The evolution of artificial intelligence (AI) is taking place at a rapid rate. It’s been adding value and changing the way we work in ways that were unthinkable only twelve months ago.
With many organisations considering ramping up their AI efforts in 2024, it’s vital to recognise that AI tools are only as good as the data they have access to.
There is no added value delivered via AI where the data it has access to is incorrect or out of date, in fact, quite the opposite. Poor data fed into an AI model will lead to poor outcomes.
For example, if someone living in a low crime rate part of a city applies for home insurance, and the postcode the insurer has on their system for them is in a different area of the city with a higher rate of crime, this will have a bearing on the premium calculated by AI. A higher-than-expected quote because of incorrect data may encourage the customer to seek a quote elsewhere.
Data decay is a concern
A major issue for all organisations is that data decays quickly - something that has a significant impact on the effective implementation of AI. The causes of this are varied, and include people moving home, death and divorce. It’s why user contact data lacking regular intervention corrupts by as much as 25 per cent a year. Additionally, 20 per cent of addresses inputted online contain errors; such as spelling mistakes, wrong house numbers, and incorrect postcodes, largely due to the typing of contact information on small mobile screens. These factors are why 91 per cent of organisations have common data quality problems.
To avoid the issues caused by incorrect data the answer is to have verification processes in place at the point of data capture, and when cleaning held data in batch. The good news is the delivery of such practices usually involves simple and cost-effective changes to the data quality process.
Adopt address autocomplete
A valuable piece of technology to use at the customer onboarding stage is an address autocomplete or lookup service. It’s able to deliver accurate address data in real-time when onboarding new customers by providing a properly formatted, correct address when they start to input theirs. It also reduces the number of keystrokes required, by up to 81 per cent, when inputting an address. This accelerates the onboarding process and reduces the probability of the user not completing an application to access a service.
Such an approach to first point of contact verification can be extended to email and phone, so that these important contact data channels can also be verified in real-time.
Deduplicate data
With the average database containing 8-10 per cent duplicate records, data duplication is a significant issue. It’s usually caused when two departments merge their data and mistakes in contact data collection occur at different touchpoints. This duplication of data not only has the potential to confuse an AI application, but it adds cost in terms of time and money, particularly with printed communications, and also negatively impacts on the sender’s reputation.
The answer is to have access to an advanced fuzzy matching tool. This technology can merge and purge the most challenging records to create a ‘single user record’, enabling the delivery of an optimum single customer view (SCV) that AI can make learnings from. Additionally, organising contact data in this way will maximise efficiency and reduce costs, because multiple outreach efforts will not be made to the same individual. A further benefit is that the potential for fraud is reduced because a unified record will be established for each customer.
Undertake data suppression/cleaning
A crucial part of the data cleansing process, and therefore in supporting efforts with AI, is undertaking data suppression, or cleaning. This technology highlights people who have moved or are no longer at the address on file. Along with removing incorrect addresses, these services can offer deceased flagging to prevent the distribution of mail and other communications to those who have passed away, which can cause distress to their friends and relatives. It’s only by applying suppression strategies that organisations can save money, protect their reputations, avoid fraud and support their AI efforts.
Source a SaaS data cleaning platform
Evolving technology means it’s never been easier or more cost-effective to deliver data quality in real-time to support AI and wider business efficiencies. It’s possible to source scalable data cleaning software-as-a-service (SaaS) technology, such as our Unison platform, that doesn’t require coding, integration, or training. This technology cleanses and corrects names, addresses, email addresses, and telephone numbers worldwide. Also, records are matched in real-time, ensuring no duplication, and data profiling is provided to help identify issues for further action. Unison’s single, intuitive interface provides the opportunity for data standardisation, validation, and enrichment, resulting in high-quality contact information across multiple databases. It can deliver this withheld data in batch and as new data is being gathered. Additionally, such a platform can be accessed on-premise, if required.
AI models are only effective if they have access to quality data. If the quality of data is good AI can deliver a competitive edge to organisations. If it’s poor it will lead to unreliable predictions and poor outcomes. It’s why implementing best practice data quality procedures must be a New Year’s resolution for those serious about harnessing the power of AI in 2024.