
Let’s talk about data and data quality.
A study by IBM has revealed that poor data quality costs the US economy up to $3.1tn each year.
The same study suggests that 90% of all data today has been created in just the last two years and that 80% of all data is unstructured.
But what does this mean for businesses?
As we’ve previously written, business data is an asset and should be treated as such.
Poor quality data can lead to mistakes, inefficiencies, and potential losses in revenue.
It’s important that your business data is fit for purpose.
Profit from Your Data
The road to quality data can seem like a long, difficult journey. But it’s one that every business eventually needs to make. As data ages, the business becomes less efficient; it’s not possible to maintain operational efficiency if you can’t make robust decisions.
Like many tasks, data quality is best viewed as a long-term concern, and something that will be tackled in controlled phases. We’ve split the journey into seven distinct parts. Each one is a stop you’ll make on the way.
1. Get Out of the Pit
When tackling a data quality problem, taking the first step is often the most important part of the process. With a long, difficult challenge ahead, it can be surprisingly difficult to get the wheels in motion.
To make progress with any problem, we have to recognise it and stop denying its existence. Own up. The organisation has a data quality problem; it’s not impossible to fix.
Most – perhaps all – businesses will eventually find themselves in the same position; they need to put their foot on the gas. The smart ones are already on the starting line.
2. Make Everyone Drivers
Once the business accepts that there is a data quality problem, it’s natural to try to identify the perpetrators or assign a single problem solver to the cause. Contrary to popular opinion, data quality is not a problem for the IT department, and incorrect addresses should not be assigned to the marketing department for a fix.
Every department uses data to make decisions; several departments interact, change and update data as part of their daily processes. The entire organisation needs to approach the data quality challenge, from the CEO downwards. That means tackling data quality holistically, and with full buy-in from management at every stage.
With data quality, everyone is in the driving seat; everyone is responsible for steering the business towards change.
3. Assess the Bodywork
For some organisations, it takes a while for the cracks to show. How deep does the damage run? Have employees been silently coping with it, developing workarounds to manage risk? Often, the true extent of data decay only becomes apparent with a full data MOT.
This is where the challenge can seem insurmountable since the data quality problem could be hidden away. But don’t despair. If your data quality problems are extensive, it makes sense to find the biggest problems. Fixing these will yield the fastest progress as you travel towards a resolution.
4. Talk Tactics
Data quality problems affect living, breathing databases, such as Customer Relationship Management (CRM) systems. They also affect siloed data just the same. Even though siloed data is not being changed or corrupted actively, it is gradually becoming out-dated, and that’s just as bad. In a data warehouse, the extent of decay may be hidden away, and there may be fewer reasons to tackle the problem urgently since the data is not urgently needed.
Your organisation must come up with a data quality strategy, treating every silo or repository as an individual checkpoint. By looking at data cleansing methods, the scale of the problem, the urgency of the problem and the impact of the cleansing operation, you can plan your journey tactically and efficiently.
5. Find Broken Parts
Data naturally decays, but in some cases, broken processes are continually corrupting it. A simple example is a database input operation that causes invalid or incomplete data to be injected into a database. Any movement of data from analogue (such as a paper form) to digital is also prone to attract error.
Identify processes that are defective, either because they never worked, or because they are out of date. Pick out workarounds that are designed to smooth out the bumps in the road, but which may be creating a cycle of ‘correction and corruption’ instead.
Use the data quality initiative as a change driver so that these processes are eliminated or improved.
6. Research Data Quality Tools
To tackle a data quality problem, you need a tool that is fit for purpose. Data quality software is the tool you will use to bring about change.
Look for software that is cost-effective, fit for purpose and efficient, and ensure the tool can tackle the problems you’ve identified at the right stage in each process. Remember: the data quality journey requires involvement from all areas of the organisation, from non-technical to technical staff. The software should have no learning curve if manual intervention will be required from a non-technical user, but it needs to be powerful enough to give developers full control of the data cleansing, deduplicating and matching process.
7. Cleanse, Deduplicate and Continually Improve
Once your data is clean, the journey to recovery is complete – at least for now. But that doesn’t mean you can switch off the engine. Data quality will begin to decline again without on-going maintenance, and that’s why your data quality tools should sustain your efforts long-term. Without frequent check-ups, you cannot be sure that your data is supporting business objectives.
DQ Global’s data quality products are fully integrated for the lifecycle of your database or CRM system. Choose from a suite of products that enable fast, cost-effective resolution of data quality problems. By using modular data quality software, organisations save time and money without continually having to export data for cleaning. These clean, healthy databases allow businesses to obtain reliable information that results in accurate business decisions.
Don’t despair when looking at the road ahead. With our help, your data will be fit for business.
Contact us today to learn more about how we can help you on your route to great quality data.