Becoming data-driven: A 5-step approach

Have you ever reflected on how we build companies? How growth always means more people and more systems, organized in multiple silos. Instead of horizontal and lean, we turn into these complex and vertical organisms. To compensate, we spend enormous resources on coordinating tasks, people, and data: More meetings, IT-systems, culture programs, policies, and control. The spiral of increasing cost and effort to keep it all flowing just keeps on spinning.

Fundamental change
Now it´s time to fundamentally rethink how we configure our companies. It´s time to become truly data driven. It´s time to move from a human centric to a process, data and ai centric way of operating. But how? This transformation may seem hard to grasp and immense to carry out.  In reality it is a different way of thinking and organizing your business logic:

Becoming data driven: A 5-step approach:

Step 1.
Understanding and recognizing the fundamental (and potential disruptive) implication of what data and new technology means to my business

Understanding and embracing how data and new technology will fundamentally reshape our way of operating is core. Crucially, we are not talking about how a bunch of new apps can change isolated pieces of work within different functional areas of the business. We are talking about how it can be used to fundamentally redesign core end-to-end processes.

Forget about current way of working and of the inherent limitations – it only cements your mind.

Instead, imagine how we could use data and technology for all its worth to automate tasks, flow, and ignite exponential learning (with the help of machine learning). How we can shift from a manual, sequential, and linear process configuration to an automated, parallel, and exponential process configuration. Electrifying your processes – make them come to live.

Output: A visionary end state picture of how our data driven core end-to-end process would look like.

Step 2.
Where to start – The order of the sequence is not insignificant

Instead of trying to design and build the whole complete end-to-end solution in one classical major push, try to divide the big picture into logical business components. Then ask yourself which of these components would be a good starting point. A first MVP (minimal viable product) that could inspire and trigger a fundamental shift of mindset in our organization (NB! it is not necessarily from left to right of your process). Involving users, learn and adjust. Get value out fast. Build understanding, enthusiasm, and momentum.

Output: High level blueprint, split up into logical components. What component (MVP) to start out with.

With a visionary end state up on the wall, and a qualified guess on where to start, you are ready for the hard work to begin.

Step 3.
Digg into the details and build your data driven process model

Now it is time to deep dive into the detailed business logic of your core process. Sprint based and in detail, get a systemized picture of how every activity and step of the process could be done, and how each individual case can flow through your company from start to finish. For each individual case, the right task/information flows to right role – at the right time. Involving humans only when involving humans is right, otherwise automate. Choose a technology with right fit for your type of process complexity, and build your data driven process model step-by-step as you sprint along.

Important: Challenge yourself in terms of:

  • Potential for using data to automate tasks
  • What roles really needs to be involved, and what info/data/task should flow to this role/group, and when
  • How much can be done in parallel
  • Front-end loading: Capturing, and using, the right data early in the process can have a huge business impact
  • What information/data and what applications are essential to support the new data driven process. Which of the current applications are used/not used/has the right fit/attributes?
  • Keeping strict focus on automation, flow, and orchestration. Do not fall into the trap of including functionality that should be done by best of breed applications
  • Not falling into the classical waterfall trap

Output: A detailed break-down of each step of the process, with tasks, rules, how the task is performed (automated or by role), dependencies and data sources.

Step 4.
Plug in your data

Now that you know which activities that make up the way you want to work, it is time to plug in your strategic applications and/or data platform and make your data go to work.

This step is all about making sure that you are utilizing your data operationally, not just storing data in separate silos, but in a consistent, structured, and transparent way sharing data and connecting business activities with relevant data. At the same time, you should introduce a learning loop for your data, build in machine learning capabilities. The aim is to have the systems and data that govern your core processes work together in an optimal way, learning and generating as much value as possible.

What  are the relevant data?
Overall, the key to effectively using data in business is to have a clear understanding of what data is relevant to your operations and how it can be used to achieve your business objectives. The key to do this is an activity-centric approach using maturity, experience and continuous improvement to structure the way you want your business to work.

To incorporate data and application functionality into your business processes, you will need to first identify the data and functionality that you need. Start by identifying the specific data and functionality that you need to support your business processes. This will help you determine which applications and systems you need to connect to and what data needs to be collected and analysed.

How to access the data?
Second, determine how to access the data and functionality. Once you know what you need, you will need to determine how to access it. This might involve integrating with existing systems, building custom connectors, or using APIs to connect to external services.

Next, you will need to collect and integrate the data into your systems. This might involve setting up automatic data feeds, writing custom scripts to extract data from various legacy sources, or using third-party data integration tools. Or, you can use this opportunity to further strengthen the use of an API management tool to make end-points generic and available to other data consumers in your company.

Once you have access to the data, you can start using it to improve your business processes. This might involve using data analysis tools to identify trends and patterns and integrating the data into your process in order to make them available to the right users when they need them.

Finally, you will need to monitor and maintain the data and functionality that you have integrated into your business processes. This might involve periodically checking for errors or issues, updating data sources as needed, and making any necessary changes to your integration processes.

There are several reasons why businesses might want to connect their strategic applications and data. By integrating strategic applications and data, businesses can streamline their operations and reduce the time and effort required to perform various tasks. This can lead to improved efficiency and productivity. By integrating data from multiple sources, businesses can gain a more comprehensive and accurate view of their operations and make more informed decisions. By connecting strategic applications and data, businesses can facilitate better collaboration between different teams and departments. For example, sales and marketing teams might be able to access customer data in real-time to improve their outreach efforts. By leveraging data and technology to optimize their operations, businesses can gain a competitive advantage over their rivals.

Output: API management enable you to administrate efficiently your data integrations.

Step 5.
Learn, adjust, scale

After going live, pay close attention to user feedback and look for improvements. Let the organization feel how fast the improvement wheel now spins. How they are involved in the process and the improvement work. How their suggestions are taken seriously, and the response comes in matter of hours or max 2 weeks.

Measure process performance, learn and adjust. Then you are ready to take the next step, extending the solution with the next component.

Output: MVP is alive and kicking, and the continuous improvement wheel spins fast

An alternative to do all of this on your own and from scratch, is to start with pre-configured process logic solutions for your industry. That’s what Insert:Logic do. We delivers component based end-to-end solutions for different Industries. Solutions you quickly can adjust to your company, connect to your data sources, and start using and improving fast. Data driven.