Ok Google!

Blau-weiße Tonwellen auf schwarzem Hintergrund
Author: Marcus Pentzek | Donnerstag, 18. Mai 2017
Category: Performance

How intelligent personal assistants are revolutionising the digital world

Since the new millennium, internet search engines have been displacing printed encyclopaedias bit by bit from their pole position in knowledge transfer. Fine dust is now settling slowly on computer keyboards and smartphone displays because the written word now has some competition: the intelligent personal assistant.

Until now, the associated technology of speech recognition has been in its infancy. In the future, however, it will present us with a whole new world of possibilities. It has already made a small-scale impact on our daily lives; we use speech recognition to activate apps, dictate messages while driving or make calendar entries. In the field of search engine optimisation, voice search opens up new opportunities – in particular the nature of searches should be more closely examined. 

Behind the friendly voice 

Speech recognition for end users has been around since the late 1990s – then in the form of software with a dictaphone function. Since 2011, voice control has also been available on smartphones. It began with Apple bringing Siri to the market. This was followed by rivals such as Google Now or Cortana from Microsoft. They are usually invoked with a so-called hotword such as “Ok Google” or “Siri”.

But there is still great potential for development in this field. The aim of the technology is to recognise and process natural language. And although speech recognition is already capable of classifying words semantically, the error rate cannot be dismissed out of hand. For example, dialects and neologisms are currently setting challenges for developers. However, this technology continues to be further developed.

The Chinese search engine, Baidu, must be considered the leader here. They have developed speech recognition software that can recognise the spoken word almost better than people themselves. This is a huge challenge in Chinese as the language has numerous dialects and only a few syllables that could be used to facilitate mapping.

Use of intelligent personal assistants

The technology of voice control is interesting to use both in one’s own home as well as on the move. However, there are significant differences between searches made via desktop computers and mobile devices. On a PC or tablet at home, people often visit numerous sites to obtain detailed information or to draw extensive comparisons between products. Other goals are more relevant when on the move: directions to a meeting location, opening times, weather forecasts or factual questions. Requests and results are mostly situation oriented and local. The user wants to obtain relevant information as quickly as possible. Digital assistants like Amazon Echo or Google Home offer new possibilities in ther way to use voice commands not only for search but to control lights, blinds or schedules.

As smartphones are used to control digital assistants the interaction doesn’t need to be situation oriented but can be used to plan ahead. The digital assistant decides respectively, whether the query has a search intention (in this case using a classic search engine like Google or Bing) or whether it’s a special request that is better served by one of the installed apps.

Larger market share – opportunities for smaller search engines

According to Statista, over 90 percent of Germans use internet search engines – market leader, Google, is particularly popular. Google can also defend this position among intelligent personal assistants and with its Google Now, holds first place ahead of the Apple competitor, Siri. In the next moths Google Assistant will replace the old app Google Now and strike a path to a new overall assistant.

These assistants have already been implemented on their respective smartphones. Siri sets the tone on Apple products and uses the search engine, Bing, as its source of information. Microsoft’s Cortana also uses Bing to find relevant answers. Google Now, however, draws its information from the Google search engine. The new home assistant Amazon Echo which is activated by the command “Alexa” helps itself by special search requests with information from apps that compete against the search provided by Google or Bing. has been developing app for finding recipes. While cooking the user can make Alexa read the recipe step by step.

But what influence does voice search have on search engine optimisation?

Search engine optimisation is becoming more and more relevant to intelligent personal assistants. Voice searches are often made while en route in order to obtain information relevant to the searcher’s situation. The structure of these requests is significantly different from those of written searches. Although, as a rule, these are limited to one or more individual key words, e.g., “Directions Cologne Zoo”, searches using voice function more like conversations. Requests are more linguistically sophisticated and complex. Searchers are increasingly asking open questions such as, “How can I get to Cologne Zoo by public transport?” or express their wishes in a complete sentence. This knowledge can be used for optimising your own web site as follows.

The closer the wording on a web site is to the searcher’s wording, the more likely it is that the search engine will consider that site to be relevant and include it in the search results. So an optimisation in terms of long tail keyword phrases or open questions promises potential. Instead of struggling through lists of keywords, open questions or Google suggest tools may provide interesting questions to include.
More frequently Google presents a Answerbox above the organic results that answers the question citing the source directly on the results page. For voice search queries SEO should be aiming to have its own website as source for such a box. Voice initiated search requests more and more a single spoken answer.
There are also special technical optimisation options. For example, the design of a page should be responsive – able to adapt automatically to match the screen being used. Alternatively, a mobile web site version will prove itself useful.
Content elements such as podcasts or audio streams can have a positive effect on the visibility of a page. Their content will not necessarily be evaluated by the search engine but they do send the signal that the page contains audio material relevant to the user.
To rely on SEO for Google means ignoring the trend of digital assistants. shows how it works and becomes native source for recipes in Amazon Echo. Other apps will offer companies the opportunities to become a source of their own. information supplier will compete for the best answer and SEO for the respective app is the key.

The future of the intelligent personal assistant – it could look like this

What still seemed futuristic to us a few years ago is already a reality today: spectacle lenses that we can use as display screens, personal digital assistants that process orders for us and many other innovations. But how could speech recognition develop in the future?

One question is how search results, for example from a news search, can be presented in meaningful speech. The idea that they should be spoken in a monotone by the intelligent personal assistant is neither an exciting nor a practical solution.
It is conceivable that only a reduced number of search results will be returned and the information limited to the essentials. It is likely that the search engine will make use of habits and proffer pages most often visited by the user. Should facts or definitions be requested then Google’s knowledge graph (in which Google collects and bundles up the most important information and returns it to the user) will doubtless come into play. Thanks to speech recognition, it would be possible to offer search results proactively – for example in the area of video chatting. The intelligent personal assistant would work in the background picking out key words from statements and then use them to collect information and keep it on hand. For example, you mention the name of a restaurant in the course of a conversation and get all the key facts about it such as opening times, reviews or the menu. Special request won’t be served by classic search machines but will be answered by special apps (named “skills” in Amazon Echo).
The challenge for SEO is to identify all business relevant queries and classify them according to their intention. Which query is best served by a classic search machine and which kind by which special app? Are there queries that can’t be answered well by Google and Bing and neither by Alexa and Google Home? In such cases it’s a successful way to close this gap by offering an own app and to promote it as useful skill for the digital assistants.

Smart-home technologies: The intelligent home = the butler of the future?

Digital assistants like Amazon Echo or Google Home have the potential to become man’s new best friend. They stand in your living room and respond to spoken commands. For example, they play your favourite song on request, read the latest audio book, or order dinner. They are also able to operate light switches and other connected devices.
Will this take off here in Germany? The problem in this country is users’ lack of trust and the resulting strict data protection laws. People feel uncomfortable at the thought of a device constantly listening to them and may even be recording and evaluating what they are saying. Perhaps this subjective trust in technology could be improved if the assistant were to appear as a projection on a wall or windowpane – to get a face, so to speak.
Digital assistants already make our lives more comfortable and there are many possible scenarios in which this could become more so in the future. Someday, they might perhaps open the door for a visitor before you get home, having identified himself by voice, facial recognition or mobile phone number. It is also conceivable that assistants could detect moods by interpreting certain characteristics of facial expressions, gestures or voice. Then they could adapt to them and respond to needs and desires accordingly.