In this guide, I’ll go into detail about how artificial intelligence is impacting marketing right now and how it will continue to impact it in the future.

At the end of this post, you’ll be excited about the possibilities of AI and probably a little nervous about the implications!

And it’s alright to be nervous because the role of marketers in organizations will change but….

…you’ll still have an important role to play.

Table of Contents

Download a PDF of this Guide


Would you prefer to read this later? If so, download a PDF version of the AI Guide.


C H A P T E R – 1

Introduction to Artificial Intelligence (AI)

Artificial Intelligence in Marketing is real and now is the time to sit up and take notice.

Artificial intelligence is accelerating marketing toward a more intelligently automated future in which smarter (i.e. AI-powered) solutions enable marketers to solve problems and achieve goals more efficiently. You have a choice. You can sit back and wait for the marketing world to get smarter and change around you, or you can embrace AI now and be proactive in creating a competitive advantage for yourself and your company.

Paul Roetzer, Founder of Marketing Artificial Intelligence Institute

However, not all software companies really have AI that say they do.

There’s just so much hype surrounding AI Tech companies want to capitalize on it by saying their software is powered by AI and investors will give higher valuations to them because of the AI in their software.

But there are many great software companies building true AI applications and this is set to grow massively over the next few years.

MRFR research predicted the AI market to be worth 25 billion by 2025.

AI marketing growth

If you’re a marketer, it’s time to get up to speed and understand the potential impact that AI will have on marketing. I’m pretty sure that this guide will help.

So, what is artificial intelligence?

We all know what human intelligence is…I hope so anyway!

Artificial intelligence is when a machine demonstrates some human-like intelligence.

For example:

A machine processes data and learns from it so it can make smarter decisions about the data it will process in the future.

Instead of just repeating the same instructions, the machine automatically learns new instructions based on experience.

Alpha Zero, the game playing AI developed by Deepmind, learned Chess in 4 hours and then was able to beat the best computer program available for playing chess.

Learning a new game is mimicking human intelligence, but the AI can learn in 4 hours what a human may take months doing.

Computer science describes the study of AI as the development of intelligent agents.

Look:

This is really about smart programming.

Our intelligence helps create artificial intelligence.

As some tasks become very routine they may not be considered artificial intelligence anymore.

Here’s an example:

Optical character recognition is often excluded because it’s a routine task expected from computers.

What is the difference between narrow and strong AI?

Narrow AI (also called weak AI) is artificial intelligence focused on one task.

Strong AI is everything else!

Strong AI has the ability to apply intelligence to any problem rather than a specific task.

For example:

A spam filtering tool performs one task well. A self-driving car is also described as narrow AI but I think this is a bit of a stretch!

Will Artificial Intelligence Replace Marketers?

Yes…. some!!!

Marketing is a time-intensive process with a lot of repetitive tasks which machines can help with…

…but there are certain tasks that machines will never be able to perform at the same level as human marketers.

I can imagine, in the future, sitting across from a robot discussing a business proposition but I can’t imagine I’d build the same relationship with a robot as with a real human.  It’s relatively easy to build software to beat someone at Chess and…

…the software gets better at beating people.

But…

Building relationships is the most important part of marketing and computers suck at it.

Also, who is going to build a strategy for a company?

An AI enabled machine can provide inputs into this strategy but strategists will still survive.

I watched a movie called ‘Her’ recently where the actor builds a relationship with an operating system.

Such a ridiculous movie!

Currently, though, there is a serious problem with implementing AI within organizations because of the lack of knowledge amongst marketers.

In a report done with CMOs (Chief Marketing Officers) by Deloitte in 2018, the major factor that could slow down the organic growth in marketing is lack of talent.

And because AI is more technical than most other areas of marketing, this is going to be a major issue.

Driving future growth in an organization

I wrote this guide because there is so much technical information on AI online that it’s quite difficult to understand.  I’m hoping this guide will help marketers understand what AI is really about.

Once you understand AI, then you can work out how to replace the systems you use internally with AI software. And if you decide it’s smart to replace the existing software solutions, you need to figure out what functionality will be gone and what new functionality will be added.

