facebook-pixel
  • Artificial Intelligence
  • Michael Baxter
  • JAN 25, 2019

How machine learning can help brands develop more personalised conversations with their customers

Machine learning can help brands develop more personalised conversations with their customers, says Mariángeles Noseda,from intive-FDV

How machine learning can help brands develop more personalised conversations with their customers image

How can machine learning create more personalised conversations?

It is more important than ever for brands to keep up a steady conversation with their customers. Those who become complacent with client communication could soon find foot-loose customers wandering in the direction of their competitors. As they say, out of sight, out of mind. That is why personalised conversations with their customers is vital. Machine learning can help make it happen.

The challenge that personalised conversations and machine learning can help fix

Churning out a steady flow of interesting, engaging content for different customer demographics over a range of channels is challenging. After all, the same blog post which clicks with a millennial is unlikely to resonate with a baby boomer and vice versa.

Luckily, if marketers have access to the right tools, the web offers a data goldmine which can help them plan, and develop communication strategies based on data points rather than guesswork. Many forward-thinking companies are turning to artificial intelligence (AI), and in particular machine learning to help them make sense of this data.

Recent studies highlight that almost 40% of marketing departmentsreport that they have already adopted AI into their strategies, and the majority of brands interviewed agreed that AI could make positive improvements to their overall marketing plan.

So, how can machine learning AI help marketers create exactly the right content which will resonate with their various audience demographics? How can machine learning create more personalised conversations?

experfy-blog  "Live chat: Backing up conversations in the cloud"

How can businesses backup conversations had by live chat conversations and chatbots in the cloud, as GDPR waits to pounce?

Personalising social media marketing with machine learning

Personalisation of messaging is extremely important for the modern consumer, but it’s a pain-point for modern marketers. In fact, recent studies show that more than 80% of marketers expressed difficulty in effectively personalising campaigns.

After all, finding the information needed to personalise messaging and content is not as easy as it sounds. Social media plays an important role in any marketers toolkit, but generally, marketers use social channels as an ‘outward’ facing marketing channel to reach the right people with the right information, whether it be through content, or targeted advertising.

However, thanks to advances in machine learning technology, social media could also be used to gain real insights about target consumers, which can drive future marketing and communication campaigns.

Machine learning AI normally requires large data-sets based to train an algorithm. However, matching ML with a huge wealth of information stored online on social media platforms could allow ML algorithms to be trained without using stored data.

In a recent slideshow, Sitecore architect Una Verhoeven shows how a machine learning  algorithm could use data insights farmed from social media profiles to create a recommendation engine for books on an online store. Using a similar system, marketers could effectively harvest information from the social media profiles of their target consumers to create personalised content strategies for particular customer demographics. By matching big data analysis with machine learning techniques, brands can not only work out what type of content will be most interesting for users based on their online activity, but also the demographic they are a part of, the style, and language which will resonate with them too.

experfy-blog  "Chatbots and human customer services; does artificial intelligence replace or complement humans?"

Gemma Baker, from Click4Assistance looks at chatbots in human customer services and asks whether they compete or complement each other?

Keeping the conversation flowing with machine learning powered ‘smart’ chatbots

Continuing the conversation with existing clients doesn’t mean just sending out the occasional email newsletter, and updating blogs once a month. The best conversations are two-sided, and the modern consumer expects to be able to communicate with their favorite brands over a range of channels, and devices, at any time.

A recent Forbes article highlights that 57% of enterprise executivesbelieve the most significant growth benefit of AI and machine learning will be in improving customer experiences and support.

Instead of chaining their customer support teams to their desks 24 hours a day, 7 days a week, many brands are instead turning to machine learning trained chatbots to lighten the load. These bots can be trained to deal with everything from scheduling visits to a store, to answering queries about commonly asked themes.

These goal oriented applications use similar technology to well-known AI assistants such as Siri, Alexa or Cortana, and are programmed to schedule an appointment with a human rep should they be asked something which they cannot give the answer to.

But there are many more uses for chatbots, there is more to personalised conversations than just just telling users when a store closes. Machine learning powered chatbots are increasingly being used as engines to not only communicate with users but also drive them towards content they are likely to find interesting, based on the conversation they have with the chatbot.

BetterBrand has created a bot and management platform which enables publishers and marketers to create and scale chat campaigns across multiple chat apps simultaneously and also to share different types of content (from blogs, social media, outside publications) to users via the same chat conversations.

Aside from chatting with users, the bot can also gain feedback for content they have shared with the user, and recommend further content based on their likes, dislikes of content which they have enjoyed in the past.

After all, if you want to find out what content users like, what better way than asking?

experfy-blog  "CIOs believe business will move to a ‘voice first’ strategy"

New research from Red Box suggests that businesses will pivot to a ‘voice first’ strategy within the next five years

Using machine learning AI to monitor and analyse existing content to drive future campaigns

But all this hard work shouldn’t end as soon as a brand publishes a piece of content, or has a conversation with a client. Instead, brands are increasingly using AI and in particular, machine learning powered content intelligence data management platform or DMP systems to gauge the effectiveness of certain content themes and channels and using this data to drive further campaigns.

A data management platform  imports, stores and compiles customer or target audience data from a range of various sources, and then cuts through the noise to offer actionable takeaways which will allow marketers to put the right messages, out on the right channels, at exactly the right time.

A data management platform can collect unstructured audience data from any source, including various different types of content, from blog posts to videos to social media posts. As more and more content is published, and more and more clients interact with the content, this offers marketers the chance to learn from their previous behavior, and ultimately predict the likelihood of conversation from different content types and themes.

Smart data management platforms can organise all of the data into segments — which in this application will most likely be based around client demographics — and then create detailed reports about the likes and dislikes of each customer segment. Aside from offering these actionable insights to marketers, the data management platform can actually be connected directly to content management systems (CMS) to pump out different types of content to different segments automatically.

So, if brands and their marketers want to have the right conversations with their different customers, it is high time that they update their tech toolkit. Making the leap into developing machine learning  applications in-house, or paying for the service via a third party vendor, will be a worthwhile investment over time. After all, McKinsey & Company highlighted that personalisation can deliver as much as eight times the ROI on marketing expenditure. And figures like that are hard to argue with!

The Harvard Innovation Lab

Made in Boston @

The Harvard Innovation Lab

350

Matching Providers

Matching providers 2
comments powered by Disqus.