All the latest news, views and random musings from across Engine. Got an opinion? Then leave us a comment or say hello on Twitter using @EngineLondon




Thursday last week was Engine London’s end-of-year “school assembly”. Joining in with the classroom theme I foolishly agreed to wear goggles and a lab coat but did manage to get away 5 marketing technology innovations that can be practically applied in 2017.


In just 2 years, voice search volumes have grown from virtually nothing to 10% of all online searches . Predominantly because it’s quicker - we type at 40 words a minute, and speak at 150.

And it’s this ease of interaction that means next time you visit a Marriott hotel, you could be using your voice to engage with their hotel services or find out what’s happening locally. That’s because the winners of the latest Marroitt accelerator will be trialing voice-activated personal assistants in one early adopter hotel. This is made possible through Amazon Dot (the hardware) and Alexa (Amazon’s software) - a combination that makes possible voice recognition and artificially intelligent conversations.

Through Alexa it’s already possible to pay the balance on your Capital One credit card, check the fuel in your BMW and a growing number of other ‘skills’.

Imagine how helpful it would be if B&Q offered hands-free help while you were putting in a new bathroom, or you could get your Experian credit rating without using a laptop...

Designing for this new medium requires new thinking - a voice UI needs to...

- Have an open style allowing for conversational turn-taking

- Be resilient enough to retain context and handle errors

- Adopt a character, tone and humour appropriate to the task

I’ve been involved in design workshops for voice UI and they’re as fun as they are fascinating.


One third of mobile searches are local - “near me”. And these kinds of searches are growing 50% faster than general searches on mobile devices. However, 80 to 90% of retail spending still happens in physical stores and our job (as the trusted partner of our clients) is to help them drive more footfall.

So we're excited that Google Promoted Locations/Pins looks like it will arrive in 2017. These will bring the next generation of local search ads that will deliver branded and customized experiences and hyper-relevant recommendations.

This kind of super-smart targeting is already possible using ad tech platforms Engine has in-house:

Rapport provides intelligent customer engagement and media delivery

Balihoo is a specialist location targeting ad platform Engine acquired recently

Imagine if we could help IKEA get customers to go in-store and convert just 1% the items left hanging in their shopping baskets, or if we could encourage Halford's cyclists to try new off-road routes based on the weather and where they live...

These tools enable the delivery of contextual content based on geo-location, day-part and purchase history. Which is why we're introducing our clients to the wizards in our media team, and recommending they put 10% of their spend aside for localised advertising.


Machine learning is the advanced application of statistics to learning - to identify patterns in data and use them to make predictions.

Siri on your iPhone has machine learning built-in - if you ask it to show pictures of snakes (and you happen to have some on your phone) they’ll be served up. You didn’t tag your photos with “snake” - Apple’s using image recognition and machine learning to identify them. What’s interesting is this isn’t heavily publicised, it’s provided as standard - machine learning is already being commodotised.

Recently 20th Century Fox used the computing power of Watson, one of the most advanced AIs in the world, to create a trailer for its new horror movie: Morgan. Another recent example is Mr Koike, a Japanese cucumber farmer who had a problem.

In Japan, each farm has its own classification for cucumbers - there's no industry standard - and his wife was spending up to 8 hours a day sorting them into 9 different classes. So his son trained a computer to classify cucumber images with a high degree of accuracy - using TensorFlow, open source software for machine learning.

Open source meaning it’s free, available to all and can be put into practice by the technologists and data scientists we already have in our business.

Imagine if Purina created a global database of pet sounds and allowed us all to find out what our furry friends are saying, or if we helped get Better with Balance by cross-referencing women business leaders on LinkedIn with FTSE 500 boards with and made recommendations on their next hires...

It’s not hyperbole to say that use cases for machine learning are only limited by our imagination. If you want to know more about this innovation area, take a look at Google’s recent 5-minute film: ‘Machine Learning: Solving Problems Big, Small, and Prickly’.


69% of snapchatters would add known brands as friends - but what then?

Well, Huffington Post uses that trusted status to share great content through WhatsApp - and achieves 3x higher click-throughs than seen in web browsers. Facebook Messenger’s ShopBot searches eBay for you matching uploaded pics with products (using image recognition). And on WeChat you can do everything from booking cinema tickets to configuring your new Audi.

Whether through pushed content, smarts bots or full digital customer experience, being a trusted instant messenger contact can deliver:

- Direct communication with consumers

- Enhanced customer service

- More efficient purchases

Imagine if Maplin had a personal assistant to guide you through their 100s of electrical products, or it was possible to get information about your Standard Life savings and investments using this approach...

It’s already possible to create these kinds of services with zero code, using drag-and-drop publishing tools to create bots with text, photos, videos, audio, GIFs, emoji, quizzes and polls. And if we want to get more technical, we can tap into Google and IBM Watson’s open source code libraries to build more intelligent services.

2017 looks like it’s going to be a year of growth and experimentation for messaging apps and bots.


YouTube just recently starting offering 4K live-streaming for both 360-degree and standard video, and Facebook Live is taking off in a big way. Social media has become the new primetime.

During the last Olympics, Sport Chek (a Canadian sports retailer) harnessed power of Facebook Live to showcase a new Adidas’ collection.

Their “Chase the Gear” campaign challenged Canadians to find athletes through check-ins on Facebook Live and chase them to literally grab prizes off their backs

What’s interesting about live streaming is that it gives brands the opportunity to meet customers “in the moment”. These engagements are truly phygital -  where the real, physical and authentic world collides with all things digital.

Imagine if Toyota started live-streaming their rally drivers or we helped E.ON make it possible for customers to watch hundreds of Christmas lights being turned on at the same time...

Thanks for reading - those were 5 marketing technology innovations that can be practically applied in 2017. If you'd like to more about any of these, hit me up on @AlastairCole and I'll do my best to introduce you to the right person at Engine.