- The Business Algorithm
- Posts
- How To Succeed In Data
On December 26th, Peter is sitting behind his computer with his hands in his hair.
4 months ago he approved the budget of $30.000 for a new machine learning product.
He’s looking at the results that the analytics team produced.
“This has literally no use for us” he mumbles.
Peter just wasted $30.000.
He feels empty inside. “How could this have happened?”
As a data scientist and management consultant I have seen this countless of times.
Fact: 85% of data projects fail.
A big reason for this is a mismatch between the data solutions and business objectives.
Here’s why: everybody wants to jump in the data bandwagon, but they have no process in place to guarantee success.
The consequence? Data becomes a money pit rather than a money maker.
Data strategy solves this.
Data strategy identifies valuable data products that move the business forward. It is a plan for managing and using data to improve business outcomes.
🎯 The 4 Elements Of Data Strategy
One might think that not all businesses benefit from data.
You would be surprised how many problems at the root are actually data problems.
Four areas are covered in a data strategy:
1. People
People enable the data solutions. They start initiatives to solve a business problem. Below are three points to consider:
1.1 Roles
There are different skills and roles you need depending on the new data solutions. Roles & skills are defined that are needed to develop data products and support old ones.
1.2 Organizational model
The location of the skills within the company. For example, some businesses prefer to have one data team, whereas others prefer to have a data professional in every business unit. Both have pros and cons.
1.3 Processes
Business processes may need to be redesigned to include analytics. This is done by documenting the steps in a process as well as where specific data are used to make a decision.
2. Technology
Technology is necessary to support analytics. Without technology, it is impossible to perform analytics. Two areas are important:
2.1 Data assets
The data needed for the data products. This part describes which business unit owns the data, what the source is and what the quality is.
2.2 Data operations
To make the data products operational, one might need other pre-made technologies. For example, you need tools to collect the data, analyze the data and deliver the insights.
3. Governance
Governance is put in place to ensure ownership and accountability. Clear roles should be defined on whom has access to which data. The governance part of the strategy contains the following:
3.1 Data ownership
A description of the employees who are in control of the data and their responsibilities.
3.2 Data control
Guidelines on what and how data is shared.
3.3 Data curation
Existence of data definitions and a data catalog (a book where all data is found). Data definitions are descriptions of what the data means. This should be consistent across the business to reduce the chance of wrong interpretations.
4. Culture
Culture brings everything together. A data-driven culture must be in place to act on insights and covers the habits and mindset of employees. Insights have no value if there they are not acted upon. Culture in data strategy should describe:
4.1 Data democratization
It is key that those who need data have access to it. Data democratization is a major problem in many businesses, which you solve with democratization guidelines.
4.2 Data ambassadors
Data ambassadors are people who drive the data culture. They talk about the importance of data and strive to convince everyone of its importance. They are required for data adoption throughout the organization.
🤔 Why Do You Need A Data Strategy?
A data strategy makes sure you will only spend time on projects that move the business forward. During the design of the strategy, valuable opportunities for analytics can be found.
Examples of these opportunities are:
Using customer data to improve customer service
Using sales data to improve sales strategies
Using marketing data to improve marketing campaigns.
These insights can lead to great benefits for the business:
Increased efficiency and productivity.
Improved decision-making.
Increased customer engagement.
Increased brand awareness.
Increased revenue.
And many more.
🤷 How Do You Create A Data Strategy?
I noticed that most data strategy approaches put too much emphasis on the current state of the company before defining where they need to go.
In my opinion, there is a better way. When buy-in from management can be hard, it needs to be really clear what is going to be delivered before investments are made. This is why I always start with a small assessment and then move on to identifying data solutions.
I sit down with the business and get to understand their business objectives first. After that, I define data solutions that support the objectives. This is done through an interactive workshop. Together, we find solutions for business challenges that increase business performance.
The approach consists of three phases:
1. The vision
In the vision phase I will define highly valuable data products through the means of a workshop. In this workshop I will use the business objectives to find interesting data products that are valuable to the company.
The data products are used as the foundation for the data strategy. They will be the north star where we want the business to move.
The first part follows 5 steps:
Asses business purpose, sponsors, and data maturity
Specify business objectives
Identify high-potential data products (I will cover this in a future newsletter)
Prioritize data products based on value and feasibility
Create a rough plan of when the data products are developed
2. A peek into the future
After the data products are defined, a current state analysis is conducted on the four areas. The data maturity assessment from phase one will help with this. We try to answer the questions such as:
What skills are already there?
What technology is in place?
Is there a data governance plan?
How is the culture?
After the current state is clear, the future state will be addressed based on the identified data products. The goal is to move the company to this state.
3. The guide
The final phase will bring everything together. Gaps in the current state and the future state are filled by the data strategy.
The end deliverable is a document with actionable steps:
Data products roadmap
Change management plan
Technology implementation plan
Data governance plan
The strategy should answer the following questions:
Where are we?
Where do we want to be and why?
How do we get there?
How long will it take?
What will it cost?
It is important to set clear deadlines beforehand to make this successful. Especially in analytics, you can spend many hours on little-impact activities if you do not set restrictions.
This quote from Elon Musk is appropriate here:
“If you give yourself 30 days to clean your home, it will take you 30 days. But if you give yourself three hours, it will take three hours. The same applies to your goals, ambitions, and potential”
💡 Tips For Your Data Strategy Design
Always base the data strategy around your strategic objectives. If the goal is to reduce churn by 10%, identify data products on how to achieve that.
Always tie a KPI to the data products to measure performance.
Check if the newly defined data products overlap with current data products. If what you've described achieves the same goal as a current data product, while being better and less expensive, you may toss the old one.
It is extremely important to start small to get buy-in from the organization. Make a proof-of-concept and test it with the stakeholders. They need to understand the value of these products to make the adoption a success.
Always align the strategy with leadership. All four elements of data strategy must be supported from the top of the organization. Without leadership support, it is not possible to successfully implement a data strategy.
In short:
Data strategy ensures that data initiatives follow business objectives.
People enable the analytics
Technology supports the analytics
Governance controls the analytics
Culture actives the analytics
Use expertise both in business and data to design the data strategy.
Now you know how you can make sure that your solutions are aligned with business objectives. My hope is that you apply this knowledge so you can make data more profitable within your organization.
If you have any questions, you can reply to this email. I read everything. 📩
If you enjoyed it, please forward this email to friends who want to know more about data in business.
See you next week!
Thomas