Best practise report-writing methodology for service providers of Business Intelligence to SMEs – Part 2

Following on from last weeks article, now it’s over to you to decide how, who, when and at what cost you can deliver the solution and what skills will be required to do it. Listed below are some of the key questions you need to ask and answer before you start making these decisions?

Does all the information required exist in the customer’s source data?

Why is this important? Believe it or not, some customers do request information that just does not exist. This happens because sometimes the people who need the information are not that close to the operations where information is being captured or not captured. A good example of this is a manager who wants to see a profit and loss analysis by projects, only to find out that certain expenses/journals have not been captured against project codes – problem. Be careful to check this out.

If not, can it be derived from the source data?

If your choice of tools allows you to create information on the fly that can be derived from other source data then you’re OK. If not, you need to consider that there may be some form of manual entry/preparation which could challenge the customer’s expectations, especially if total automation was agreed.

Does it come from more than one source?

You need to be sure that the tools you are using cater for this in a way that data integrity is preserved through automation, and that loose Excel spreadsheet use is eliminated. Generally this leads to a more complicated setup and may cost more.

Is the information related/structured or unrelated/unstructured?

If the information is related there is good chance that you can use standard query language to create the relationships and extract the information in one process which is quicker and easier. If this is not the case, you may have to build more than one process to extract the data and then consolidate it using a union process which will mean a more complex structure.

Has the data been consistently captured over the time frame required for output?

This is important and ties back to the first step of the process where the customer’s expectations are defined. A good example of this is when a company changes its processes over a period of time. For example, if the requirements are to track sales trends by rep over a period of 5 years and the way in which sales were processed changed after 3 years, then you have may have to consider how your solution will handle the two types of data. Another example is if a company has run a debtors book using the balance forward method, and they then switch to an open item method where invoices are matched against payments, you will have 2 scenarios to cater for which may cause complications. In light of this, be sure to check the completeness of data before committing to timeframes and solution delivery.

Depending on what tools you are using, are there any relevant solution sets/templates that can help fast track the solution?

If your choice of tools comes with pre-formatted templates that cater for the ERP/accounting/Payroll system that you are getting the data from, be sure to see if any of these can help you fast track the solution delivery as it may save you from reinventing the wheel and assist you to build a customized library of solutions for your customer. If you are in contact with other consultants be sure to check if any of them have ever created a report similar to what you need to develop. Opportunity may exist to negotiate an arrangement that works for all parties. Have a look at the BI Community for access to videos, forums and groups of like-minded business people.

Who needs to get the information, how often, and how are they connected/located?

Creating the solution is one thing, but delivering it to the areas of the business that need it closes the loop on driving value to the customer’s organization. What is the use of having excellent information in a business if it is not delivered to the right people, or if they don’t have access to it? Always identify who needs what, where they are located and the potential security implications of this information. Again, depending on what tools you have, your options should cater for challenges relating to bandwidth constraints, frequency and real-time access to data, delivery to desktops, or access via WEB protocol. One of the ultimate goals of a good Business Intelligence platform is that people in the organization can “help themselves” to relevant information that they can trust. It is critical to free up resources that have historically spent most of their time preparing and sending out ad-hoc unstructured information to the rest of the business, normally the financial and IT departments. So be sure that the tools you choose cater for this.


The temptation is always there to impress your customer quickly, but compliment this attitude with a sound and thought-through methodology, and your customer will respect and trust your services as you build a relationship to add massive value to their organization. There are many ways to skin a proverbial cat, but this is a method that works well. Taking short cuts will unfortunately result in failed projects.