Using real-life examples, Julien Delvat, SAP Mentor, is going to guide you through your own Business Intelligence (BI) & Analytics journey, helping you answer a simple question: if there was one thing you could do today to improve your organization's analytics level, what would it be?
Last month I attended the SAP Innovation Forum in London. Sessions from Dan Wilde
(“Engaging Everyone, Everywhere”) and Jan van Ansem
(“Becoming a Data-Driven Organization”) triggered numerous conversations along the lines of "Now that we've spent that much effort implementing SAP BW, what's the best way to put it in the hands of our business users and increase our chances of success and adoption?". Obviously, there isn’t a simple cut and dried answer, but the following will steer you to the right answer for your organisation.
The BI Survey ’16 highlighted that, on average, customers score their BI solution at 6.9 out of 10 in “Business Benefits Realised” . In our experience, this failure of BI projects to deliver the anticipated outcomes is the result of the following two main reasons:
- Trying to do too much, too early
- Solving business problems with technology only
BI Maturity Model
The Business Intelligence Maturity Model. Source: SAP
When our consultants are invited for a BI / Analytics roadmap project, they start with the evaluation of the organization's BI maturity. In most cases, we find that business decisions (e.g. product rationalization, customer prioritization, project profitability, expense cuts) are taken either without the support of data analysis or through a complex non-repeatable manual excel-based exercise. Both solutions are not sustainable and this is readily recognized by most organizations, which is why they invest in data warehouse and analytics solutions like SAP BW and BusinessObjects.
Build your roadmap
Using the BI maturity model, organizations can evaluate where they are and where they need to be. However, it's still hard to evaluate how long it will take, how much it will cost and what needs to be put in place. In a recent analytics roadmap for a consumer product organization, we combined several initiatives over a year towards enabling a self-service BI environment.
||Creation of a Center of Excellence (CoE), creating positions for ‘Analytics Champions’ within business units
||Upskilling of these champions
||Definition of a KPI catalog
||Setup of a BW on HANA environment and upgrade of the BusinessObjects
||Alignment with other in-flight initiatives
As is frequently the case, that organization required a sequence of projects to achieve the required outcome. Looking at the analytics holistically ensured that an initial roadmap was built to drive value and ensure a sustainable platform for BI. These factors helped justify a business case and aid in key investment decisions and change management through involving the business early …
Sample High-Level Project Plan for Self-Service BI Deployment.
None of these projects would have individually solved the lack of analytics in the organization, but having built a roadmap helped validate the business case for each project and supported vital investment outcomes.
Lisa Kart, research director at Gartner, states: "a successful advanced analytics strategy is about more than simply acquiring the right tools. It's also important to change mindsets and culture, and to be creative in search of success." 
In the same way, large and long BI projects fail to deliver the expected outcome, especially if the organization doesn't have a culture of analytics. A food and beverage company we met with explained that the introduction of a KPI dashboard contributed significantly to the alignment of every function and business unit.
Achieving such a remarkable outcome is very challenging and has better chances of success when approached with agile techniques:
- Involve end-users in the process (not just super users or IT architects or external consultants)
- Create mockups with low-tech tools like post-it notes or whiteboards
- Start in Excel. If you can't agree in Excel, there's no point in moving to any other technology
- Perform iterative steps and require regular inputs, weekly or bi-weekly
- Frequently push to Quality, daily if possible, or invest in a proper set of data in the development environment
- Start from the output. For instance, simply import an Excel dataset into your reporting tool of choice and create a mockup to get end-user feedback before starting any development in the data warehouse
At the Innovations Forum, several SAP customers that I met were very interested in starting their analytics journey, but felt powerless. Budgets are hard to justify in organizations where large BI projects have failed. Timo Elliott, analytics evangelist at SAP, recommends starting small by enabling end-users to play with their own data like time sheets or vacation allowance. By playing with familiar and small datasets, users feel more confident and this creates a virtual circle that entices more ambitious initiatives.
BI on BI
How about starting with free BI? SAP offers a service called "BI on BI", through which organizations can get an audit of their analytics licenses and potentially reduce or convert their unused capacity. This is also the opportunity to rethink your analytics strategy and get more from your IT investment. Just start the process by contacting your SAP Account Executive.
While evaluating the effort required for a conversion of reports from Excel to Analysis for Office, our experts recently stumbled upon an inconsistency in the same metric between 2 reports:
- In logistics, "shipment volume" represented the quantity of goods leaving the warehouse (and therefore reducing the stock),
- In finance, "shipment volume" represented the quantity of goods reaching the customer (and therefore triggering a request for payment).
In most cases, there was no difference because the truck took less than one day from the warehouse to the customer, but in significant cases, this lag got reached one week, thus generating variances in the numbers and a lack of trust in the system and the reports. Users consequently dropped the data-driven decisions and returned to their gut instinct.
Organizations need a proper set of metrics. The most important factor here is not in the details, for example, the data source or the units of measure, but in clearly defined responsibilities. At Bluefin, we have defined a metrics template that extends the traditional information (name, definition, data source, etc.) with a Data Responsibility section assigning names to the key roles: Responsible, Accountable, Consulted, and Informed. In the above example, the solution was simply to rename the 2 metrics to "logistics shipment volume" and "finance shipment volume" since they served a different purpose.
Pick your weapon
My favorite task at the start of an analytics roadmap exercise is the listing of analytics capabilities and matching it with reality. Understandably, IT teams within organizations usually have clear guidelines about which software can be installed. Software vendors, on the other side, have become very creative to fly under the radar: free tiers (no IT budget), cloud-based (no need to install), Microsoft Office add-ons (extending already installed software), etc. Anything is possible to help solve end-user pain points and get their foot in the door. The result is a hodgepodge of solutions with pockets of champions trying to get their favorite to become the standard.
I believe that this is the result of IT organizations having limited capacity, bandwidth, budget, and knowledge to meet a demanding business community. There is also usually an unclear separation of duties between IT and the Business when it comes to analytics.
By enabling self-service BI with the right tools and by creating a culture of analytics within the organization, the focus can shift away from IT (in charge of software selection & licenses, backend systems, data warehouse datasets, and security) and towards Business Users (in charge of metrics definition, end-user enablement and deployment of reports and dashboards). To do so, IT leaders should focus on selecting and defining clear guidelines around analytics software. For instance, with the SAP BusinessObjects Suite:
- Dashboarding & Visualization: SAP BusinessObjects Lumira
- Data Analysis: SAP BusinessObjects Analysis for Office (AfO)
- Simulation / Planning: SAP Business Planning and Consolidation (BPC)
The SAP BusinessObjects Portfolio. Source: SAP
Let’s return to the initial question:
“If there was one thing you could do today to improve your organization's analytics level, what would it be?”
Feeling a little more comfortable with the considerations you need to be making to answer it? Remember that whatever you feel you need to do it doesn't have to be enormous, or expensive, or long, to start having an impact. Often small, quick to benefit projects can give the business confidence in analytics projects, making it more receptive to the larger scale ones.
We believe that it's possible to think big, start small, and act now.
 BARC – The BI Survey ‘16
 Gartner Says Business Intelligence and Analytics Leaders Must Focus on Mindsets and Culture to Kick Start Advanced Analytics