BLOG | DATA & ANALYTICS
Supercharge your Analytics Project
15 June, 2021
Henri de Bruine, Head of Data & Analytics at Decision Inc. UK
Supercharging: In the fast-changing world of electric vehicles, the race is on to produce batteries that charge as quickly as possible. There’s a similar parallel in the business world. As businesses rely more and more upon up-to-the-moment analytics to get ahead of the competition, there’s pressure on IT departments to deliver analytics projects as fast as possible.
And with the expectation to get full self-service BI up and running, often projects aren’t as well thought-through as they should be and then hit problems.
The way to supercharge these projects is to bring in external expertise with the skills to really make a difference fast!
WHY IS MY INTERNAL ANALYTICS PROJECT NOT DELIVERING?
Business Intelligence (BI) products are often marketed as ‘self-service’ and we see many organisations embracing Self-Service BI as the way to facilitate their analytics requirements. But there is a gap between Self-Service BI and the underlying Enterprise Analytics Platform that is needed to deliver it.
Because of this, analytics initiatives have a habit of fizzling out when they do not deliver or even worse, go off the rails; this is when analytics sprawl starts happening.
Here are five common pitfalls of internal analytics projects that a supercharge could solve:
1. The software is viewed as a tool that solves all problems
There are many steps in moving the organisation to become a Self-Service BI (SSBI) organisation. These include:
- Creating the data platform it will run on;
- Designing the processes and controls for the analytics environment and then
- Driving the user adoption through change management and training.
The analytics tool itself does not solve the problem. It’s like owning a cordless drill and expecting someone else to go and hang shelves on the wall, whilst only showing them how the drill works! You need the shelf, a solid wall to mount it, plus knowledge of where the wires and pipes are in the wall.
Then you need to plan and measure where to mount the shelf and mark it out on the wall. From there a total novice can hang a shelf without flooding the floor, tripping the power or having that expensive Ming Vase break the first time you put it on the shelf.
2. The underlying data integration is implemented poorly
The biggest technical issue that even power users will run into is accessing and linking data from source systems, especially in an automated fashion. Master data or data quality issues need to be addressed which will require some data literacy or technical input to solve. Modern systems are moving more toward APIs as a mechanism to extract data which can be a challenge if authentication and network access control come into play.
3. The data modelling only happens at one level
To overcome the data issues above it is better to create curated data sets in the data platform. This moves technical issues away from the self-service user. But not all modelling can be done at a data platform level. Some metrics need to be calculated in the reporting data model because they depend on user selection at the time of viewing.
A very simple example is average price, this of course depends on the time interval and the products selected. In addition, business perspectives vary that require different relationships between datasets, such as alternative levels of aggregation and hierarchies, hence why you should create specific reporting data models aligned to multiple output requirements.
4. The governance around data reporting is inadequate
Processes for reviewing and publishing self-service reports to a greater audience must be put in place to ensure that the right access control is applied to sensitive or competitive data. This ensures that the interpretation and quality of the analytics are consistent across the business.
5. User training isn’t comprehensive enough
When users get training on a tool it’s typically done on simple and “perfect” data set that has no complexity in relationships or bad data quality to trip you up. Certainly, this type of training is essential and should be the first step to educate users, but it should not stop there!
Technical training and guidance on the data platform are also essential to ensure solid foundations are in place and the right skills enabled to support and extend the platform going forward. Co-development is a very effective way to get professional help to deliver initial data and reporting models whilst also training up users on their organisation’s specific data challenges.
The final report output from a good BI tool always has a strong allure and is something all users want, but that is only the part of the solution. It is a lot like an iceberg – there is a lot below the surface that needs to be put into place to ensure that the platform is successfully adopted and serves its purpose in delivering cost-effective, high-quality, self-service BI at scale.
How to move forward?
If you have hit some of these pitfalls, it might be worth supercharging your project by bringing in experts to assist you. At Decision Inc. we’ve assisted many companies to do this and the project can be done surprisingly quickly.
HOW QUICKLY CAN AN EXTERNAL ANALYTICS PROJECT BE IMPLEMENTED?
Analytics Projects lend themselves to an iterative approach with a staged delivery. Here at Decision Inc., we often describe it as a journey with the following stages.
An initial Analytics Dashboards (the user interface) can be delivered in as little as 1-2 weeks depending on scope and complexity.
2-3x time taken for the
Connecting to the Sources
At Decision Inc. we don’t believe a solution is complete until it is connected to the sources of data, eliminating all the repetitive and manual processing. This usually takes 2-3 times the length of the initial Dashboard concept.
Total Time to Value
These short timelines are achieved based on experience in:
- – Technology including SAC, Power BI, Qlik, Tableau, …
- – Finance & Data
- – Architecture and connecting data
If you would like to get your analytics project moving again, we’d love to see how we can help. You can contact us below and we’ll get one of our analytics experts to get in touch.
ABOUT DECISION INC.
DI is a global leader in information-driven transformation with a core focus on data, digital and performance intelligence. Our expertise in technology and industry specialisations have enabled us to help companies around the world make better decisions faster. Our emphasis on local excellence and global scale enables us to bring clients best of breed technology solutions that enable significant improvement whilst leveraging our Global Research and Innovation teams to accelerate the time to insight. Our local consulting teams are experts in the implementation, support and execution of these solutions and work with our clients to deliver real results and improvement.