How ML is killing the search box. Hello mobile
The way we look into information through our mobile phone is less about Google’s search box and more about how I discover through conversations, AI, chats, cards, content aggregation or new forms of queries if there is any query at all.
This makes sense if we are moving toward a predictive data driven world. But I wonder how mobile and conversational search will break the existing Paid Advertising Model. Do we still need page ranks as a user acquisition path? Do we move towards app within app models, bot interactions or maps as the channel to contextual discovery ?
Mobile is base on conversation and those conversation happening inside certain apps. If search was the center of web, chats could be the center of mobile and the new entry portal to the internet. Mobile brings us : 1.5 Billion people connected with sensors suggesting all kinds of data points, where you use only 3 to 5 social apps everyday and the browser is not one of them.
The power will come not from information but from who has the power to use my data and predict my next move without being weird about it. Search in mobile is not just information but how the data changes how we access that information, we are now able to ask more complex questions, give implicit queries or no query at all and just wait to be dazzle by some predictions. (Let’s face it Google Now is pretty cool sometimes)
The old structure built on top of browser will be obsolete soon and here is where Google has a threat: SEO — SEM — Page ranks are rules from the past and we are witnessing how they are trying to re invent themselves. Everyone knows they need a social ecosystem to take advantage of their identity platform capabilities and be part of the mobile wave, they need chats not just the OS or keyboard button.
Search in mobile is not about ads but location, is the opportunity to evolve from the page rank towards a stream of information that you will be suggested according to your context and not just your query. Google has such a robust data baseline they can and will make this happen, they just lack a social channel, even though they do have maps which is a huge starting point seeing what Baidu maps is capable of doing.
The interesting thing that happened this year, is that they finally came out of the closet in I/O when they announced how their next generation of platforms are going to be AI based : #Allo, #InstantApps #Ads in G.Maps #Google Home is just the beginning.
We will have to wait and see how the users responds to this and to whom, since every tech giant is moving in the same direction, it will interesting to see what Apple has to offer in WWDC.
Ref |
The 5 trends in the next search model that will define UI/UX :
- Implicit signals : There is an explosion of implicit signs coming form all the sensors and data tracking that is leading to more complex queries and granular results.
- Compound queries : queries are no longer single expressions of intent followed by static responses, check out “OK Google” or HOUND.
- Keywords vs. intents: Intents are not just better keywords but is determined from implicit signals and across multiple queries.
- Shift away from Web search to data search: Direct answers and cross device search are driving us towards a data driven model where cards are key.
- Personal assistants : From Search Engine to Personal Assistant
