Each component must be scaled appropriately based on your organization's structures and decision-making frameworks. Here are its core elements:
- Everything must follow the corporate strategy and objectives.
- How will you leverage your data strengths to achieve the company strategy, and how can we overcome data weaknesses?
- What capabilities are needed to implement the data strategy, and who is responsible for achieving them?
- Data Governance is the set of business roles and people we must have to create rules for data.
- Data management consists of the people executing the data rules and ensuring they are met.
- When, where and who will make decisions in the data governance bodies and committees?
- Data management deliverables include data quality, master data, metadata, risk policies, and risk registers. We also need to adjust our data architecture to achieve our goals.
- Focus on your data's crown jewels as they flow from data creators to data consumers.
- We must help people adapt to these new roles and processes and improve their behaviors.
- It's not worth designing an organizational framework that doesn't fit into the business.
What is a Data Management Framework?
You need to know what goes into a framework before implementing one for your company. Watch this video for more information:
Strategic Objectives and Corporate Strategy
Make this principle your guideline when developing your data-management framework. All expenses and efforts related to managing and cleaning up data must align with a particular business goal, or you risk spending both time and money cleaning data that does not help achieve any tangible business objective.
Data teams often need help aligning their strategy with the companies. Engaging business peers may prove challenging if your I.T. The department consists of six layers deep. But don't allow this as an excuse for misaligning goals; most organizations offer ways for staffers to identify them without consulting the CEO or Chairman of Board interviewee.
Aligning Data Strategy With Corporate Strategy
Read your company's annual report if you need help meeting with executives directly. Every public company issues an annual report listing its major strategic goals and expected challenges; read what management anticipates or plans on how they intend to respond.
Here is an example of McDonald's strategic plan, "Accelerating The Arches". You can access this plan freely online; to identify similar strategies in your annual report.
McDonald's intends to expand via digital channels, Drive-Thrus services, marketing initiatives, and core product offerings.
How To Develop A Data-Driven Strategy
You'll need to adjust your data strategy to meet these goals. Data strategy must include at least three components.
- Diagnose the problem that your strategy is solving or the goal you want to achieve
- The principles that will help you overcome the problem are:
- You can get to your destination by following a series of coordinated actions.
Diagnose A Data Problem
Take a look again at McDonald's and consider how data could play a part. It is clear that data will be used to help support the pillars of each strategy.
- Maximize our Marketing
- Data on the effectiveness of ad campaigns, marketing channels and customers will be needed.
- Commitment to the core
- McDonald's requires data on their products including their chicken, burgers and coffee in order to maintain their core offerings.
- Do the Triple D!
- McDonald's serves its customers through these channels. Here, data can be used to optimize operations, to create a more seamless digital transition, and to ensure that delivery and drive-thru experiences are up and to date.
This website doesn't go into great depth when discussing how these pillars address problems or opportunities; for more insight you will need to speak to those responsible for carrying out this plan and ask what could be achieved with better data; also inquire what challenges exist with using said data.
Keep an eye out for customer reviews online about your business; this could reveal areas in which certain data could be turning them away from purchasing from you.
Establish Some Goals And Principles
So if our data strategy's goal is to reduce negative feedback we receive online from customers about deliveries, then all customer feedback must be assessed to find where data has caused problems and evaluated accordingly by asking frontline staff how often these incidents arise and any associated problems exist before formulating responses that adhere to principles rather than rigid regulations.
- All McDonald's systems should provide consistent data about the availability of products
- The order data must be accessible and clear to allow our staff to deliver it in the correct format and place.
- We need to be able to deliver fast food by having accurate data about your orders.
The principles we follow will ensure that our goals are met. Staff can rely on these principles in the absence of rules. To determine if we deliver on our promises.
Want More Information About Our Services? Talk to Our Consultants!
Coordinate Your Actions
How will we address this challenge? As outsiders, it's impossible for us to know exactly which methods McDonald's currently utilizes, as we don't know them well enough. Since there may not be any data management systems set in place yet, we must establish appropriate power structures and decision-making processes so as to address this challenge effectively - thus using the Data Function Target Operating Model as our starting point and moving on from there. Now let's examine:
How to Create a Data Target Operating Model
Target Operating Models are future-state diagrams that depict what people will do, decisions will be taken on, and processes will operate. We will create our target operating model from scratch in this hypothetical example whereas you may already possess certain elements for building it yourself.
Data Governance and Management are central tenets in your Data Target Operating Model, yet many firms find them confusing. Let us clarify their respective meanings here.
Data Governance vs. Data Management
Data Governance's purpose is to set rules and expectations around our data, while Data Management executes these rules.
Data governance must involve both business and I.T. departments; ownership issues often present themselves within firms when data falls into unsuitable hands (our research revealed this to be true for 25% of firms); an easy way of increasing data maturity by doing this would be putting aside these two sides and working together on data ownership management together.
