Purna Virji is the Senior Manager of Global Engagement from Microsoft.
She says we’ve been obsessed with artificial intelligence for a long time now. Movies such as Her, Terminator, The Matrix propagate this. But it’s not just in fiction. For over three decades we’ve been obsessed with using AI to automate our lives. As an example, she shows us the following adverts from the 90s.
How did we advance so far from this? It’s because of cloud computing she says. It all goes in one place now, it’s no longer disparate systems. We can create the algorithms and create entities between the data, and with that we’re getting truly intelligent AI.
Let’s define AI
It is technology that can perceive, learn and reason, assist in decision making and act to help us solve problems. All this means search is becoming more intelligent, Purna says. It means that it’s no longer just indexing information, it’s now understanding the context and being able to reason with it.
As as example, she mentions her 11 year old son who loves video games. They had an argument whether or not video games are bad or good for you. She explains how they both searched for biased keywords, “are video games bad for you” and “are video games good for you” and they both received biased results based on what they were looking for.
However, with intelligent search – she describes how now you can put in a query that is bias and search will understand the bias in the question, and search will create good and bad points, no longer allowing people to live in a echo chamber.
That’s all well and good, but how can we make money?
Intent. Search is increasingly the centrepiece of consumer engagement. Purna explains how we see this with ML powered offerings such as In-market audiences. She describes for those in the audience that don’t know that an In-market audience is those who have displayed enough interest to be in the market to buy. We can use AI to find out intent and build these, and the tenfold numbers speak for themselves, she says.
Additional question: Does getting more insight conflict with the consumer’s desire to get more privacy?
She says, no. Microsoft never track or store anything or personal interest such as political or race.
How else can AI help?
Image search: you can click on a necklace an actress is wearing and you can then purchase it. People assume this is paid, but it’s organic. Optimise your feed and optimise your images as you always have. Put up an XML image sitemap, she find that helps as well, she said.
But, curb your enthusiasm, she says.
70% of consumers have never used voice search to shop for products – it’s currently for simple informational tasks.
Should we stop? No. We’re going in the right direction.
The three things she wants us to think about before she leaves:
- Better understanding = increased personalisation
- Expand your reach via audience targeting
- Focus on conversational tactics