Almost a month ago I attended Teradata’s Third Party Influencers event, and although I’ve been remiss at writing about the event more quickly, I have to admit that it has taken me this long to gather my thoughts. After all, each year I learn about updates in the products, the advancements in technology, as well as hear from customers to see how they apply technology and what that means for their business. And until more recently, there has been a focus on showing the business value associated with technology adoption, but usually within the outlook of EDW (enterprise data warehouse) adoption as best approach. Whether or not a reality for most was always a consideration, but never appeared to be a main focus.

As the market shifted, some of this outlook changed as well, but not as much as it has since the acquisition of Aster. In addition to a focus on “big data”, expansion to the cloud, and a broader focus on services, there is a new focus on applying a Unified Data Architecture (UDA). In some ways this is similar in terms of using an integration data warehouse that is shared among many business users. The difference is that technology can now support broader use cases and more diversity, meaning that Teradata can now fit multiple business scenarios without having to conform to an EDW outlook.

Obviously, this hasn’t been the case with the appliance family, but the reality was that there was still a focus on building out an EDW eventually or based on a best case scenario. Now, things are different. 

Big data leads the way to more flexible analytics and lets organizations look at a broader range of data in more varied ways, making analytics use able to address business challenges that may have been out of reach in the past due to data warehouse limitations. Based on Manan Goel’s presentation about the Aster Discovery Platform, the solution is meant to do just that by addressing the following challenges of traditional data warehousing:

1. long time to value
2. lack of flexibility
3. requirement of specialized skill sets
4. perception of high risk and high cost initiatives

Aster Discovery Platform aims to provide rapid exploration, provide access to all types of data, provide access to all types of users, and provide a proven discovery methodology. One of the key ways Aster enables easier access is by using pre-made models through a Library of SQL Hadoop to provide single SQL statements addressing different data types. Obviously, this will never be within the realm of the business user, but does provide promise for broader use – especially within the small and mid-sized market place. 

One of the value propositions of any solution is the ability to implement a solution without having to worry about time to value or hiring new resources. Hopefully this expansion of Teradata Aster will enable solutions to be more broadly adopted within organizations, making large scale data warehousing more accessible to SMBs.

A couple of weeks I was having a conversation with a group at Enterprise Data World in San Diego about self-service BI. The consensus seemed to be that although businesses are jumping on board and implementing some form of self-service, on a broad scale, the intensions of self-service are not being met (or at least not with this particular crowd). Organizations are constantly trying to take better advantage of their BI investments and one way to do so is to take advantage of industry trends – one of which being the concepts surrounding self-service BI. Self-service BI essentially means better and easier access to analytics and data – in essence empowering the user to interact with BI in an autonomous way to explore business and data-related challenges. 

The ability to provide self-service BI has required many vendors to think about providing their solutions in a different way. Moving away from traditional models whereby only very data oriented users could get value out of BI interactivity, towards an outlook that aims to provide these capabilities for the masses. And in theory, this is great! After all, business decision makers require tools to explore what is happening, what opportunities exist, and where things might be going wrong. The reality for many, however, is that the true promise of self-service doesn’t fully exist.

Many vendors base their self-service solutions on previous iterations of their solutions, making it easy for those well-versed in data and statistics to interact fully with solutions, but leaving others in the dark. Other vendors provide access to everyone, but may not have the built-in processes necessary to ensure that data accessed provides the value or governance required to make accurate decisions. Either way, it seems as if somewhere along the way, self-service is falling short and isn’t there yet. This particular group of individuals from a variety of organizations all stated that while they had implemented some measure of self-service, it’s use was no different than more traditional forms of BI. Their reasoning included a lack of ease of use and the requirement for training. 

Odd…all of the efforts placed on making things easier, when the reality remains (in most cases, but not all) that the value of self-service has been making it easier for super users and experienced BI users to expand their BI use and do so independently. And although there are solutions that do enable some free form analysis, the easier these solutions are to access, the simpler the functions. In essence, the promise of BI for all with an easy to use interface and intuitive analytics access is premature. Potential on a broad industry wide level exists, but it seems naive to assume that most organizations are there already or have the processes in place to make this happen.

Unfortunately, the reality remains that self-service BI may be easy to use but only if previous familiarity with the tools being used exists. Yes, more interactive and “socially” designed solutions exist, but whether people are actually getting the value out if it originally expected remains to be seen. Hopefully as self-service BI becomes more prevalent through adoption, organizations will have more freedom to design the types of solutions that meet their individual users’ needs.

This post was written as part of the IBM for Midsize Business program, which provides midsize businesses with the tools, expertise and solutions they need to become engines of a smarter planet. I’ve been compensated to contribute to this program, but the opinions expressed in this post are my own and don’t necessarily represent IBM’s positions, strategies or opinions.

Many organizations are struggling with how to get the best out of their BI investments. Part of this means moving from historical reporting and analytics to BI use that takes advantage of operational data feeds to help provide a current view of the organization. In some cases, companies have been doing this for awhile, while in other cases, businesses struggle with the best way to do this. Either way, the reality is that the market is shifting. Historical BI may still be valid, but no longer provides the strategic advantage that it once did, unless it is complemented with a view of the present and a way to predict into the future. 

A look at the operational intelligence revolution

Ok, so the term revolution might be taking it a little far, but organizations are trying to find value in their IT investments, meaning that they are searching for ways to change how they are leveraging their technology – and to do that well, it might require a revolution of sorts by changing not only the technology itself, but also the way people work and interact with information assets. And the reason using revolution might be a good term is because it’s important to know that it isn’t an easy shift. It requires:

• Technology that supports near real-time data feeds and interactive development of whatever types of delivery methods are chosen
• Support from stakeholders and project sponsors to understand the importance and push required changes to the next level
• Agility in the form of infrastructure, but also in terms of people
• Process evaluation and change management
• The ability to adapt and be future oriented
• A focus on continual improvement and people who are dedicated to this

All of these areas are interrelated, meaning that to really move towards a successful agile environment, both technology related and people related activities need to collaborate together to make it happen.

The value of operational analytics

As mentioned, many organizations are struggling to get value out of their business intelligence investments. Adding to this, general competition is increasing due to the way in which people interact with organizations. Social media, online reviews, new styles of marketing, etc. give companies the opportunity to develop closer relationships with their customers, but also increase risks for churn because of broader competition. Consequently, companies require insights that are near immediate to identify what is happening – how are marketing campaigns being received, what is happening within call centers and why, and is the supply chain being managed effectively. Taking this further, there are implications for specific vertical markets. For instance, telecommunications and service outage identification, healthcare insurance fraud identification, general risk mitigation, and the list goes on. 

Overall, it’s important for organizations to note that even though the road to broader agility might not be easy – the market is demanding it now. Yes, it’s possible to work with traditional BI models, but to really get the most out of information assets and to leverage technology and move ahead of competitors, it will be essential to start looking at operational analytics and broader interactivity and agility within the organization at large.

This post was written as part of the IBM for Midsize Business program, which provides midsize businesses with the tools, expertise and solutions they need to become engines of a smarter planet. I’ve been compensated to contribute to this program, but the opinions expressed in this post are my own and don’t necessarily represent IBM’s positions, strategies or opinions.

The BI market place is flooded with messaging from vendors saying they can provide analytical platforms and access to information faster, better, easier, and cheaper. This is true in many ways due to the shifts in technology:

• storage is more cost effective
• processing speeds are quicker
• operational intelligence and access to information in near real-time is available more broadly
• self-service models make it easy for many different types of users to interact with solutions
• time to value is occurring more quickly due to quicker implementation times on a broad level

But even with all of these shifts and advancements in technology, many organizations are still left out in the cold because of the fact that it is difficult to sift through all of the solutions available to decipher where real value will lie. After all, many of the solutions available in the market place will meet the needs of many organizations depending on what the intended goal is. Identifying real value, however is something different to different businesses that depends largely on scope, goals, and expectations. 

