“Alexa? What does the growth of Voice Recognition Software mean for the marcomms industry?”

And why not? For users – especially the more fat-fingered among us – it makes perfect sense. Voice recognition is quicker, more convenient and, in a world in which speed of response and convenience are highly valued, of huge commercial potential for both brands and services.

Voice-triggered computing is a dream that is as old as computers and robotics themselves, but until recently, has been the stuff of sci-fi. Work began in the early 50s, but despite big funding and the participation of the likes of IBM and Bell Labs, results were always disappointing and progress was agonisingly slow. The problem being that whilst computers can be staggeringly fast, the statistical and algorithmic ways in which they work cannot cope with the (still) poorly-understood ways in which humans formulate and use speech and language.

Trek geeks will recall the 1986 movie Star Trek IV: The Voyage Home, in which the crew of the USS Enterprise travel back two centuries in time to avert disaster in 2286. Mr Scott, confronted with a then state-of-the-art Apple Mac, attempts to get the computer to respond to his voice commands. He is shocked when he finds that he has to use the keyboard instead.

Even if that movie had been set two decades later, Scotty wouldn’t have had much more luck; it is less than 10 years ago that voice recognition started to take off as the technology improved. Interestingly, it was in the B2B sector where it was perfected. By around 2010, voice recognition software and hardware had become sufficiently sophisticated to allow it to be used in environments such as warehouses and wholesale, where it was used for tasks such as picking, stocktaking and goods inward. Hardware – such as headsets – and software were also sophisticated enough to be able to cope in noisy environments and with different accents.

It may have been the success of VRS in these industrial environments (as well as military applications) that prompted Apple to launch its speech-triggered “assistant” Siri, followed by Google’s Assistant. Spotting the potential, others have got in on the game – VRS is starting to be built into apps; Microsoft has integrated its Cortana into Windows 10, while consumer electronics giant Samsung’s Bixby system – which claims to go ‘beyond voice recognition’ – theoretically, anything you can do via a touchscreen or keyboard can be done with voice commands – is imminent.

And of course, there’s Alexa. Anyone who’s tried Echo, the Amazon device powered by Alexa, cannot fail to be bowled over by its ease of use, accuracy and sheer cute smartness. It feels just a small step away from the kind of AI (Artificial Intelligence) that has up until now been the preserve of futuristic sci-fi movies. As well as acting as a diary, and answering simple questions, it can do really useful stuff like controlling your lighting and heating. It can even run “digital errands” for you.

What Amazon’s done that is really smart is to think about usefulness, not tech specs (something that helped Apple become all-conquering with the iPod and iPhone). Tech friends tell me that Alexa isn’t as high-spec as Siri, but it’s not tied to a mobile device (an Echo can sit in the kitchen, study, hall or be moved around to where it’s most useful). It has been known to help the disabled by being able to undertake commands such as turning on lights and the television (via linked connected devices) and even call emergency services. It does, however, also have its downsides – being linked directly to Amazon, there have been reports of Alexa accidentally ordering stuff as it listens in. People have had unexplained Amazon purchases cropping up – including Rocking Horses – on their doorstep.

Most importantly, Amazon is treating Alexa as open-source. It needn’t be confined to just an Echo – it could “sit inside” or power any device with a mic and speakers. As we move into an “Internet of Things”, the big winner(s) will be those with a system that works best with the majority of connected home devices coming onstream. Amazon might well be up there with the winners – the more devices that use it, the more useful it becomes.

One can see why marketers, brands and agencies are getting excited about all of this. Especially if it’s open-source or built into all the gadgets we interact with – it no longer becomes a device or a program, it becomes part of life, of the environment.

But there could be downsides as well. Consider this scenario. You’re running low on chicken soup, your favourite food. You ask your device to order more. Naturally, it will order what you’ve ordered before (let’s say it’s Baxters Cream of Chicken). This is a nightmare for marketers, especially the people at Heinz, who would dearly like you to switch to their chicken soup, or at least try it.

Because the ultimate aim of marketing and advertising is to change behaviour (to get you to do or buy something, or to stop doing something), the marcomms industry (apart from the marketeers at Baxters, who will be delighted), would quite naturally find this an extremely concerning development.

If it’s so easy to re-order the products you already buy without even browsing other options, how likely are you to try a new brand? Not very, I’d argue. All of a sudden, the Internet of Things becomes a kind of anti-marketing. Rival brands’ hopes of catching your eye via packaging, a TV ad or a special offer are wiped out. If you’re able to have your favoured chicken soup delivered direct to your door without (literally) lifting a finger, why bother shopping around? And why, if you’re a manufacturer, bother spending money on branding and marketing?

Of course, you could put ads into the answers. So, when you tell your device to order soup, it could suggest that you try Heinz this time instead of Baxters. This is what happens when you do a search via, say, Google. You have the paid-for results at the top, the organic ones underneath. The two are clearly delineated and you can make a choice.

The problem is, on a voice assistant, this is incredibly disruptive, and in an age of ad blockers and folk fast-forwarding through ad breaks on their PVRs, we all know how intolerant people are of interruptions. I can see people shouting “JFDI” at their chosen device.

Fundamentally, advertising also undermines the whole point of these devices. If your assistant is constantly challenging your choices, how likely are you to interact with it? Does anyone want to want to live with a salesperson constantly pushing all manner of wares? These devices are supposed to make your lives easier, less complicated and consumer resistance to advertising on them is likely to be extremely high.

That doesn’t mean ads can’t play that part. Alexa/Echo could offer an opt-in setting to offer coupons when available, so that in the aforementioned example, “she” presents Heinz as an option because there’s 50p off this week. Or if a product isn’t available, brands could pay to have Alexa offer theirs as an alternative.

It’s a case of new tech, new channel, same problem – resistance to advertising interruptions; so you have to reward the consumer. Only in this case, the interruption is more irksome, so the reward has to be bigger.

Companies like Amazon, Facebook and Google operate what are fundamentally advertising (or ad-funded) businesses, and marketers are always looking for new ways to get in front of potential customers. They may eventually find it, but it’s difficult to see how much money could be made.

Until then, brands must rely on other avenues for advertising to make sure their product comes to mind.

As an echo (pun intended) to our article last week regarding the pressures on Google following the YouTube ad-placement debacle, perhaps the screen (or poster, radio or print product) isn’t dead yet!