As Information Builders emerges from transition to new ownership models and new leadership, the New York-based provider of data and analytics solutions has adopted a ‘steady as she goes’ approach in efforts to build new onto older roots. Never the easiest course to chart, this high wire act is especially challenging in the rapid pace world of information technology, which dangles a carrot before those who seize the helm and the heave ho to those who hesitate. But at its customer Summit event in Orlando last week, Information Builders (IB) demonstrated the art of the possible – an updated version of the company’s traditional vision that features pragmatic capacity building as opposed to flash modernization.
At Summit’s opening keynote session, Information Builders CEO Frank Vella described an “IB ecosystem” consisting of “loyalists” or legacy customers who built on the company’s original FOCUS platform, new customers, many of whom have deployed on the IB Cloud, which launched only a year ago through partnership with AWS, and partners, who are receiving more attention and investment as IB begins to refine its go-to-market strategy. Speaking to the needs of these disparate groups, Vella acknowledged that there’s “lots of change going on at IB,” while affirming that the leadership understands what needs to stay the same: the “customer-centric approach is what everyone seems to value, this is in our DNA, it’s in our culture – if our customers succeed; we succeed,” he explained.
At a product level, this prudent stance has translated to innovation aimed at easing the use of IB’s platform with an overall goal of driving adoption – as opposed to technical wizardry for its own sake. Citing a BARC analyst survey, Vella noted that analytics products are used by only 17% of employees in most organizations: “the tools are hitting the walls,” he added. And though IB’s record has been better than this norm, addressing loyalist customers’ requests for improved look and feel has animated much of the company’s development focus over the past year. As example, at the event IB announced several updates to WebFOCUS Designer, its BI and data visualization tool for creating and sharing content. The product now features a more intuitive interface to streamline users’ experience when building and deploying analytics dashboards, as well as faster, and more integrated search capabilities to enable users – including those who may not have extensive training or knowledge of deep data manipulation – to quickly generate insights from analytics.
Product announcements targeted at easing adoption aligned around two themes at Summit. User need for ‘simple’ was address through update of the IB platform’s interface, but also through new integrations and capabilities designed to enable customers to orchestrate use of IB tools and advanced applications. A new integration with Kafka, for example, an open-source stream-processing software platform allows customers to access and analyze real time IoT data streams, build dashboards and alerts that respond to immediate needs, automate decisions and initiate downstream operational actions to speed and improve performance. Similarly, Information Builders’ BI and analytics technology has been containerized to support deployment to container platforms such as Docker, and to enable container management through Kubernetes – with a goal of easing management of future architectural changes such as a shift to the cloud or other platform.
According to Jake Freivald, VP of product marketing for Information Builders, these innovations also address customer need for ‘scale’, a second key theme in Summit product announcements. Support for scale is perhaps most evident in the IB Cloud offering, introduced in May 2018 with IB participation as an Advanced Technology Partner in the AWS Partner Network, and launch of on-demand delivery of fully-managed BI, data integrity and integration services. The IB Cloud managed service is also about ‘simple’ though: “that’s what this model is all about…. It’s about saying that you don’t want to have to manage the iron, you don’t have to manage the BI environment and you don’t have to manage the data. It’s about saying you want to do WebFOCUS in the cloud, and then effectively, and metaphorically, swiping a credit card to have WebFOCUS running. And when there are upgrades that have to be done, they just happen. That level of simplicity is what we are looking for,” Freivald explained. “And because we are managing this from soup to nuts, we are able to make it as simple as possible for the customer.”
To support the success of cloud delivery, Freivald explained that IB must ensure “quality is super high,” and that upgrades, which roll out frequently with managed services, are fully engineered. “You also need to make sure that you have the right scale for the cloud,” he added, to allow customers, with the help of containerization in some cases, to take advantage of the cloud service elasticity to address peak resource demand needs, or to create a temporary, cloud-based sandbox for experimentation.
Going forward, IB intends to offer different levels of its managed service on AWS – platinum, gold and silver, which will support different numbers of users with varying levels of service, including customization. And to reach many enterprise clients, it plans to work towards deployment on Microsoft’s Azure platform as well. According to Freivald, a cloud-based managed services strategy will remain a strong focus for the company as it delivers customer success: “[with a cloud-based managed service] the success rate is higher. People can get their projects off the ground faster and see returns faster. So they are happier, and our margins are good. It’s a combination of better success and ROI for customers, greater simplicity, good margins, and better awareness and outreach with AWS.”
In its discussion of scale, however, Information Builders moves beyond issues of peak resource demand and data volumes. Scale is about enabling IB partners – who can take advantage of cloud and containerization capabilities to quickly create, replicate and deploy containerized applications featuring IB and their IP. But in the IB schema, scale is primarily about helping customers extend business intelligence to different worker roles throughout the organization. Freivald explained: “Scaling [a BI application] lies in deploying the model so that a non data scientist can present the data set, can select the model as a value in a report, and then take this a step further by entering the value in a dashboard for another employee – a customer service rep, in a marketing data example.” This approach can feed 6,000 people in the organization, as opposed to 6 people in marketing, or one person who is the data scientist, he added. “Scale lies in the curation and deployment of models” that extend beyond top level management.
The range of workers who can be given access to improve data value was described in the keynote by IB CMO Michael Corcoran: in operationalizing business intelligence, he explained, “the needs of the many outweigh the needs of the few.” While executive management uses dashboards to support high level decision making and analysts/data scientists do data discovery and advanced analytics to generate insights, to operationalize analytics and achieve data monetization, non technical employees must be provided with decision support applications, metrics, alerts, and IoT data to drive culture, efficiency and performance, business partners can use decision support portals and performance metrics to optimize supply and service chains, and customers/citizens can be engaged in customer portals, sentiment analysis and Infographics to drive revenue, loyalty and organizational differentiation. “Scale is not only about how many terabytes you are running in a technical sense: we are also talking about business scale which relates to the number of users, the number of use cases a model might be useful for, the volume of queries, and the volume of data – as in a sensor-based, data streaming application,” Freivald concluded.
IB vision of the future
Discussing this vastly more complex scale value proposition, Vella argued it’s critical that Information Builders respond to the changing relationship between data and the organization. Traditionally, in the “one to a few” model for distribution of information that characterized most businesses, risks associated with incomplete data were mitigated by limited usage and management’s gut instinct. Today, data is everywhere, and the relationship is now between data from an infinite number of sources, to many or all employees, and extends out to all partners and devices. In this complex scenario, ML and AI come into play, limiting data sets, and making hundreds of automated decisions a minute to provide competitive differentiation for adopting organizations.
In building capabilities to address what Vella called IB’s “vision of the future,” the company is now in the process of conducting due diligence, investigating customer need and leveraging expert resources to better understand how AI can contribute to IB’s platform development and how the company can support users who may be looking to take advantage of the technology in their own environments. Freivald noted: “There are a lot of pieces in AI that are coming together right now. But one of the biggest challenges is to have the right customers looking at the right use cases who can say this is where AI is really going to make a difference for us. Until we have that, it’s difficult to state that we are going to do AI in this particular way.” To help, Information Builders has launched ‘Voice of the Customer’ research led by VP of field technical strategy and engineering Kabir Choudry. This research is designed to streamline product engineering by articulating common AI use cases that would serve as good candidates for productization, and feeding this market intelligence in a structured way to product teams.
Similarly, IB has engaged data scientist Aditya Sriram, a PhD engineering candidate and member of the KIMIA Lab at the University of Waterloo, to provide vision to product engineers on what the market is looking for, and to help IB orient its portfolio around an AI vision. Sriram sees three potential directions that IB can follow to normalize AI and improve access to the technology: decompose the level of complexity for prescriptive analytics, creating a platform that allows users to easily deploy ML models without needing to write or understand code; leverage recommendation engines to apply AI to data quality challenges (automate data cleansing); and explore NLP potential in the provisioning of content based on search requests initiated with natural language. Sriram’s second direction is closest to realization. Freivald noted that there are multiple areas where AI could be employed effectively today. For example, IB has already used a form of AI to automate content generation – a near term roadmap involves watching interactions between the content consumer and the content that is created, and applying machine learning to this interaction so that the system can learn and surface content that is personalized to the user. Another near-term application might involve the monitoring of data remediation in IB’s Omni-Gen solution: based on corrections are observed, the machine could develop a recommendation, or rule for data cleansing that would be reviewed by the business user and then implemented for automated response.
Sriram sees his individual role as helping customers understand how to embed AI in their workflows, and in the creation of pathways that allow customers to use AI to bring value into their organizations. In articulating the use cases that can guide productization, though, Sriram recognizes that the application of AI must address business challenges, including issues around bias, transparency, accountability and accuracy, in order to establish data trust, which serves as the basis for driving user adoption. To achieve this, Sriram is focused on a concept that he calls “explainable AI.” “Traditionally, AI was a black box,” he noted, but it is possible to help people understand how features or patterns in the data are extracted and then visualized. Sriram described his work with clients as “a process that a team, led by the data scientist, engages in,” which relies on prescriptive analytics to confirm that the machine learning model is healthy before the algorithm is deployed to a dashboard. Another tactic he advises is to “start small in a PoC with a defined sample” to establish the accuracy of the model and then scale. As new data is added to the network, it can be fed back to the ML system in a 'learn-as-you-go' approach that fine tunes the algorithm, improving accuracy, and thus validating the original assumptions and model.
