Skip to content
Advertisements

Archive for

Why IT Projects Fail

There are many reasons why IT integration projects fail.  From the lack of senior management support to imprecise, inaccurate goals, IT integration projects fail more often than they have to. Based on consulting I’ve done with system integrators, distribution providers, financial services firms, logistics providers and manufacturers, five core lessons emerge.  One of the most innovative companies taking on these challenges is enosiX, whose customer wins at Yeti Coolers, Vera Bradley, BUNN and others provide a glimpse into the future of real-time integration.

  • Middleware forces IT integration projects to focus only on moving data instead of improving business processes.
  • Not having a clear idea of the goals the integration needs to attain in the first place.
  • Sacrificing application response times, data accuracy and user experience in never-ending middleware projects.

Five Lessons Learned From IT Integration Failures

The following lessons learned are based on my experiences and work with IT departments, Vice Presidents of Infrastructure, Enterprise Systems, Cloud Platforms, CIOs, and CFOs. The lessons learned from them are helping current and future IT integration projects increase the odds of success.

  1. Selecting middleware or an integration platform not capable of offline, mobile use with the ability to synchronize in real-time once connected. The fastest growing areas of Customer Relationship Management (CRM) are being fueled by the real-time availability of data on mobile devices. In Configure-Price-Quote (CPQ) and Quote-to-Cash (QTC) workflows, tethered and untethered use cases dominate. To be competitive, any company relying on these two strategies to sell must have an integration framework capable of delivering data in real-time that enables quick app response times, higher performance, and a better user experience. IT integration projects that don’t take this requirement into account nearly always fail.
  1. Selecting an integration solution that requires time-consuming, expensive training and has a steep learning curve. When a given middleware, integration technology or framework is too difficult for IT to learn and use, projects fail fast. The middleware landscape is littered with companies whose marketing is covering up products that have non-existent to mediocre documentation and learning materials. One of the primary factors behind Salesforce’s exceptional growth is their commitment to making the user experience on their platform immediately scalable to each application developed and launched on it. Within 30 minutes, sales teams are often up and running with new apps, successfully selling as a result. Integration frameworks that don’t force system users to change how they work are the new gold standard and are driving the market forward.
  1. Using middleware for business process logic integration when it is designed for data only. Attempting to use middleware for business process logic workflows can get complex and costly fast. It’s one of the main reasons IT integration projects don’t deliver results. In reality, the most valuable aspects of any integration project are the business processes and supporting logic that is automated, streamlined and tailored to a businesses’ unique needs, revolutionizing it in the process. This point of failure happens when IT architects push middleware beyond its limits and attempt to do what more streamlined integration frameworks are designed to accomplish. Business process logic is core to the future of any IT integration project. It is surprising that more organizations don’t look for integration frameworks that have this capability designed into the core architecture.
  1. Failing to consider how data transfers can be minimized or eliminated in the planning and deployment of an integration project. The more customer-centric a project, the more the variety and depth of data transfers required for the integration to be complete. Data transfers grow exponentially and can challenge the scale of a middleware platform quickly. The most successful IT integration projects aren’t data transfer-intensive, they are business strategy driven. One of the most effective best practices of integration is not having to move the data at all. Using an SOA-based framework as a means to enable data consumption without having to perform lengthy ETL processes is the future of integration. By definition, middleware relies on a series of tightly-coupled integration points designed to move data asynchronously. In contrast, SOA-based frameworks are designed to enable real-time synchronous communication through the use of loosely-coupled connections that can flex in response to business process requirements.
  1. Failure to plan and anticipate how a change in one cloud platform or enterprise application including those running on Salesforce’s Force.com, a SAP R/3 system and other platforms impact the entire company’s IT stability. The VP of Infrastructure for a globally-based gaming and hospitality chain told me he and his team often are given the challenging task of bringing up new casino and hotel operations offices globally in two weeks. He sends in an advance team to determine how best to integrate with any legacy on-premise systems. The team also works to integrate any unique Salesforce apps that need to be included into the main Salesforce instance at the tab level, and to determine how best to integrate into the SAP R/3 procurement system. System security is the highest priority during the integration pilot and go-live work.  The company has standardized on a series of network adapters and connectors that are designed to shield all traffic across the network. He told me that just one API change in the IT stack supporting their SAP R/3 integration would cause all adapters to quit working, report an error condition and force debugging to the line level.  They learned this during a go-live with a Reno property. Today all changes to middleware are run in a pilot mode in a sandbox first, and the company is looking to get away from middleware entirely as a result.

From the enosiX blog post, Why IT Integration Projects Fail.

