Feb 2021 | Data Quality |

Improving data quality could be the most important aspect of your data management strategy; however, you’ll rarely see it at the top of the Board’s agenda. Crucially though, all processes, decisions and outcomes the Board makes are based on the data your organisation holds and, if that data is inaccurate, then the downstream impact on critical differentiators like customer experience, can be costly, both from a resource and expense perspective.

Conversely, organisations with better data can explore a wealth of insight-led opportunities over their competitors.

In this article, we lay out five questions to consider when thinking about data quality improvements.

1. What is your data requirement?

Are you collecting the data you need? It’s one thing not collecting the right information, it’s another to collect data that you do not actually need. Collecting more data could simply mean there’s more to manage and more opportunities for inaccuracies. Take webforms for example, many have preset capture fields out-of-the-box, so spend the time to decide what data you really need and how valuable it is to you. You should also consider how you securely store and manage any personal and sensitive data that you collect.

It’s worth reviewing your processes and ascertaining what data elements are required to run your organisation or build valuable insights and no longer collect the rest. Don’t forget, you can always append enrichment data later on if you need it. More on that later.

2. How do you capture data?

Typically, first party data is captured in two ways. The first is digital web forms, most commonly on your website, and the second is via call centre staff or other employees.

Both methods are subject to purposeful and accidental misinformation. Some typical examples are detailed below.

We have all entered invalid data on a webform, normally due to wariness about how the data will be used by the recipient. We enter the basic amount to achieve our goals with little concern for the quality of the rest.

Likewise, we have all provided personal information to call centre staff over the phone only for at some later date an irregularity or issue to surface (often details presented incorrectly in an official letter).

However, there is a lot that can be done these days to monitor data entering your business. Be that validation checks of digital web form data or profiling data collected in your call centres. If you have no measures in place to understand the accuracy of the data at point of capture then it is something worth addressing. If you are measuring the quality levels and your data repeatedly falls beneath your thresholds, then explore whether the data you are collecting is susceptible to inaccuracies.

3. How accurate is the data you already hold?

If you cannot answer this question then you should act fast, as your organisation currently interacts with customers, and makes decisions, based on this data.

We discussed profiling your data earlier, and this, combined with a better data collection methodology will go the furthest to improving the health of your data as you’re fixing many of the issues ‘upstream’.

You could also look to cleanse the data you already hold. For example, 17% of Australians move house every year, so regular cleansing of your data can help you keep on top of these constant changes.

A common issue in many organisations is duplicate contact records. Even with data monitoring in place, duplicates can still sneak through as everyone tends to mix use of email addresses and other identifiable information. To deal with historic duplicate challenges a more robust matching exercise may be called for to help create that single view golden record. The downstream impact of poor data quality can affect all aspects of your business and create a huge amount of manual correction work.

4. Can you add insights from third party data providers?

There is a limit to the amount of data you can collect from your clients simply due to the fact it will be deemed invasive, or they simply will not know the answers you are seeking.

Official sources of data can be an invaluable resource when profiling and segmenting your customer base.

There are a wealth of data sources available, as well as numerous data files made accessible by the government that can help you append crucial insights to your existing portfolio.

Working with accredited third parties, such as Experian, is another great way to enrich your data. We can add demographic and behavioral information that will help you interact, target and delight your customers.

5. How connected are your systems?

A common cause of data error is from the movement of data through your business. Often moving data between systems is a manual effort which, as we have discussed, is prone to human error. Improve automated ‘data flows’, and you improve the quality of the information you hold in multiple places.

Connecting internal systems is becoming common practice to ensure the integrity of data moving through your organisation. For example, having your CRM system speak directly to your BI instance will ensure there is a single source of truth and that multiple systems are using the same source data points.

Experian has a range of integration tools that work at the point of data capture to ensure your data flows through your organisation seamlessly.

Don’t underestimate the importance of data quality

It’s easy to push data quality initiatives down the pecking order due to other business priorities and limited understanding about how important it really is.

As we’ve discussed in this article, the downstream impact of poor data quality can affect your customer relationships, your decision making and your ability to grow.

Get data quality right and you could unlock a wealth of opportunities for your organisation and power your decision making going forward.

If you’d like to discuss your data quality initiative, then please get in touch.


Contact Us

By providing your personal information you agree that we may collect and process it in accordance with our Privacy Statement.