AI & Automation: Takeaways from the Google Partners Masterclass 2017
Emmeline Dorsey Emmeline Dorsey
Analytics Consulting
November 9,2017

AI & Automation: Takeaways from the Google Partners Masterclass 2017

Each year, Google holds Google Partners Masterclasses to help bring agencies up to speed on:

  • New products
  • Changes/evolutions
  • New learnings in the search world.

This year’s Brisbane event was held in October and I was fortunate enough to attend.

The 2017 masterclass focused quite heavily on Artificial Intelligence (AI) and the benefits it can lead to via automation. In this article, I’ve covered 3 of the key takeaways that stood out.


1. AI is growing up

Google spoke about AI being in its ‘teenage years’ (i.e. not fully mature yet, but much more developed than its early days).

AI is smart enough now that we should be using it for automation. Ideally, this would reduce the number of manual tasks we have to do (like bid adjustments and ad split testing) and give us more time to work on strategy.

In other words: We should get AI to ‘paint by numbers’ while we humans focus on understanding which colours we need to paint.

If you need another reason to justify automation, consider the consumer journey. It’s more complicated than ever, because:

  • People frequently consult 10+ sources before purchasing something.
  • 90% of people with multiple devices complete a task by switching screens.
  • Consumers are 2 times more likely to be influenced by targeted digital ads.
Consumer journey is more complex than ever
Image credit: Slide 43 of Google Partners Masterclass 2017

And even though consumers are making themselves tricky to keep track of, they still expect tailored ads (i.e. ads that are timely, fit for their device, appropriate to their location, relevant to their interests, and a good match for their current intent). That’s a LOT of info for a human to manage; imagine trying to juggle all these factors for every single person who could potentially view your ads!

So yes, maybe leaving this work to AI isn’t a bad idea after all. Especially since it can:

  • Enhance traffic and click-through rate
  • Boost conversions
  • Make campaign management easier by optimising ad rotation (see slide below), among other thing.
Optimise ad rotation
Image credit: Slide 51 of Google Partners Masterclass 2017


2. AI is evolving fast

According to Google, AI is coming ahead in leaps and bounds. This is particularly evident in Google Attribution, which will be available in the paid version* of Analytics 360 in early 2018.

*We’re expecting this to appear in the free version of Analytics at some later point.

This will have a significant impact on how accurately and efficiently we can track, measure and optimise campaigns. Anything that makes this easier to do is great news too, because the phenomenon of fragmented data is working against us. Consumers have myriad channels, devices, platforms and content to select from, and this makes it increasingly tricky for us as advertisers to track them. So it’s exciting that AI is evolving quickly enough to help us overcome this hurdle.

Data Fragmentation
Image credit: Slide 108 of Google Partners Masterclass 2017

Data fragmentation: When 30% of customer journeys involve multiple channels, tracking them and serving the best ads isn’t so simple. But AI can and will help with this.


3. Dynamic Search Ads (DSAs) are still worthwhile

An important stat to keep in mind: 15% of searches are unique. That is, out of every 20 Google enquiries, 3 of those enquiries have never been searched before.

15% of searches are unique
Image credit: Slide 54 of Google Partners Masterclass 2017

What does this mean? Dynamic Search Ads (DSAs) are still important today. Or, as Google says, they’re old but gold.

Oh, and they deliver:

  • 30% more clicks*
  • 30% higher CTRs* (click-through rate)
  • 30% lower CPCs* (cost per click)
  • 35% lower CPAs* (cost per acquisition).

*For a typical advertiser, when compared to regular keyword performance.

DSAs help you answer enquiries that you couldn’t have predicted, while also saving saving time and improving performance. What’s not to like?


Wrapping up

There are literally millions of possible variations to consider when serving ads to people, based on factors including:

  • Time (hour, day, week, month)
  • Location (radius, postcode, region, country)
  • Device (mobile, tablet, desktop)
  • Audience (page visitors, cart abandoners, loyal shoppers).
Millions of variables
Image credit: Slide 83 of Google Partners Masterclass 2017

Why wouldn’t you take advantage of AI in every way possible to optimise your results and efficiency?

For our clients, AI means we can get better data and improve attribution, allowing us to automate and optimise (e.g. we can deliver even better optimisation recommendations). This also means we can spend less time on the nitty gritty manual work and focus on the bigger picture for our clients. Or, in other words, #WINNING.

All this, and AI hasn’t even fully matured yet. Exciting times ahead.

Analytics Consulting
November 9,2017
Emmeline Dorsey

Author: Emmeline Dorsey

Emmeline is a Senior Paid Media Manager at Search Factory | iProspect. She’s been in the industry for just over a decade now, probably because she’s a bit of an Excel geek and enjoys analysing data to gain insights and drive optimisations. In her spare time she enjoys spending time with her family; which with a four year old means lots of playing with Thomas trains, watching Cars movies and endless hours of anything and everything to do with Teenage Mutant Ninja Turtles (pretending to be ninjas, drawing them, practicing writing their names, playing with TMNT toys and of course watching the shows/movies).

Get In Touch
Get In Touch