You’ll then need to educate your team about AI and train them on the new software.

Plus, the marketplace for AI solutions is growing so fast that, without understanding AI, you’ll have a hard time finding the right vendor.


C H A P T E R – 2

The elements of Artificial Intelligence

Artificial intelligence is a complex field that includes various elements.

It is focused on the following:

  • Learning – Acquiring information and rules for using that information.
  • Reasoning –  Thinking about something in a logical and sensible way.
  • Doing –  What’s the point in learning and thinking if you don’t do?
  • Self-correction – Understanding mistakes and correcting them.

Here’s a breakdown of the main areas that AI has been implemented in.

Note:  There are some overlaps in each of the areas. For example, a self-driving car uses a combination of machine learning, image recognition, and deep learning.

Neural Networks

A brain takes an input (external or internal), processes it and then produces a result.

A neuron is the basic unit of computation in the brain and it’s responsible for processing those inputs to produce the outputs.

Chemical signals are passed from neurons to neurons.

There are over 100 billion neurons, on average, in a human body and it’s an extremely complex web of interconnections between neurons. Some neurons can be connected to up 10,000 other neurons.

Imagine if someone was putting their hand near a hot stove. This is an input. The neurons would process this causing the hand to move from the stove.

Here’s how this would look internally:

The sensory neuron feels the heat, passing the information onto other internal neurons and eventually to a motor neuron which causes the reaction of moving away from the heat.

A single neuron doesn’t do much on its own, but using a complex web of neurons gives you amazing capabilities.

The neuron consists of input, output, and weight. Weight is really an indicator of importance in the overall scheme of things for this particular piece of information.

For example,  you want a machine to work out how valuable a car is.

You take in a range of inputs e.g. year, make, model, condition, mileage, etc. and these are passed through neurons. Each input is weighted.

The make and the model are weighted higher than the mileage or the year.

And then:

Through a series of complex calculations, the machine comes up with a result.

Here’s a simple example of a neural network.

The initial inputs are weighted (e.g. characteristics based on importance), they are then sent to the hidden layer for processing, and the result is the output.

Machine Learning

Machine learning is a branch of AI which enables computers to become progressively better at performing existing tasks or become able to do new tasks without any need for human intervention.

The computers are continuously analyzing data so they can produce better results in the future. Simply put, they’re becoming smarter.

Machine learning is typically broken down into 3 parts:

Deep learning

Earlier we talked about neural networks. Deep learning uses more advanced neural networks.

So instead of an input, hidden, and output layer, you may have many hidden layers.

Meaning there is a lot more processing done than with a basic neural network. The same system of weights is passed between the neurons.

Deep learning is typically categorized in the following way:

Supervised

Supervised learning is where you provide the computer with input data and then the output data (i.e. the results you’d expect). You then build an algorithm around this so you can start providing new input data and the computer will automatically create the output data.

For example, imagine if you had a spam filter. Instead of giving the computer a set of rules to determine whether an email is spam or not, you provide it with a set of emails and then tell it which of those emails is spam and why. The algorithm would then be used to work out a new set of emails.

Unsupervised

With unsupervised machine learning, you provide the input data but you don’t provide the output data. The input could be a batch of test data at first.

So, the computer doesn’t have any example data to help it generate the answers. It needs to do a bit more work.

Semi-supervised

This is a happy medium.  It’s not completely unsupervised but the output data is not enough to accurately predict all results.

So, the computer processes the data and uses the output data as a guideline that it improves over time as it processes more data.

You may want to use semi-supervised ML in cases when you have to manually classify the data but there’s so much to classify that you just classify a piece of it and leave the rest to the computer to deal with.

Natural Language Processing (NLP)

This is what natural language processing is about…

Alexa

Alexa is an Amazon device.

You ask questions in a conversational way and Alexa is able to process them and give a response.

Well, it usually is…..

Natural language processing (NLP) systems have become more advanced over the last few years but there are still many challenges.

For example, it wouldn’t be unusual to say the following:

Alexa – Who are Man U playing?

Manchester United supporters often abbreviate Manchester United to Man U or the Red Devils or just saying United. There’s a slim chance that Alexa would understand these abbreviations.

