Voice Assistants

Alongside the race to implement textual based chatbots, moves are also being made towards achieving natural language processing technology (NLP) which is software that can interpret human speech and respond in everyday language.

This technology is somewhat new, with Amazon only very recently releasing its smart speaker, ‘Amazon Echo,’ in October 2016.

This intelligent personal assistant can perform a variety of functions in the household, including managing schedules, carrying out internet searches as well as controlling light and temperature in the home all through voice activated commands.

It is estimated that in less than a month, 11 million Echos were sold globally, an estimate which supposes that over 8% of US households now use NLP technology.

Innovators are keen to remove the ‘manual’ aspect of using technology. Although laptops and android phones are impressive, they still require the user to manually input data through typing, or physically tap through websites to find accurate information.

Voice automated technology attempts to remove this by perfecting the accuracy of their software. For example, Apple’s ‘Siri’ has an impressive rate of 95%+ accuracy, and is able to sift through false news articles and illegitimate websites when providing the most recent news.

However, this is not to say that voice automation is perfect, or even close to what innovators hope NLP technology can achieve. There are many limitations and defects to voice automated products which only proves the potential NLP still hold in transforming the way in which we live in the future:

Ambiguity

Adam Cheyer, one of the software developers that created Siri claims that tackling the ambiguity of human language is the most difficult technical challenge to overcome.

For example, when asking Siri to ‘book a four-star restaurant in Boston,’ many problems arise from such a simple request. Whilst ‘Book’ and ‘Star’ are both cities in the US, there are also four Boston’s!

Activation

Since Amazon Echo is ‘always’ listening when switched on in the household, it is very easy for the program to pick up normal conversation and perceive it as a command.

In the US, a young child managed to use Amazon Echo to order a £130 doll house and a tin of biscuits! When the local news station reported this story, viewers’ Echos were activated, triggering many systems to buy more doll houses.

Accents

Although NLP systems are well suited to listen to an array of languages, these programs are not the best at understanding different accents. Amazon has begun scouting for linguists to help solve the problems the Echo has had with understanding the Australian accent.

Time, Cost and Productivity

There’s an assumption that voice automated services will make life easier, quicker and more productive. However, this is not always the case. For starters, the development of this software is pricey and lengthy. Not only this but the lack of accuracy can actually make tasks longer for the user.

For example, if using speech recognition to write an email, the user has to be sure to speak at a reasonable pace that is grammatically accurate. The user must then go back to proof read the email before sending it off – surely it’s easier to type it up and send manually?

Despite the setbacks, voice automation is being implemented in different industries and has been proven to be a success:

Banking Security

HSBC now allows users to use their voice as their password when carrying out telephone banking. To pass through security, users simply repeat 'my voice is my password' twice before being passed onto an agent.

Household

As already mentioned, Google and Amazon have both created programs for the home. These voice recognitions machines can perform an impressive number of actions - you can use them to check the weather, find new recipes, exercise routines as well as carry out online purchases for you.

Crime Management

Voice biometrics are increasingly being used in police investigations, allowing audio evidence to be used in court cases.

Only seconds are required to match live or recorded voices, and boasting an accuracy of over 99%, 'Voice ID' brings a new angle to forensic investigation.

Education

Voice recognition technology has proven to have many benefits for young children with learning disabilities.

Not only this, but it provides an alternative for individuals who are physically unable to carry out manual based tasks, such as typing on laptops or phones. Therefore, it allows technology to be something that is accessible to the many.

What does this mean for the pricing industry?

Like any other industry, natural language processing technology is set to inevitably change the landscape in the pricing industry. Though the possibilities are endless, we are able to make some educated guesses as to what changes may occur.

Voice Automated Virtual Assistants

Managing pricing is definitely a tricky task - especially if you implement dynamic pricing, which means that it is likely that the value you sell your products at will change often. It is easy to forget to update your website, or to regularly check to see if price recommendations have changed.

Through NLP, the user will be able to set voice automated commands and ensure that their pricing is regularly updated and accurate.

Security

With so many competitors prevalent in the market, it is necessary to keep your pricing strategies and business information confidential. Voice automation can easily lend a hand in improving the security of businesses.

Sales Support

What's better than a great face to face sales pitch? A virtual one, where the NLP is able to provide you with accurate information based on reviewing a large source of data. This program will be able to recommend you products based on previous user experience, case studies and specific needs of the individual.

And Beyond?

It's difficult to know how natural language processing technology will impact and change the pricing industry when considering how fast it develops!

But we can suppose that it will be something exciting to look forward to in hopes that it will make pricing software faster, more accurate and user-friendly.