The promise of artificial intelligence finally came good in 2018 and 2019, with a wider adoption of AI - from its use in detecting and combating fraud in financial institutions, through to sophisticated analytics tools in contact center. There are a host of use cases showing the value of a future-facing AI strategy, leveraging accurate and collectable data to save time, improve efficiencies, and reduce operational costs.
In fact, a recent KPMG report states that five of the most AI-mature companies are spending $75m annually on AI talent, indicating the increasing importance of using AI by business leaders. The same report also finds that analysis of voice data is a high priority AI initiative, but there are some critical foundational elements that are maybe not being given the consideration they should.
Organizations interested in adopting this new technology - and those that already are - must remember that AI and analytics tools are fueled by data, and the output is directly correlated to the quality of the input. The proper foundations must be in place to capture, access and control this data to unlock and extract its value in order to maximize the ROI from AI. While organizations have benefitted from AI in recent years, a post-pandemic world will truly highlight areas in which their AI strategy is lacking - and where their competitors are excelling - and voice data in particular will be pivotal for any organization looking to innovate and gain a competitive advantage.
Your data is your business
One of the richest and most useful types of data is voice data. Historically, voice data has been captured for compliance and regulatory purposes - and with the introduction of GDPR, MiFID II, and Dodd–Frank Wall Street Reform and Consumer Protection Act, this is still true - but organizations are increasingly seeing voice data as a key strategic asset.
It denotes sentiment, tone and provides context to conversations in a way that no other dataset can. Even with the increase of conversational AI in the form of chatbots and online assistants, customers still seek out human-to-human interaction, particularly for more complex issues. This is something that’s evidenced by the significant increase in call volumes during the pandemic, with the NHS receiving over 95,000 calls a day throughout March, an increase of over 6000 percent when compared with the year before. Given the circumstances, contact centers are generating more data than ever before - data that can provide valuable and actionable insights when analyzed by AI applications.
Secure access to and sovereignty of that voice data, however, is just as paramount. The majority of businesses agree, with recent research showing 85 percent of IT Directors and CIOs want access to voice data collected for use in AI, but more than half (51 percent) of this data is either locked away in legacy platforms or siloed, making it difficult to access. This impacts organizations’ ability to feed high-quality voice data into best of breed AI and analytics solutions to drive their desired business outcomes.
However, even for those that do have access to their data, they may not be getting the full picture. Fewer than half (49 percent) of organization-wide conversations are being captured, while those that are capturing their data aren’t using their voice data adequately; nearly two-thirds of businesses (62 percent) are failing to use transcribed voice data to fuel their AI engines - a significant missed opportunity.
For example, insights from recordings are, in some cases, still being derived by individuals who are listening to and transcribing calls manually; not only does this bring with it a greater chance of human error, it is impossible to scale in any meaningful way, especially for larger organizations with larger contact centers.
Marrying AI & voice data
If organizations are able to overcome the barriers preventing them from fully accessing and utilizing their voice data, then there is huge potential to leverage this to fuel AI initiatives. High-quality recordings combined with state of the art speech-to-text transcription software makes analyzing unstructured and structured voice data at scale a true possibility, improving the customer experience, increasing staff retention, and driving revenue.
For example, call center workers often have to document important information while they’re on a call with a customer. Not only does this mean some information will inevitably be lost or transcribed inaccurately, but it also means that the customer receives a poor experience.
But, those who feed their voice data into sophisticated automatic speech recognition engines are able to create accurate, text-based files of those conversations for AI engines to reason over both in real-time and post-call analysis, which can drive improvements in quality assurance programs and better training of staff, ultimately benefiting the end user.
Furthermore, identifying trends and patterns of frequent requests, looking at root cause analysis, as well as best practices of how they can be solved effectively, is only possible at scale by using analysis of the data.
There are benefits for staff that go beyond training, too. While many look at technology and humans as mutually exclusive entities, AI and analytics tools in the contact center are a support, not a hindrance. For example, by automating the aforementioned manual - and sometimes laborious - processes, staff are in a position to take on more useful tasks. With the additional time available, staff are also afforded the opportunity to up-skill or, should they want to, take on more challenging tasks, which ultimately improves the skillset of an organization’s employees and ensures that they hold on to their top talent.
Voice data is the largest and richest source of insights most organizations have, and is increasingly recognized as a strategic data source to be leveraged by AI and ML engines. In the age of Covid-19, organizations are being pushed to digitize faster than ever before to support new ways of working and meet and exceed customer expectations.
As such, organizations must take action at the earliest possible opportunity to leverage their richest data set to drive improved customer journeys, process optimization and increased efficiencies, and happier and better skilled staff.
In both a current and post-pandemic world, any organization looking to differentiate, gain a competitive advantage, reduce costs and maximize profitability, must look not only to sophisticated technology, but to the foundations that this technology rests upon.
Richard Stevenson, Red Box