Session-based matching

More conversational search means that you’re more likely to ask Google: ‘how old is David Beckham?’, rather than typing ‘David Beckham’s age’.

Your next search query might be: ‘What club does he play for?’ Google uses session-based matching to understand that you’re still looking for information on David Beckham, even though you didn’t specify it in your search.

In other words, it uses your search queries from a previous session to interpret what you’re looking for now. 

Take travel as an example. If you’ve done loads of searches for ‘skiing’, and then you search for ‘holiday’, Google’s going to give you results for ‘skiing holiday’. If you’re an advertiser and your keyword is ‘skiing’ or ‘skiing holiday’, your ad will show, even though the actual query was just for ‘holiday’.

This is a move away from traditional keywords, which is dominating how AdWords is developing. Yes, they’re still the core of how advertisers bid. But what’s becoming more important is how Google is wrapping other stuff around those keywords to improve context and meaning. 

It’s working out what we’re thinking. 

Moving from targeting filters to layers

For a while now, we’ve used filters to exclude groups of people from being shown ads: predominantly, location, time of day and what device you’re using. But with the switch to enhanced campaigns, these hard filters are shifting, and becoming layers that overlap, to improve targeting. 

For example, if you’re a shop that sells trainers, one of your keywords might be ‘Nike trainers’. You have a default bid for that phrase.

But you start layering information on top of your bid, to change what you bid for that term depending on various different filters, such as ‘is this person near my store?’ or ‘is my store open right now?’ to set your bidding strategy.

This adds multipliers to your bid: a person near your store, looking for your product, at a time when your store is open is worth a higher bid than someone who’s several miles away, at a time when your shop’s shut.  

Targeting by situation, not device

The biggest change as part of enhanced campaigns is that we no longer target ads by device. There’s still a tendency to think device first: does my website work on desktop, tablet, mobile?

But with enhanced campaigns, Google’s switching us away from device, and towards situation and context. You might be at home using a mobile, or out and about using a laptop. Thinking like that, choosing based on the device doesn’t make much sense.

What matters, is what information the person wants to see. That’s not informed by what device they’re using, but by the situation they’re in. 

Rather than thinking ‘what device is that person using?’ think: ‘how does that person want to interact with me?’ If they’re at home, researching your product, you might want to take them to your home page to find out all about you.

But if they’re out on the street, you want to tell them where you are, when you’re open, and where’s the nearest store.  The important decision is ‘do I send this person to my home page or to my store locator?’ 

It’s not just for B2C businesses

This is relevant for B2B businesses, too. As always, you start with your keywords, but your targeting layers might include ‘what device are they using?’ and ‘what time of day is it?’ to get an idea of their situation.

If someone’s searching for your company on a mobile phone at 6pm on a weekday, and they’re not at home, they’re probably commuting and still in a work mindset. Again, you can work out how valuable that bid is to you, and where to send them on your site.

If they’re searching from home in the evening, they might be a business owner or senior team member who’s really busy and working long hours from home. 

These aren’t just another different service from Google. They’re the beginning of a whole new way of approaching AdWords. For advertisers, they hope to build a profile of the person searching for your products, what their situation is, what they’re looking for, and what you can show them that’s going to deliver the most value.