Here’s another challenging example for NLP:

“I was at a pub the other night with my mates and it was deadly.”

When we use the word ‘deadly’ in this context in Ireland we mean that it was great fun. NLP systems are still not good at detecting the sentiment of text or spoken word.

So NLP will continue to evolve but it will never be perfect because of:

  • Accents
  • So many languages, variations of languages and slang used
  • The tone of voice and body language

Evolutionary Computation

This is the definition of evolutionary computation from Wikipedia:

“In computer science, evolutionary computation is a family of algorithms for global optimization inspired by biological evolution, and the sub-field of artificial intelligence and soft computing studying these algorithms.”

But what does this actually mean…

It was called evolutionary because it’s a continuous process of optimization of results which ‘evolves’ better solutions over time.

It was also called evolutionary from Darwin’s theory of evolution.

For example, one of Darwin’s theories was about survival of the fittest. The weakest members of a species will die over time.

With evolutionary computing, you come up with many potential solutions to a problem. Some may be good and some may be completely random.

With testing, over time, the best solutions evolve.

With deep learning, we are focusing on models we know already. Evolutionary computing is coming up with solutions to problems where we don’t have any sample results we could use to help.

Vision

We’re talking about the ability of computers/machines or robots to see, process, and act automatically based on images.

AI for vision it’s generally split into:

Computer vision –  A computer extracting information from an image to make sense of it.

Machine vision – Machines using visual methods to improve things in areas such as a production environment. They could be visually identifying faults, reviewing food labels, and/or detecting flaws in a product.

Robot vision –  This is where vision is used to identify something to be worked on and the robotic capabilities perform the necessary action.

Robotics

Robots are physical machines.

Robotics is the field of study of robots.

Sometimes you’ll hear people talking about robots automatically creating content for marketers but these are not actually robots. There’s no physical robot involved.

Most robots do not have AI but this is changing.

For example, I used to own a robotic lawnmower called ‘Robomow’. The tagline was ‘It mows you don’t’. I actually used to sell them but that’s a whole different story.

Robomow sits on a charging unit and every few days it would come out and cut the grass. There was an electrical cable around the edge of the garden and the mower would go back and forth at different angles to the edges.  It recorded where it had been so it knew when everywhere was cut.

It even had rain sensors so if it was raining it wouldn’t come out to cut the grass.

But it didn’t have artificial intelligence.

For example, it could have learned about obstacles in the garden and built different routes based on those obstacles.

Unfortunately, mine just kept getting stuck underneath the trampoline…

…every time…

Look:

I’m not saying these devices are not useful.

But…they could be a lot smarter.

Expert Systems

An expert system is a computer program that emulates the human ability to make decisions.

i.e. it replaces the need for or supports an existing expert.

It typically contains a knowledge base with a set of rules for applying the knowledge to each particular situation.

With machine learning capabilities, it’s building its knowledge base over time and adapting or creating new decisions based on its working knowledge.

Speech Interpretation

In the not too distant future, it will be unusual for someone not to have a device such as an Amazon Echo in their home so they can voice questions and instructions to this device and get immediate answers.

Voice interpretation is getting better all the time and some of these devices are leveraging artificial intelligence to learn over time and produce better responses.

Imagine if a speech recognition system was able to predict if a sale was going to be generated from a call center and then make suggestions to agents to improve the conversion rate?

And they did this by analyzing the conversation and the acoustics in this conversation.

A company called OTO systems studied 4,000 hours of inbound sales conversations with 50% conversion rates.

They trained their deep learning models to capture the ‘acoustic signature’ of a successful sale.

They managed to predict 94% of the call outcomes.

They then implemented this system in a call center and seen a 20% increase in engagement with a 5% increase in sales.

AI Planning

According to Wikipedia, these are strategies or sequences of actions automatically created for intelligent agents, robots or unmanned vehicles.

So, its all about analyzing a problem and producing a plan of action.

AI planning is taking into account things like:

  • Dependencies – does one task require another task to be completed
  • Milestones – specific dates that have to be met
  • Constraints – for example, if you only have 10 people available you can’t throw 20 people at the problem.

