Siri, Why Should Google and Microsoft Fear You?

Funny, I had an article entitled "what happened to speech recognition?" in the works, which I started before the launch of the iPhone 4S.

I've been involved in speech recognition technologies for the greater part of the last 20 years - and despite the never-ending slew of technological advances, while speech remains the fundamental means of communicating with each other as humans, it still hasn't taken off as the means of communicating with machines.

It's quite hard to put a finger on exactly why, in the sense that there's no single obvious reason; mainly it's a complex recipe of issues involving over-expectation and under-performance, not to mention the relevance of the alternatives.

Could that be set to change? It looks disctinctly possible - It won't happen overnight, but Apple has a track record of generating the momentum for technology adoption, even where the technology is not necessarily entirely new (mp3 players and tablets immediately spring to mind). Here are some interesting views on the topic...

 

As I watched The Wall Street Journal’s All Things Digital Asia interview with Android’s Andy Rubin, I was highly intrigued by his comments about Apple’s Siri. Rubin told Walt Mossberg, "I don't believe your phone should be an assistant." He said, "Your phone is a tool for communicating. You shouldn't be communicating with the phone; you should be communicating with somebody on the other side of the phone."

Furthermore, when questioned about Siri, Microsoft’s Andy Lees said it "isn't super useful." At the same time, he noted that Windows Phone 7 has a degree of voice interactivity in the way it connects to Bing. Thus, it harnesses "the full power of the internet, rather than a certain subset."

What are these two guys smoking? They both seem to ignore the fact that Apple has just introduced voice as a major user interface. Its use of voice, coupled with AI on a consumer product like the Apple iPhone, is going to change the way consumers think about man-machine interfaces in the future.

But I think their responses were rooted in jealously and the fact that, based on what it will soon become, Siri will ultimately threaten their businesses.

 

[continue reading]

..On Self-Service technologies

I found an old interview I'd done with Contact Centre World back in 2005 about self-service technologies (while I was at BT). Everything that was said then is just as relevant today.

Nik Sargent - Self-Service Product Manager, BT On Self-Service Technologies

What do you think are the three biggest mistakes managers make when choosing a self-service technology?
A common mistake is getting too bogged down in the technology. It's important to remember that the technology is a means to an end, a way of delivering a better service to customers. The three most important things to consider are usability, usability and usability. It is also important to have a framework that can grow and adapt with your business without holding you back. But that's as much about your people and processes as it is the technology.

In your opinion, what criteria should be used when evaluating different self-service solutions?
It's easy to be lured by cost alone, both in terms of cost of the solution and the promise of cost savings.  This should absolutely be part of the decision-making process, but it's also necessary to understand that your self-service solution is an important customer touchpoint: it can win and lose customers for you.

From a service point of view, how are you going to be able to deliver a slick customer experience: is designing this your core business, or do you need to buy this in? What flexibility will you have to modify your solution as your business evolves – and who will ensure your service works as well after it is changed as it did on day one? From a technology point of view, are you getting robustness and flexibility to shrink and grow and handle the volume you need to? If customers like your solution, then they will start to depend on it.

Within the next five years, which self-service technology do you think will have the biggest impact on the contact centre industry?
Speech Recognition will have a huge impact. Adoption rates of the technology are steadily increasing and organisations are seeing significant benefits. The availability of mass-scale solutions demonstrates that the technology is totally viable, and also educates the public about these types of services and how to use them. As we have seen, anything that can transform the cost base of the contact centre industry – such as outsourcing – can have a big impact, and Speech Recognition fits this profile.

What types of questions do you believe online self-service technologies should be able to answer?
I think the days of the Internet solely being a glorified brochure-publishing medium are over. Users are increasingly savvy and looking for more. What organisations shouldn't forget is that consumers are using this medium to educate themselves and to ask and answer complex questions. As a result, consumers that call your organisation may actually have more knowledge than some of your employees, which then results in frustration and wasted cost.

Traditionally, self-service only catered for the "average" consumer, yet the technology can now segment and analyse customers to provide personalised experiences. This is a major boost for companies that want to offer a more comprehensive web service. For example, if I'm looking for car insurance, but cannot get an answer online because there is no way to ask if my particular circumstances are covered, I will have to revert to speaking to an agent. This costs the organisation a series of calls that could have been answered online, and costs me my "web discount" – so effectively results in a lost sale.

What do self-service technologies aim to accomplish?
From a user's point of view, self-service should be simple, speedy and satisfying.  From our own personal experiences we know this is what makes for a good self-service experience, be it a vending machine, a cash machine or a telephone transaction. From a service provider's point of view, self-service technologies shouldn't divert you from your core business – they are a means of extending your customer touch points to offer choice, capacity, flexibility and reduce costs. A good self-service solution reduces costs, but a great one also delivers a rewarding and effective user-experience.

How can a contact centre utilise its self-service technology to increase contact centre capacity?
It's about finding the right business processes, or parts of them, that are simple and repetitive enough to automate, and driving volume through this. This might be self-evident for certain businesses, but for complex technologies like speech recognition it usually needs an expert to analyse processes.

The important thing to realise is that a self-service channel and an agent are not the same thing, and do different tasks differently. Rather than using self-service to try and fully replicate your agents and create capacity that way, have it do the tasks that it is suited to, and take those tasks away from agents in their entirety. This changes the balance of how your run your contact centre, and the roles agents perform – to get maximum value from them.

Once again, it's about having a user-centric view. Getting it wrong can end up generating more work for your contact centre. This was one of the great "shocks" of the web for early adopters – they ended up generating even more support calls.  Businesses should also tread carefully when thinking about forcing customers to use a self-service solution. This means you may take your eye off the usability ball, and in the long run this could backfire. On the other hand, a well-designed solution will have customers sailing through it and completing tasks that used to require an agent. If you can automate 30 seconds of a 2-minute call, then you've effectively increased your capacity by 30%.

Call centre automation could save economy £23bn a year

I don't even need to think about this heading in order to come to conclusion it could very well be true. Call Centre automation technology has the ability to rip out upto 95% of the costs of conducting contact centre transactions, and yet it has been deployed and designed so badly in the past that it is almost universally hated by everyone. Almost everyone has a tale of woe about a bad voice self-service experience and it's even de rigeur for comedians to make fun of it.

When I saw Kevin Bridges at the Edinburgh festival this year, he was at it regarding a cinema booking line. (Which incidentally does have a fundamental flaw that I spoke about at a conference almost 10 years ago, and it hasn't been improved). It's exactly this kind of lip service to good design that needs to be challenged. And there's no excuse not to do so, and do so well, when the savings can be so high.

For what it's worth, I think Gartner are totally wrong on this one. The technology is absolutely mature enough for the big time - what isn't mature enough is the commitment to user-centric design rather than cost-oriented and departmentally siloed project mentality.

Here's some of the article:

In an interview with Computing, local government CIO Jos Creese said local authorities should be looking to move as many services as possible into self-service. However, a report from Gartner last month argued that the technology was not yet sophisticated enough, and that self-service struggles to solve more than one eighth of IT problems.

A step towards self-service, at least from the perspective of the consumer, is call centre automation.

A study released yesterday and carried out by the Centre for Economics and Business Research estimates that UK organisations and consumers could save more than £23bn a year by reducing inefficiencies in public and private sector call centres using call automation technology.

The study, called The Economics of Call Automation, said firms could recoup £14.8bn through call centre automation, and consumers could save £8.3bn a year through not being put on hold at call centres.

The potential public sector gains rise to £13bn per year if increased call automation is applied across doctors' surgeries, universities and government departments, as well as call centres.

 

[source]