Getting Your Priorities Straight: Five MUCH Better Ways to Deploy BI

You did all your research on BI vendors, and you’ve compared what they offer against your key objectives. You’re looking for self-service reporting and analysis, users working independently of IT, analytics aligned with strategic goals and being able to use a performance management application across your entire organization.
You’ve finally made the leap into the world of BI. Your organization is on board and they’re sold on the benefits the industry has been preaching forever.

That doesn’t seem like too much to ask for—all of this sounds pretty standard, right?

So where do we go from here?

This article addresses a few of the key priorities your organization should have on its radar as you approach your BI deployment. It should help you internalize a few crucial lessons in a way that’ll garner maximum returns and minimum risks.

Understanding the circumstances

The definition of business intelligence seems relatively simple. In its basic form, BI entails gathering transactional data from one or multiple sources and using it to provide a model that business people can reference to obtain a better understanding of their operations. However, this may also require a range of highly skilled technical resources, a deep understanding of the source data, an understanding of the business requirements, and most importantly, an ability to perform this in a timeframe that’s acceptable to the business and its major stakeholders.

Organizations that have implemented or are considering implementing a business intelligence solution to support their Dynamics implementation experience a number of challenges, some of which are specific to the Dynamics application, while others are generic to any BI project. I’ve narrowed them down to five key areas that need to be prioritized and explained:

  1. Complexity: How to understand the underlying data model and the configuration of the system.
  2. Extensibility: How to keep your business intelligence solution in sync with your business system considering ongoing customizations, add-on solutions and other data sources.
  3. Lifecycle: How to deliver the business benefits in an acceptable timeframe and continue to do so on a regular basis.
  4. Adoption: How to ensure that what you deliver is used within the business.
  5. Skills: How to remove the resource bottleneck to manage and maintain your business intelligence environment.

Extra class name

Overarching all of this, we must also consider elements of data governance that are intimately linked to each of the mission-critical priorities above. We can’t ignore data quality (ensuring the right versions of data being deployed to the user community) or data security and compliance, which is comprised of protection for regulatory reasons, and the management of the appropriate levels of data access.

Now let’s look at each priority individually in hopes of drawing a few lessons during the BI planning process.


What we’re plainly dealing with is a relatively complex base model for a standard ERP or CRM system in the Dynamics family. With each new release of the application, new functionality is also added. The existing data model is extended and enhanced to make processing even more efficient and responsive. This key issue of is how to understand this complete data model and identify the specific data source to support the reporting, analytics and BI needs of the business.

Separate to the database model, each Dynamics application has a core component that defines and manages configuration details such as security, table relationship and the customization that have been made to the standard system. The challenge is therefore not just limited to understanding the data model, but also the ability to connect to the metadata or configuration layer to enable you to fully understand the design of the system.

To address this issue of data complexity, you should be seeking a deeply integrated solution that understands the database model and the configuration details of each of the different Dynamics applications and versions of those applications.

Many BI projects will struggle when they’re not able to provide an integrated solution that includes a pre-built data warehouse or analysis model that understands the underlying business system. You should be thinking about the ability to merge multiple instances and multiple databases as they’re deployed for performance, geographical and/or for business reasons. This will create safeguards as the complexity of the model continues to evolve.


One of the great advantages behind deploying Dynamics is the ability to modify the system to support your own business processes and unique requirements. From my experience, there’s usually a large degree of customization that’s made to any standard Dynamics model. In addition, the Microsoft Partner community offers a large range of third-party add-on solutions specific to vertical sectors or to a specific business function. It can be a payroll system, a complex pricing model or an HR system.

The point is that we are very rarely dealing exclusively with Dynamics data. 90% of the customers I’ve interacted with have data sources (both internal and external) that need to be added to the Dynamics model. This is essential to build a complete model to the business community.

Therefore, you need a solution that allows for automatic data discovery and profiles the implemented instance of Dynamics ERP or CRM. The solution must be able to identify these extensions and modifications and easily extend the data warehouse and analysis model

It also needs to be done in an agile way. Your business is constantly changing as new applications are being added, and, at that point in time, your underlying data model, which is supporting your end-user community for BI reporting, needs to be extended quickly as well. There should always be the option to accommodate other disparate data sources.

You ultimately need to be able to augment Dynamics with data from other Excel spreadsheets, Oracle databases, Salesforce, Azure and other data sources which may or may not reside within your own data center.

An optimized solution will effectively need to embrace a process that enables you to build, modify and enhance the underlying data model that the business will use for BI and reporting. This should be done using a process that firstly provides you with a prebuilt data warehouse, a prebuilt model and—probably most importantly—a position where the data model has been greatly simplified.


Let’s be honest: BI projects have a reputation for taking too long to show a positive ROI. Historically, these projects take some 24 months before meaningful analysis begins. This is simply too long since projects can quickly lose momentum and the original business objectives can change significantly over this time.

Understanding your timelines, project management and the time to demonstrate value are key factors in a successful BI project. I’ve seen many cases where an organization is unable to demonstrate visible results and value to business users in the first 8-10 weeks of a BI project. The business users then simply find another way to do access the information they require.

If we just take a minute to consider the waterfall approach to a BI project (which we unfortunately still see a lot of), we see that the timeline to deliver the requirements to the business can be extensive. We start with defining the user requirements by picking at the brains of those using the reporting subject, then we move onto the technical build component and prepare a development specification based upon those requirements.

Defining every last specification and building the analytics from scratch takes time, energy and money. Left unmanaged, every step in the waterfall approach to implementation can often bring entire projects to a halt, leaving project managers scrambling for cover, sponsors looking for remedies and budgets wiped out.

Apart the time taken to deliver value here, there are some items inherent in the process which should be monitored. Often projects can become very bogged down due to constant change requests because requirements are changing over this development period’s time.

Also, I’ve seen the user community becoming alienated due to the lack of involvement and the time to see the deliverables.

You can minimize the long stretch in the project lifecycle by employing a solution where the development specifications, the understanding of the data lineage and the design of a starting data warehouse model have already been prepared. The process is simply a matter of an initial build that allows you to connect to your Dynamics system (regardless of what application and version it is) and to populate a standard model that clearly identifies the customizations and maps it to pre-built reporting templates that surfaces that data up so that the user community is actually able to see it quickly.


In relation to user adoption, research from TWDI and BIScorecard suggests that BI deployments roughly experience a 25% adoption rate. The key contributing factors include the poor performance of the system itself, un-compelling user interfaces, lack of mobile accessibility, and the existence of multiple tools.

Many BI solutions ask the user community to use one tool for financial analysis, another tool for reporting and another for building dashboards. The fewer tools that are involved in the deployment of a BI strategy, the higher likelihood for the success of the overall project.

Making user adoption a priority truly necessitates a single application for the development and consumption of reports and visualization components, including pre-built templates that are useful in demonstrating capability and accelerating the implementation.

Without this, your users might feel bogged down by a system rich in complexity but poor in getting them answers to their critical business question.


The traditional approach relies heavily upon a multitude of highly technical skills. Without a solution that understands the underlying ERP or CRM data model, relational and multi-dimensional data experts are required to simply understand and map the source data required. Experts in data warehouse design and developers with the knowledge and skills to write the code to extract, transform and load the data into the new model are also needed. At the front end, the business needs skilled resources for each individual tool being used for the design and development of the reports and visualization components.

In additional to the development skills, the end user community also needs to possess the skills required to user the tools deployed. Most organizations don’t possess this range of skills in-house and therefore rely upon external consultants and contractors.

Your BI solution should reduce the reliance on this set of highly technical resources. The solution should support rapid and agile development that simplifies the data modeling, the ETL development, the design of the data warehouse AND the report development without the need for highly technical resources.


As seen from the factors that we’re prioritizing above, BI projects can be challenging and there are many different facets to consider in order to streamline a BI deployment.

The main areas that need to be held as priorities are:

  1. Complexity: understanding the underlying data model and the architecture of the system.
  2. Extensibility: how do you keep your business intelligence solution in sync despite customizations and extensions?
  3. Project lifecycle: the ability for the project to demonstrate benefits quickly and frequently.
  4. Adoption: what happens when you get the BI content into the hands of your users?
  5. Skills: the use of multiple tools and technologies required to implement and maintain the BI solution, and the skills required to support them.

In the most fundamental sense, an organization can enjoy higher success rates with a deeply integrated solution that understands the database model and the configuration details of each of the different Dynamics applications. That’s really the common denominator.

It should be able to accommodate the extensibility of the Dynamics database along with other sources of data, demonstrate a short time to value and provide trusted pre-built analytics to ensure user adoption.

Comments are closed.