According to Sriram, algorithms do not necessarily provide truth, rather they explain accuracy levels: “AI is complementary. It’s not a fact; it’s a likelihood. It’s an association of likelihood on your predictions.” As a result, it must be implemented with thresholds “but as you scale, and as the machine learns, the model begins to prove itself though outcomes in real time, allowing the adopter to begin to move from a predictive to a more factual representation of the data. It’s is a process,” he explained.
How to guidance supports traditional customer solutions focus
If this pragmatic ‘how to’ approach is helping to inform the development of future forward platform capabilities, it is also a hallmark of IB’s legacy relationship with customers. As did his predecessor and IB co-founder Gerry Cohen in previous keynote presentations, Frank Vella described customer and analyst feedback on IB performance, noting that in industry awards, “IB scores disproportionately high,” due to the fact that IB is “not a product,” rather it “builds solutions with the customer.” So while Summit featured multiple labs and presentations aimed at tech skills development, presenters such as Michael Corcoran also focused at a high level on ‘how to’ align strategy and execution to drive business opportunity. According to Corcoran, in implementation of BI and analytics, data quality management, and master data management are key to establishing data trust, but since this is difficult to execute on, IB has built automation into its platform, prepackaging MDM and AI in healthcare and insurance bundles (law enforcement is next). “When you bring all these together, that’s when you transform the business,” he explained.
In a separate session focused on helping customers build out analytics solutions, Data Challenges Along the Analytics Journey co-presented with IB director of market intelligence Lyndsey Wise, Corcoran offered a framework designed to guide data strategy. This framework listed the four, commonly recognized levels of analytics, mapping different goals and assets customers need to have in place from a data perspective to each level. Corcoran’s ‘journey’ is as follows.
- Descriptive analytics – customers need reporting dashboards, a data warehouse, and may engage in analytics initiatives. “Most of us have been doing this for decades,” Corcoran added.
- Diagnostic – answering the question ‘why did something happen’, this is achieved though tools for drill down, data mining, data discovery, and the visualization of correlations. In many enterprises, Corcoran explained, “this conversation turned towards data lakes,” which proved problematic as lakes were more swamp than a system of record. To achieve trusted insights and operational BI, customers needed real time data, data governance, and master data management.
- Predictive – when diagnostic management requirements are in place, its possible to move to automation, including algorithms, ML, embedded analytics, IoT anything – and streaming data in edge applications where knowledge of outliers in real time becomes important to building thresholds and alert notification systems.
- Prescriptive – this form of analytics not only identifies what will happen in future, it also prescribes activity, through data scoring, content scoring, which can be built into recommendation engines and control functionality.
While this is not the only journey, Corcoran believes it is an optimal one as “the more quickly you can achieve data quality, and master data management, the more quickly all the journey will fall into place.”
In her presentation, Wise also focused on helping customers to “think strategically” in order to “get value out of analytics activity, but also to cope with the organization’s change management needs.” The Wise view of strategy is a set of policies designed to realize a vision, which is achieved through goals, planning and tactics that recognize the importance of both data and analytics operating in a “closed loop.” This cycle begins with data, and moves through analytics, information, insight, and action, aligning inputs and outputs to create value. A key requirement, in her view, is to organize the company so that technology, process and people work together – and are managed to ensure data is suitable for its purpose, and so that people can leverage the data to solve business problems. To support data and analytics strategy, the right stakeholders must be involved, including those with specialized skill sets, project sponsors/financial decision makers, subject matter experts, external perspectives, customers, partners, suppliers, technical expertise; the right structure must be in place; and change management must be effected. According to Wise, the “5 Key Components to Change” are “align data and business requirements, identify data assets/perform gap analysis, involve stakeholders, create cohesive strategy that crosses IT and business units, and enable people.
Aimed at IB customers, this high-level guidance reflects the company’s ongoing commitment to helping users define or refine the implementation of trusted BI and analytics solutions. It may also inform IB’s own strategy as the company looks to blend old and new – advancing its solutions orientation with more structured process for standardized product development, using consultation, collaboration and change management to ensure people, process and technology continue to align with customer needs.