At a glance
Good data can create a competitive advantage for any type of business.
As a result, organisations are investing in cloud systems with centralised data pools, and advanced analytical tools and technologies, to glean key understandings, personalise customer connections and better understand their markets.
Bad data, on the other hand, produces only negatives.
What is bad data?
It can simply be incorrect information – an old phone number, a defunct email address, a cat owner recorded as a dog owner.
It might stem from a lack of understanding of context. It can come from poor interpretation, from relying on old data for decision making in a new and fast-changing market, or from comparing good data to an irrelevant benchmark.
And it can cost businesses and the economy millions.
Causes of bad data range from human error and unreliable collection methods to technology issues such as difficult-to-access data silos.
An error as simple as a misspelled name or an incorrect gender can create tears in the data fabric of the organisation, leading to negative financial outcomes.
The real cost of bad data
According to research from Gartner, bad data costs organisations an average US$12.9 million (A$20 million) per year. This average is amplified by the potential effects on large corporates.
Perhaps more alarming is that 60 per cent of businesses – predominantly large businesses – don’t measure the cost of poor quality data.
But even without the time and resources involved in measuring the costs accurately, most businesses are likely aware of the negative effects of bad data.
Kitman Cheung, technology sales leader at IBM Asia Pacific, says part of the problem is that today few organisations have a real understanding of where all their data lives, with most data remaining unanalysed, inaccessible or not trusted.
Bad data can affect an organisation in four crucial ways:
- Time: costs incurred because of lost time
- Income: opportunity costs related to lost income
- Expense: additional, unnecessary expenses related to poor-quality data
- Reputational: costs resulting from damage to the company’s standing.
“Let’s take bad billing data as an example – investigating a customer’s claim is a manually intensive and costly process,” Cheung says.
“If an error is found, additional fees may be incurred to correct the error.
“Most importantly, each occurrence of billing errors will impact trust in the customer relationship. A company’s reputation can significantly suffer as a result of too much bad data.”
According to Melody Chien, senior research director at Gartner, the concept of reputational damage means it can be difficult to accurately measure losses from bad data.
At the same time, great positives can come from small changes in data management.
“If companies do something very small, the impact can be tremendous,” Chien says.
“The means to fix a data quality problem can be extremely simple. It typically depends on attitude and approach.”
Compare your data to a benchmark
Gavan Ord, senior manager business policy at CPA Australia, says good data is about quantity and quality.
“Data needs to be from a reputable source and must be of a sufficient quantity to be able to be interpreted,” he says.
“You should compare your data to other data sets.”
“While data, by itself, is good, when you compare it to past trends or to other businesses in your market, it becomes much more valuable. That’s when it starts revealing trends and patterns.”
5 steps to becoming a data superhero
Finance professionals can adopt practical solutions to improve their data collection, presentation, literacy and interpreting skills.
Good data gives organisations what some describe as “superpowers” – such as the power to identify otherwise invisible trends, and the ability to see into the future.
To become a data superhero:
1. Be proactive
“Ensure all critical data is being captured correctly,” Chien says.
“For instance, I have a new customer. They tell me their phone number and email. I can easily validate whether this information is correct.
“Don’t wait until the problem manifests itself. Put safeguards in place to ensure the data is clean.”
2. Be interested
“Get to know your data better,” Chien says.
“Be familiar with the trends, the patterns, the nature and the shape of your data.”
3. Be aware and mindful when collecting data
“When organisations collect data, it must serve a specific purpose or strategy. And always ask whether this data is necessary.
“Once you collect personal data, you have legal and ethical obligations to protect it,” Ord says.
4. Be clear in presenting data accessibly
“This point often applies to accountants,” Ord says.
“If you’re communicating figures and people don’t understand them, you’ve missed the point and so will they. Instead, ask the client how they would like the data presented.”
5. Understand the data analytical technologies you use
“If you can explain to clients and stakeholders how the analytical tool your company has employed works, it builds trust and credibility,” Ord says. “It’s vital to understand how the AI technology has analysed the data and produced particular results.”
Bad data can cause grief within an organisation and among its customers. Organisations that use poor quality data are less likely to benefit from data use in the future.
However, several simple, practical solutions can mean data collected by a business is relevant, meaningful and rewarding for the organisation and its customers alike.