Advertisements

Business Intelligence And Analytics In The Cloud, 2017

  • 78% are planning to increase the use of cloud for BI and data management in the next twelve months.
  • 46% of organizations prefer public cloud platforms for cloud BI, analytics and data management deployments.
  • Cloud BI adoption increased in respondent companies from 29% to 43% from 2013 to 2016.
  • Almost half of organizations using cloud BI (46%) use a public cloud for BI and data management compared to less than a third (30%) for hybrid cloud and 24% for private cloud.

These and many other insights are from the BARC Research and Eckerson Group Study, BI and Data Management in the Cloud: Issues and Trends published January 2017 (39 pp., PDF, no opt-in). Business Application Research Center (BARC) is a research and consulting firm that concentrates on enterprise software including business intelligence (BI), analytics and data management. Eckerson Group is a research and consulting firm focused on serving the needs of business intelligence (BI) and analytic leaders in Fortune 2000 organizations worldwide. The study is based on interviews completed in September and October 2016. 370 respondents participated in the survey globally. Given the size of the sample, the results aren’t representative of the global BI and analytics user base. The study’s results provide an interesting glimpse into analytics and BI adoption today, however. For a description of the methodology, please see page 31 of the study.

Key insights from the study include the following:

  • Public cloud is the most preferred deployment platform for cloud BI and analytics, and the larger the organization toe more likely they are using private clouds. 46% of organizations selected public cloud platforms as their preferred infrastructure for supporting their BI, analytics, and data management initiatives in 2016. 30% are relying on a hybrid cloud platform and 24%, private clouds. With public cloud platforms becoming more commonplace in BI and analytics deployments, the need for greater PaaS- and IaaS-level orchestration becomes a priority. The larger the organization, the more likely they are using private clouds (33%). Companies with between 250 to 2,500 employees are the least likely to be using private clouds (16%).

grouped-bi-cloud-platform-graphic

  • Dashboard-based reporting (76%), ad-hoc analysis and exploration (57%) and dashboard authoring (55%) are the top three Cloud BI use cases. Respondents are most interested in adding advanced and predictive analytics (53%), operational planning and forecasting (44%), strategic planning and simulation (44%) in the next year. The following graphic compares primary use cases and planned investments in the next twelve months. SelectHub has created a useful Business Intelligence Tools Comparison here that provides insights into this area.

cloud-bi-use-cases

  • Power users dominate the use of cloud BI and analytics solutions, driving more complex use cases that include ad-hoc analysis (57%) and advanced report and dashboard creation (55%). Casual users are 20% of all cloud BI and analytics, with their most common use being for reporting and dashboards (76%). Customers and suppliers are an emerging group of cloud BI and analytics users as more respondent companies create self-service web-based apps to streamline external reporting.

cloud-bi-power-users

  • Data integration between cloud applications/databases (51%) and providing data warehouses and data marts (50%) are the two most common data management strategies in use to support BI and analytics solutions today. Respondent organizations are using the cloud to integration cloud applications with each other and with on-premises applications (46%).  The study also found that as more organizations move to the cloud, there’s a corresponding need to support hybrid cloud architectures. Cloud-based data warehouses are primarily being built to support net new applications versus existing apps on-premise. Data integration is essential for the ongoing operations of cloud-based and on-premise ERP systems. A useful comparison of ERP systems can be found here.

cloud-data-integration

  • Data integration between on-premises and cloud applications dominates use cases across all company sizes, with 48% of enterprises leading in adoption. Enterprises are also prioritizing providing data warehouses and data marts (48%), the pre-processing of data (38%) and data integration between cloud applications and databases (38%). The smaller a company is the more critical data integration becomes. 63% of small companies with less than 250 employees are prioritizing data integration between cloud applications and databases (63%).

use-cases-of-cloud-management-by-company-size

  • Tools for data exploration (visual discovery) adopted grew the fastest in the last three years, increasing from 20% adoption in 2013 to 49% in 2016. BI tools increased slightly from 55% to 62% and BI servers dropped from 56% to 51%. Approximately one in five respondent organizations (22%) added analytical applications in 2016.

bi-tools-growth

  • The main reasons for adopting cloud BI and analytics differ by size of the company, with cost (57%) being the most important for mid-sized businesses between 250 to 2.5K employees. Consistent with previous studies, small companies’ main reason for adopting cloud BI and analytics include flexibility (46%), reduced maintenance of hardware and software (43%), and cost (38%). Enterprises with more than 2.5K employees are adopting cloud BI and analytics for greater scalability (48%), cost (40%) and reduced maintenance of hardware and software (38%). The following graphic compares the most important reason for adopting cloud BI, analytics and data management by the size of the company.

most-important-reason-for-adopting-cloud-bi-and-data-management

%d bloggers like this: