5 things marketers need to know about machine learning

Blog | 07 Oct, 2016

At our recent Digital Journeys conference Microsoft, IBM and Google talked about digital, search and the future. The theme of the day was around digital no longer being a ‘thing’ as technology is an integral part of everyone’s daily lives. 

One common thread was the undeniable pace at which algorithms and machine learning are impacting the way different industries operate - as well as our experiences on the ground, as consumers.

One question lingered in my mind… what will this mean for marketers like you and I? 

1.    It will increase personalisation, relevancy and as a result, loyalty in marketing
2.    Customer service will be more automated – another *robots will replace people's jobs* alert
3.    It won’t compensate for new trends and innovation (yet…)
4.    It could be more accessible than you think
5.    You’ll have to adapt along with it, or better yet, drive it: The skills to develop

 

1. It will increase personalisation, relevancy and as a result, loyalty in marketing

 

Dave Coplin said of algorithms 'we no longer tell them what we want – but teach them through our behavior'. 

Algorithms use a ‘best guess’, whereas machine learning takes it one step further using personalised, varied interactions with people based upon their responses. Machine learning learns from each experience as it goes, introducing better topics and solutions all the time. 

Dave Coplin shared Microsoft’s current prime example; Cortana.  Cortana has conversations with people, explicitly answering back with more sophisticated and useful answers over time as it makes predictions based on users responses. Watch the video from 17:51.



An example of how machine learning will improve marketing is using the ever-growing data brands will be able to collate or access.  

Data such as what food and drink you’ve consumed in an evening, what you’ve watched, your internet searches, demographics, who lives in your household and your sleep pattern.

Brands will be able to use this information for better targeting.

Example 1 – A beer brand may want to target all the people whose team won at football and who drank a beer by giving away a celebratory free bottle. 

Example 2 - A sleep aid may want to target those who’ve had little sleep the night before, have searched to purchase a new mattress and are new parents by serving them a personalised discount with a tailored new parent message.

Essentially all of this learning will reduce irrelevant marketing experiences more and more over time.

 

2. Customer service will be more automated – another *robots will replace people's jobs alert*

 

How and why will consumers and businesses give their data? Freely! In exchange for services and products that enhance their lives. 

(Think your Google account hoovering up all your search data…)

You’ve likely seen some or all of these – but here are real-world machine learning examples of how it is transforming customer service:

 

Chat bots  

Not yet mainstream

Pizza Hut and Facebook, H&M and Kik and Skyskanner and Skype are just some of the retailer / messenger chat bot partnerships testing the machine learning approach to customer service.



They engage in two-way customer service, like Cortana. 

Job taken from humans – Customer service agents.

 

Uber 

Mature, successful tech consumer brand investing in machine learning

Uber use machine learning for ‘everything’ - i.e. they’ve built their own platform to concentrate on real-time data.  Efficiency is vital for Uber, as getting riders working on time and doing as many trips as possible makes them more profitable.

So analytics, R&D and their whole product can use their built machine learning platform to find out vital customer satisfaction data to free up their time to improve the product and customer experience so they remain competitive.
 
One great example is that their customers don’t like surge prices during peak times but drivers do. I wonder if they’ll just scrap the term and promotion of ‘surge prices’ and amend the cost.
 
Job taken from humans – Taxis **industry disruptive**.

 

The Self-Replenishing Fridge 

In development

Microsoft is developing a fridge that knows what it's holding and can monitor what you consume.
 
The machine learning algorithms it will use “will be able to learn new food types based on its experience from processing millions of generic food packaging images.”



The idea is it that it can then have automated re-stocking capabilities when something runs out. 

Watch this space for when they’ll be able to teleport the replenished goods into the fridge in your absence!

Job/s taken from humans –  Cashiers (the second hit following self-service), webmasters and website customer service operatives.

 

Not so exciting examples!

Traffic lights and your email spam filter.

 

3. It won’t compensate for new trends and innovation (yet…)

 

Machine learning can’t YET completely originate ideas out of nothing.

Given that 15% of Google searches are new every day,  we are in an era where disruptive business ideas are prevailing and marketing campaign ideas don’t just come out of thin air – we still need to innovate. 
 

4. It could be more accessible than you may think

 

Adopting algorithms or machine learning into your business to provide a better service and / or product for your customers may not be as complex as it sounds. 

This is because, like any other industry, demand will encourage supply and there are already Open Source machine learning platforms you can get your data scientist to look into.  

(Uber’s head of machine learning also shares his tips for approaching a machine learning strategy.)  

Failing that, Joseph Sirosh the corporate VP of the Data Group at Microsoft shared at a recent event:
 
“When the effort to build and deploy machine learning models becomes a lot less and you can ‘mass manufacture’ it — then the data to do that becomes widely available in the cloud. That cloud platform can then be like a department store.”

“We’ll have huge app store-like marketplaces for APIs and applications that can be used to build software to help automate more processes.”



So it sounds like it’ll be just like popping down to PC World!

 

5.    You’ll have to adapt to it, or better yet drive it: The skills to develop


Machine learning could even learn how to be a marketer. It could decide the best audience and messaging for your campaign, the best channel to promote it, through to even calculating the ROI!

We’re not quite there yet. Essentially machine learning will be the new way to free up CMO time to focus on strategy, campaign ideas, new partnerships and audiences. 

Therefore it could enable marketers to focus on driving growth and business strategy developing their place in the C-Suite thanks to not being bogged down by process and administration. 

In light of martech stacks growing in components and variants, knowing about and instigating new technology is never a bad thing as a marketer. 

Three experienced marketing leads at Digital Journeys emphasised that aligning yourself with CIO’s or IT provides a distinct advantage.  Namely for their knowledge, buy-in and technical support.

Aside from that the other main skills that will be a tick in the ‘compliments our ML strategy box’ are:

Creativity - Being the ideas person

Analytics - Not necessarily the skills, but the experience of mapping data and using it for strategy

People management / directorship - The number of technicians, analysts, technical project managers and marketing operations people may increase in order to use AI strategically.

I’m sure there's more, I’d love to hear your thoughts. 

All in all, greater autonomy and more relevant real-time marketing will be achievable with machine learning. 

But there’s still one thing it doesn’t have. And that’s soul. 

Please enter a valid Name.
Please enter a valid email address.
Please enter a valid Phone.
Please enter a valid Company Name.
  • Select Service
  • Analytics
  • Brand
  • Consultancy
  • Conversion Rate Optimization
  • Display
  • DoubleClick Partnership
  • Email
  • PPC
  • SEO
  • Social
  • UX
  • Video
  • Websites
  • All Services