scale, given the shortages in the analytics talent pool and varying business need for data scientists. The need for such people is often greatest when the solution is being developed but may decrease after adoption.. Also, many companies will not have sufficient work across multiple areas to justify hiring a full-time expert in each of the different scientific disciplines.
However, the case for AaaS is a positive one. Part of the allure is the ongoing innovation that AaaS delivers, as subscribers are able to pool their own data with an ever-expanding range of external data sets and as providers continually develop their libraries of offerings.
Analytics as a service also offers great flexibility in an operational context. Businesses may have very different requirements of the service. Some will want to have their own analytics teams working with the tools provided, taking a close interest in how specific insights are delivered as well as how the business acts upon them. Others may choose to outsource the whole process, with the provider designing, building and operating the system, and even working with the business to implement the actions identified as necessary to deliver the desired outcome.
In practice, most businesses will sit somewhere in the middle, and many companies will want to have enough in-house expertise to lead analytics, even if they do not want to do all of the execution work. It may also be that different executives within the same business will want more or less support, depending on available internal talent. The AaaS model can be designed to deliver that flexibility through varying subscription offers.
These designs will evolve as the AaaS model gains traction and buy-in rates increase. Today, consumer-facing businesses are at the vanguard of AaaS – particularly in retail and financial services, and especially in the U.S. But there has been a dramatic increase in the last two or three years in the range and number of organizations creating related internal roles, such as chief data officer or chief analytics officer. This will further drive adoption of AaaS, especially as more providers emerge and offerings continue to mature.
At times, businesses will need to contend with and manage very specific issues. For example, data security is one consideration currently holding back some financial services firms. For this reason, some AaaS providers are now developing private cloud facilities that sit behind the firm’s own firewalls rather than using the public cloud.
But other challenges will be more universal. For example, some providers are now beginning to develop payment models that tie subscription rates to the actual value produced at outcome stage.
So what will differentiate these providers? With analytics success depending as much on the questions asked as the way the answers are generated, one crucial differentiator will be the ability to offer a holistic service. Providers will need technology skills and modeling abilities, but also a deep understanding of the industries in which they’re working.
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access