Who knows more about customer data? Who better knows this topic - such as customers - than the Chief Customer Officer? Which data they require on my customer and if any is stolen or lost. Businesses have an obligation to set expectations around data governance which in essence involves setting rules regarding this area of knowledge.
Who best understands how Point of Sale integrates with CRM? Who understands more about how these two systems can communicate together? Who could help us explore data models more thoroughly and create scripts to check whether all records contain telephone numbers as required? It is evident that I.T. experts play an essential part in carrying out data rules, assessing compliance with data policies, and data management in general - thus being an essential element in data governance processes.
Who Is Accountable For Data Governance?
Our data management framework clearly stipulates the requirement for three levels of engagement.
- We need an executive sponsor to help us prioritize and fund our goals and data strategy.
- Establish business accountability for data of high quality. What is the risk to their job if they don't provide us with data that will support our enterprise strategy.
- Who should document data meaning and expectations and enhance our understanding and trust of the data that we have collected and who is responsible for this?
What Is The Responsibility For Managing Data?
It is important to get the data management role filled. The three-tiered structure is also followed:
- I.T. Most likely, the sponsor will be either Chief Information Officer (CIO) or Chief Technology officer
- I.T. leadership - I.T. heads. Heads of Enterprise Architecture and Data Owners work together to resolve issues and prioritize them.
- Data Management Office: data modelers, experts in data quality and other professionals who can help us to make our policies and rules a reality
Data Management Framework
The roles have no meaning unless they are grouped into functions. Our framework has three major functional groups:
- The steering committee is the senior leadership of your data department.
- The program will be funded and the corporate objectives and strategy must be met.
- Take some time out of your meetings to make decisions about the data that you use and create.
- The Data Owners and I.T. Leaders need to get together to discuss the priorities of their teams.
- The group is responsible for measuring and maintaining progress towards objectives
- Senior leaders are able to break deadlocks that cannot be resolved by the teams themselves.
- Data Governance Team: Data Stewards, Data Management Office "doing work"
- IT and business experts are needed to model the data, establish quality rules for it, and resolve technical data challenges. We need IT resources and business Subject Matter Experts to model data, set up data quality rules, resolve technical data challenges, etc.
Framework For Data Management Activities
We need to assign them a task now that we've organized the team into roles. What are your key data management activities?
Improved Data Quality
Data quality outputs from any program for managing data are of critical importance to its management and will save both time and money, protecting both brand equity. To enhance data quality it's vitally important that business functions define exactly what expectations exist from data and develop data quality rules accordingly.
To succeed, you'll require four things:
- Data quality log to track the problems bad data can cause your business
- The business should have a list of rules for data quality.
- Data quality dashboards can be used to show how your data meets your criteria.
- Data triage is a process that helps you prioritize the issues with data quality.
Our research indicates that organizations are failing to deliver on this front. Of those we contacted, 29% had no processes in place for improving data quality while 54% are still looking into what process might best work. Utilize your data-management framework and stay ahead of the game!
Policies And Standards For Data
It is not surprising that we must define standards and policies to ensure all users of data in our company understand the changes we expect them to make. In order to do so, we need to turn the principles that we laid out in the phase of data strategy into policies with teeth. You may remember that in our McDonald's case study, we outlined three key principles.
- All McDonald's systems should provide consistent data about the availability of products
- The order data must be accessible, and clearly visible to ensure that our staff is able to deliver orders in the correct format and place.
- We need to be able to deliver fast food by having accurate data about your orders.
What are the steps to turn them into standards and policies? We could start off by outlining our policy.
- The point of sale system will become the definitive record of product availability
- The franchisee is responsible to ensure the consistency and accuracy of data.
- The accuracy of the product data will be the responsibility of all McDonald's employees
As we possess the premier source of product data, we should leverage that advantage by making certain all systems utilize this source when determining if an item is in stock or unavailable.
Franchisees are held accountable for providing data that conforms with our standards, such as maintaining consistency in POS data accuracy. Any discrepancies would lead to negative online reviews which harm our brand; and are held liable if their products don't conform with them.
At our organization, each employee is held responsible for this quality. By instilling this importance into them, their behavior changes accordingly and they become alert for any bad data that might enter our systems.
Read More: How AI is being Used in Data Management
What Are The Data Standards?
Per our company policy, product data must remain constant within our system. But what exactly is meant by consistent? For staff to know whether they're meeting the standards we set, these terms must be clearly described and defined.
Ensure the SKU data in our Point of Sale System accurately represents how many units exist - otherwise, it won't work! For example, if it says I only have enough lettuce for 12 Big Macs, this system won't function effectively.
McDelivery App and POS data must coincide with matching product numbers from both systems. Otherwise, we will only meet standards if an App sells products we currently possess in stock.
Master Data Management
Master Data represents those entities which our business relies upon as important. "Golden Records", truth-based sources that serve as decision-making pillars. Managing Master Data allows us to identify customers and better comprehend our relationship with them uniquely.