When looking at the time to value specifically, the following should be noted:

1. Realistic expectations should be set related to initial implementation times. These will differ based on new implementation or upgrade, technology used, complexity of data, and development of business rules and delivery platform.
2. Many factors require consideration when implementing a solution that may affect timelines.
3. Some solutions will require an iterative approach, meaning that value will increase over time. Businesses need to identify what is realistic and what they can accept.
4. The level of value will differ based on the targeted audience. 
5. The meaning of time to value and value itself needs to be identified as it will differ based on stakeholder. For instance, does time to value translate to implementation times? Or does it rely on goals set to save costs or increase profits?

The reality is that there is no single definition identifying what “time to value” means within the market. What this translates to for companies evaluating solutions is that much of what they hear will relate to implementation times and not how that translates in terms of time to the iterations required to get BI right and provide a framework for overall value. The value being actual results, whether they be the ability to lessen wasteful spending by targeting customer needs better, lowering customer churn rates, identifying issues before they become problems, or increasing profit margins. Therefore, it stands to reason that organizations require the education and tools to develop their own expectations surrounding time to value. It will always be possible for vendors to estimate the implementation of various solution components, but they will not be able to identify how BI will be applied, what business questions will be asked over time, or how decision makers will leverage their information assets to improve overall efficiencies. This remains the realm of BI stakeholders and those in charge of asking the right questions and delving deeper into the information at hand. 

For many organizations, the goal is simply to get a dashboard or set of analytics up and running, thinking that the value they achieve will come naturally. The truth is a bit different. A solution can only go so far without getting into the hands of the right people. The right people asking quantifiable questions to get to the heart of business challenges are what leads to true time to value. After all, technology is meant to support our business operations and not make the decisions for us.

This post was written as part of the IBM for Midsize Business program, which provides midsize businesses with the tools, expertise and solutions they need to become engines of a smarter planet. I’ve been compensated to contribute to this program, but the opinions expressed in this post are my own and don’t necessarily represent IBM’s positions, strategies or opinions.

A couple of weeks ago I attended Predictive Analytics World (#PAWCON), a multi-city conference focusing on how organizations are leveraging predictive analytics within their organizations. Much of the discussion surrounded “big data” and how to leverage analytics within environments that contain large, complex, and real-time related data. I really enjoined the conference because of the fact that most of the attendees were trying bridge the roles of technology and analytics with business insights through predictive modelling and most of the speakers made a point of speaking about the ways to use data to leverage business insights. For instance:

• Jane Griffin from Deloitte talked about the ability to move beyond traditional software towards a more agile environment and understanding that managing big data requires the right people with the right skill sets to develop statistical realities while coupling them with the ability to ask the right questions
• Emma Warrillow from the Data Insight Group and a marketing strategist, discussed churn modelling as applied within predictive models and the complexities that arise based on the variables of customer behavior through churn and general segmentation
• Johan Forman from MailChimp looked at leveraging data to achieve a singular goal – achieving compliance and mitigating risk. This involved shutting down bad users and developing a strategy to identify what that meant and how to identify what that means specifically and who qualifies.

All of these areas relate specifically to people – customers, decision makers, data scientists, etc. The reality is that the use of predictive analytics requires the ability to translate the breadth and depth of existing data into valuable insights in a way that extends beyond simple reporting or dashboarding applications. One of the areas that was very obvious is that the newer iterations of BI are touting self-service and data discovery in a way that is easy for business users to interact with. The reality for predictive models, however, is that statisticians, data scientists, and those willing to ask complex questions and delve deeply into an organization’s information structure are the types of resources required to develop and maintain a successful predictive analytics strategy.

As more organizations mature within their BI use, the push towards predictive analytics will become commonplace. Businesses need to understand the requirements to expand into this area and make sure that they have the relevant resources to support these initiatives. After all, built-in or out-of-the-box analytics will only help an organization go so far. Without the proper skill sets which include a mix of business and technical, it becomes quite difficult to build a strategy that supports a strong predictive analytics environment.

Here is a listing of the next set of Predictive Analytics World conferences in the United States – I think it’s a worthwhile conference to check out:

• April – San Francisco
• June – Chicago
• September – Washington, DC (Government)
• September – October – Boston