When the plan and the schedule are created, it is automatically adjusted based on results and changes to inputs.

For example, if a resource is not available any more then the plan has to be adjusted.


C H A P T E R – 3

AI Applications in Marketing

There are so many potential uses of AI in marketing that would make it more efficient and help deliver better results.

We have talked about 1 to 1 marketing for many years and, even with advanced marketing automation systems, this is still not a reality.

But…with artificial intelligence, we have a much better chance of delivering what feels more like a one-on-one customer communication.

Let’s take a look at some examples of how marketing can improve with AI.

AI and Content Marketing

To survive on the web we need to produce content.

Content attracts visitors, engages our audience, and gives them an incentive to come back.

Content comes in many forms:

  • Blog post
  • Testimonials
  • Factual data e.g. reports
  • Video content
  • Tweets
  • Company information

AI will never take over the full role of Content Marketer but it can certainly help.

Can computers automatically create content that doesn’t sound like it was created by a computer?

Yes!

A 2017 report by Statista found that over 90% of people surveyed said that getting personalized content was ‘very/somewhat’ appealing’.

Attitudes towards personalization
Content personalization is on the rise

Its no surprise that people want to feel like you are providing information and content that is just relevant to them. They don’t care about anyone else!!!

Marketers don’t have the time to personalize all content but luckily AI can help.

Here’s how:

Content research

MarketMuse is a software platform that gives users guidance for creating the right content. It uses big data and AI to understand how search engines rank content.

It crunches all your data and compares with other companies’ ranking for similar content.

It then organizes your content into topic clusters, defining the topics that are easy to rank for and provides recommendations on how to improve your content.

Performing a content audit is a really time-consuming process and a software like this can save you massive amounts of time.

Here’s an example where MarketMuse analyzes the top search results for marketing tools. It extracts the most relevant terms within each of the top ranking content pieces and compares this with your content.

The tool displays the number of mentions of these keywords in competitor content compared to the number of mentions in your content. You get a content score that you can improve to rank higher.

Marketmuse

By analyzing your content, MarketMuse determines your ‘topic authority.’ These are the topics you could easily rank for by creating more content around them.

Content creation

Neurolinguistic generation (NLG) is a technology that transforms data into human-sounding narratives.

Automated Insights is a company that does exactly what their name suggests.

They analyze the data and automatically produce text that describes the data.

Imagine if you were in a stockbroking firm and you had to create 1,000 different reports for customers. That’s a dreadful thought, isn’t it?

Now, imagine clicking on a button and generating those reports automatically.

AI may not write a book or replace me as a blogger but it can certainly help a lot with content creation.

Content amplification

Content amplification is the process of promoting and distributing content through paid and unpaid tactics to achieve greater reach.

With so much noise online, even the most epic content won’t perform well unless you promote it.

Content promotion used to take up a big chunk of content marketers’ time but now there are some really smart tools out there that can help automate this process.

Here’s one example.

Inpowered is a tool that lets you select the content you want to promote across many native advertising platforms and then automates the process of placing the promotion and getting the best pay per click rates.

It will cancel promotions on certain platforms, increase promotions on other platforms, and analyze what’s working and when.

All fully automated.

This platform is interesting because the technology is very good and you only pay for engaged users. If someone views your content and immediately bounces you won’t get charged.

Content optimization

How about optimizing content to drive more traffic from Google?

In the olden days, you could stuff the same keyword many times into your article to rank.

But now…Google does semantic analysis of your content to understand what the content is about.

It uses machine learning (Rankbrain) to understand the content you write.

Also, it’s not just looking at keywords it’s looking at topic authority.

Here’s an example of how to demonstrate topic authority on your site.

You create a pillar piece of content like this piece of content.

You then create related pieces of content which link to the pillar content (and the pillar links to the related).

You may even take one step further and create guest post content on other websites linking to the related or pillar content on your site .

Pillar and cluster content

This shows topic authority which is more important than one post targeting a specific keyword.

Google uses AI to figure out your topic authority so it makes sense that we need tools that leverage AI to figure out if we are providing the right signals to Google.

