The new rules of search engines: using data to predict real-world actions
Traditionally search engines were built on a paradigm that looks at the web as a topical world, where pages are connected by hyperlinks and anchor tags. This was the world of 10 blue links, a search engine built on a web of documents.
Cedric Chambaz
We are now engaging with the internet at an exponential scale, producing a social graph that is growing at an unprecedented rate. We are no longer just reading the web; we are building the web around us, making it more intelligent every time we’re online.
The average person has six connected devices in the home, and the increasing type of those devices such as wearables, connected cars, fridges, thermostats, glasses… I could go on… means the number of internet touchpoints is rapidly increasing.
This newly created web calls for a new connective fabric. For the first time, we have the technology to bring this wealth of data to life, understanding what people are looking for, when and wherever they are.
Bing stitches together pieces of the digital web to reassemble them into a clear picture – one that learns from the past to seamlessly connect to people, places and other things that users care about. For search, this is a major accomplishment. But it isn’t enough.
Today people don’t just want to type, they expect to interact with information in natural ways (hence the adoption of voice recognition and gesture control technology) and the same expectations apply to search. Why would you have to open a browser to search if you can search directly from your operating system or your hardware? Why would you type when you can speak? And ultimately, why would you have to search for information if it was given to you before you asked?
Search technology has had to mature, understanding where to point people from just a handful of words rather than perfectly written sentences. An evolution dictated by necessity. New functionality like query suggestions or semantic search were created and social, geographical or device signals were integrated into the algorithms. Search is now contextual.
Today, these interactions – such as voice search – remain contrived. Search will only truly become intelligent when the engine can anticipate what the person using it needs, even before the intent has been revealed. That is one of the promises of digital personal assistants such as Cortana who is powered by Bing intelligence, paired with machine-learning to anticipate someone’s needs. One of my personal favourites is when Cortana nudges me to leave for my next appointment by making sense of my current location and the traffic conditions to my destination.
So if that’s where we are today, could we take anticipation to the next level and let search take us into the world of serendipity?
Search engines understand intent, they receive more interactions than any other technology – people search things they might not tell anyone. At the same time, social networks are the depository of sentiments. If technology has developed the ability to process, analyse and understand these two humongous, historical and real-time information sets, it has the opportunity to discover user sentiment for certain events or entities, estimate popularity trends, as well as predict outcomes of future events.
Bing Predict uses data from social channels over-layered with search queries to understand the “wisdom of the crowd”. This machine-learned approach has proven to be more reliable than traditional statistical methods on several occasions. Bing accurately predicted the Scottish Independence Referendum outcome from the very first day whilst the official statistic was oscillating between the Yes and the No.
What can brands learn from these forward-looking experiments?
Machine learning models are already making their way to the advertiser toolset. Bing Ads for example, includes an opportunity tab which allows brands to evaluate the future impact of actions taken on their search marketing campaigns based on auction and competitive behaviours. That is just a first step.
Brands need to think outside the (search) box, and harness the full potential of search data to inform their marketing strategy and as a result, new advertising models will come to fruition. Trajectory marketing, for instance, would geo-target consumers based on the location they will be at rather than the location they are in. After all, marketing is about seeding the right message to the right audience, at the right time.
And that time is in the near future.
Cedric Chambaz is international marketing director at Bing Ads