McDonald's could find it extremely valuable to know exactly how much I spent across multiple channels: in their store, on their app and through delivery services. By tracking how I spend my money across these various avenues, they can tailor marketing messages so I purchase from my preferred channel(s). To accomplish this feat, they would require unique identification as a customer and connecting all my online purchase data with purchases made offline in-store.
They cannot use all the information about me if they lack adequate master data, have duplicate records, or cannot link identities across systems.
Metadata Management
This phrase may sound tedious to businesspeople; it provides information on another piece of data. I use metadata within column titles to understand what each row or column entails.
Documenting metadata via a business glossary or data dictionary should include documenting all important elements within your company and asking subject matter experts (SMEs) to enhance it for non-expert users.
Information Risks and Controls
Hackers and other bad actors have an eye out for our data. Yet, we entrust this trusting information with customers and staff, trust that has to be protected at all costs. each record of the customer will cost $180 - how many clients do you have, how valuable is their information, what are its implications if mishandled, and will it affect their identity if mismanaged?
Establish a data risk register for your company. Speak to regulators regarding their expectations regarding data. Document, triage and implement controls that reduce risks whenever emerging risks emerge.
Data Architecture
This map details your company's current systems in use. Each system serves a particular function and integrates with other business systems for work to flow seamlessly from team to team; for instance, having two systems recording sales data while another creates invoices could present problems if their respective databases do not align perfectly.
Data architecture should always be essential when developing your data management framework. To be truly effective, systems must work harmoniously together.
Why is a Data Management Framework needed?
Data Crown Jewels form the cornerstone of data management. Your most crucial data can put your business at serious risk should it become compromised through loss, corruption or theft; using this measure as a gauge allows you to easily understand which files should be prioritized as important or otherwise.
To avoid "bloating up the ocean", it is crucial that CDEs be identified. Your resources are finite; therefore, it would be impossible for you to manage all data elements at the same time effectively; focus instead on those data elements which truly matter for optimal management results.
What Is Data Governance, And How Do You Keep It Up?
It's a good question because many programs are canceled. A Chief Data Officer's tenure is usually between 2 and 2 1/2 years. Data governance can be a challenge.
To maintain data governance, we need to make measurable improvements. It is important to set up metrics and KPIs demonstrating our work's value. You'll align your corporate goals with the data you use if you properly set up your strategy. It's important to demonstrate that the value your team brings is quantified.
You must make a critical shift in the data value chain.
Governing Data Creation
Your staff, customers, systems, and partners all create data. The data creators can type whatever they want in the absence of rules. It impacts our critical data flows and negatively affects our business processes. Sending correct invoices can be costly and efficient.
The majority of root causes for bad data are bad data generation. If we want to see lasting changes, training staff and changing data collection processes is important.
Data Consumption And Its Regulation
Whom do we hold responsible if there's no guidance as to what "fit-for-purpose" means? Am I to blame if I enter customer details into an open field instead of searching and using existing records instead? To set expectations accurately, ensure satisfaction among data users, and set realistic expectations accordingly.
Additionally, consumers of data need to be informed. Middle management often needs help.
At one company, the director of delivery requested we implement Project Start Date tracking into their CRM system so he could predict when projects would commence and plan resources accordingly. It made perfect sense - deals were being closed three weeks post-project start date!
This request could have been better managed had the company implemented an appropriate framework for data management. Such information can be gained by setting a rule that only projects may start three weeks after an order closes - this enables delivery teams to forecast demand based on close dates, which are often inaccurate, saving sales teams the effort of trying to create more accurate records themselves.
Change Management and Continuous Improvement
As in the previous sections, we compared data producers and consumers. It shows us that employees must develop different relationships with data; otherwise, it will remain negative without the change being implemented through behavior modification alone or technology solutions being found.
The data management framework is at the core of all our efforts. It is rarely considered, which explains why Chief Data Officers only last for a short time in organizations.
Data training isn't top of mind for companies; 61% still need to allocate sufficient budget for training employees on this critical matter.
Work closely with your employees to ease their lives as much as possible. Behavior change takes time; to ensure its longevity, it's key that we find ways of making it less scary for them.
The Corporate Organizational Model
All this work will go down the drain once and when your data management framework matches up with your business model and structure. Otherwise, any attempt at developing one that starkly contrasts with what already exists could quickly spiral downward into chaos.
Most organizations fall somewhere in between; their Data Management Framework can consider both approaches by centralizing certain core decisions while permitting business units to diverge from standards when needed.
Your chances of success increase significantly if the framework you create for data management fits seamlessly within the company structure.
Want More Information About Our Services? Talk to Our Consultants!
The conclusion
Data governance frameworks will help your organization reap the rewards of becoming more data-driven. Simply collecting information isn't enough - data governance refers to an overarching framework ensuring its management and compliance - you can build one by following this blog series; creating such a framework will meet organizational and industry-related requirements to achieve desired business results.
