5 Steps to Improve Data Quality
Demand More from your Customer Data: 5 Steps to Improve Quality Data
In the digital era, almost all customer experiences and processes are fuelled by data. New technology has led to some huge advances in productivity but has created a reliance on data that can be double edged sword. If data is incorrect or inaccurate, these efficiencies are quickly eroded. Moreover, the more processes that need data to run, the greater the impact of poor-quality data.
According to the Harvard Business Review, 47% of data has some sort of integrity issue. This translates into an average financial impact to companies of $15m (source: Gartner). For example, money is lost through misleading “insights”, wasted time and resources, missed opportunities, poor customer experience and reputational damage.
For these reasons, improving data quality has risen to the top of the to-do list for many COOs, CIOs and CMOs. However, tackling data quality can be a daunting task, especially when customer data is spread across a myriad of systems in your organisation. This blog outlines Acrotrend’s approach to auditing data quality and putting an action plan in place to fix the root causes of poor quality data.
1). Focus on Your Most Important Data First
Not all data is created equal and fixing everything at once is not an option. The best starting point is to figure out how your company uses data to support its business strategy, and where the biggest problems are occurring.
Key questions to ask are:
- What are your business strategies and goals?
- How do you measure progress against these goals?
- What customer data is required to support these objectives?
- Which teams and processes need access to this data?
- How effectively is customer data being used today?
- Which systems collect and store customer data?
At Acrotrend, we start out by mapping the main touch points in the customer journey. After that, we overlay the teams/systems involved in the customer engagement process and where teams are having data issues. The result is a heat map showing where in the customer journey data issues cause the most pain.
This will often highlight marketing, sales or service teams as having the biggest data challenges, but data problems in finance, product or compliance teams could also have a big impact on business performance.
2). Quantify the Size of the Problem
Once you have identified the systems, data sets and teams where data quality is slowing the execution of your business strategy, the next step is to size the problem.
Acrotrend recommends assessing the quality of data across 7 axis. For each key entity / object (e.g. account, contact etc.) in the system in question (typically a marketing automation, CRM, database or data warehouse system), measure the variation in data for that entity based on:
- Accessibility & Usage
Once armed with this information you can visually represent and compare data quality across systems to quickly highlights problem areas.
3). Think Ahead – Plan to Join and Deduplicate Data
As part of the audit, make note of any fields that could be used to join or stitch data across systems. This is important because the more fields that are available and populated, the easier it will be to merge data in a single customer view or data warehouse system. As a result of the complexity of today’s digital journeys, a 360º customer view is desirable because makes them easier to understand and visualise. Even if this is not in your immediate need, it could be in the future.
Data points that can be used to stitch together customer profiles include:
- Address / postcode
- Contact telephone number(s)
- Login ID
- Customer reference number
- Device ID
- Social media ID
- Data of birth
When more join fields are populated, it is easier to merge and deduplicate data across systems (or even within a system if possible duplicates have been identified).
4). Make the Business Case for Change
Evidence shows that poor data costs money, but how much is it costing your business?
By now you know which teams are having the greatest trouble with data and have an idea of what problems there are. In addition, you should validate that the business case for the improvement program adds up. In order to justify funding to fix the problems, explore the benefits of improving data quality. Typically, these benefits occur in 4 areas:
- Cost savings/ efficiency gains
- Avoidance of rework or exception processes
- Faster access to data or insights
- Increased service/staff utilisation and scalability
- New revenue/ growth opportunities
- Responding to new cross or upsell opportunities
- Better customer engagement through more relevant, personalised marketing
- Improved sales planning, forecasting and territories assignment
- Risk reduction/ compliance needs
- Improved decisions making/ avoidance of poor decisions
- Greater trust and loyalty in your brand
- More accurate compliance reporting
- New use cases / strategic capabilities
- Readiness for Data Science and Machine Learning
- Increased adoption of systems like CRM by staff, boosting ROI
- Higher staff satisfaction and retention
5). Define an Action Plan
Using your business case, you can prioritise what actions to take. For instance, these could be quick wins that are fast to implement, or more wide-ranging strategic changes to your customer data.
Some examples of quick win data quality changes include:
- Making key fields required
- Using types and pick lists to restrict variance of data
- Help data built into forms
- Implementing validation rules at the point of data entry or in processes
- Use a naming conversion for things like marketing campaigns and events
- Leveraging 3rd party reference data to confirm the accuracy of or enrich data
- Reviewing ownership and permissions to access and update data
In addition, remember that sometimes less is more – removing options from UI screens can encourage employees to make better use of the remaining fields.
- Remove non-vital fields from forms, pages or CRM screens
- Clean dashboards and reports to focus attention on the data points that matter
- Guide form completion with pick lists, flow logic and help text
Strategic data quality changes could include:
- Defining a conceptual data model, relationship map and metric glossary for your business
- Link accounts or contacts across systems with a unique customer ID
- For example use a DUNS number, CRM ID or unique generated ID
- Put in place a data governance board to define data policies
- Nominate custodians or guardians to own data quality for key systems
- Workflows and automation processes to find, clean and merge records
- New infrastructure to manage, catalogue and master data
- Create a single customer view, giving insights, dashboards and clean data to employees and downstream systems
- For example, use an existing system such as your CRM
- For example, implement a new system such as a data warehouse
Starting Your Project
In this guide we have provided some tips and inspiration for your data quality project. If you need more assistance, Acrotrend provides a comprehensive range of data quality services to solve your customer data pains!