This is what MarketMuse and other tools in this area do.

Content curation

A content curation tool is great for finding relevant content you are interested in.

For example, you set up a set of keywords and it finds content that is popular related to those keywords.

But….

…the AI version of the content curation tool takes an extra step.

Take Frase.io as an example.

This finds content but then uses AI to summarize the content so you don’t have to read it all.

I don’t know about you but that sounds awesome to me!!!

In terms of content curation, AI should assist in the following workflows:
– Making more targeted queries and removing noise when monitoring the media
– Summarizing information to help knowledge workers consume content faster and only dig deeper when relevant
– Identifying relationships between topics and drawing trends over time
Improved content curation through AI should help marketers create better newsletters, incorporate more research on their original content, scale their social media posting and create richer internal microsites. Digital publishers may use AI-driven content curation to automatically generate reports and enrich their editorial workflow.

Tomas Ratia CEO Frase.io

AI and Analytics

We typically break analytics down into descriptive, predictive, and prescriptive analytics, but let’s add a fourth dimension:

  • Descriptive – Looking into the past to understand what has happened
  • Predictive –  Looking at the past and figuring out what could happen in the future
  • Prescriptive – Figuring out what we should do next
  • Action-oriented – Automatically implementing, testing, and adapting.

Descriptive analytics has been around for a long time.

An example of this would be seeing Google Analytics data but not knowing what to do with it.

Predictive analytics gives you ideas of what you might do and prescriptive tells you what you need to do.

Action-oriented analytics is where actions are automatically taken and tested based on what is prescribed.

Sometimes I log into my wife’s Netflix account by mistake and most of the recommendations are not the movies I would watch!

But when I log into my Netflix account it always shows something of interest to me.

Netflix automatically groups people into different categories and ratings are based on the feedback within the category you are placed in.

So, when I see a percentage rating indicating how likely I am to like a movie, this rating could be different for my wife as she’s in a different category.

Netflix continuously tries to provide better recommendations to market better movies to their users.

But they don’t just look at the movie/show you started watching. They will also look at:

  • Did you watch some of it and stopped watching
  • Did you watch it over a couple of nights
  • When you watched it i.e. a month ago, a year ago, etc.

And, of course, much more.

These are machine learning algorithms that are learning over time and automatically adjusting.

A UK company called Datalytyx have patented an AI solution which solves a major problem of analyzing large volumes of data, for example, analzying billions of records.

It’s AI software identifies the most relevant 1% of the data and you run reports based on this.

AI and Marketing Automation

A typical marketing automation task is sending a series of emails to users after they opted in to an email list.

And then, based on their interaction with emails, route people to a different path.

For example, the click on a link about a new product in the second email in a sequence triggers a different email.

This is smart email automation but it’s not AI.

AI adds a whole new layer of intelligence. Here are some examples:

Watson is an IBM platform that uses AI to learn more about your data.

Watson marketing‘ is a part of the Watson platform focused on…you guessed it…marketing.

One of its components is creating targeted email campaigns.

It uses AI to understand more about each individual in the campaign and tailors the communication based on this data.

For example, instead of just putting people into a bucket based on a form they fill out, it pulls the data from many sources and creates micro-segments based on lifestyle, social behavior, life stage, location, etc.

But it will also continuously evaluate this data and automatically move people between segments based on new data and performance analysis.

When you are working with large data sets you need AI to automate certain tasks and make sense of data.

For example:

Compile data from many sources and create micro-segments based on lifestyle, social behavior, life stage, location, etc.

Discover flaws in original campaigns and change segments and offers based on this.

AI and Conversational Marketing

A chatbot is a computer program designed to simulate a conversation with another human.

There are many tools available (e.g. mobile monkey) which allow you to easily create a chatbot.

They have a builder program which allows you to automatically create actions based on inputs.

However, these chatbots are not AI-enabled. They are trained to recognize specific user intents and they tap into a knowledge base to retrieve answers (retrieval-based chatbots).

We’re still far from seeing chatbots that can provide users with an unlimited amount of answers that they can generate on the fly. This would be the true AI at work.

Most chatbots today operate in a specific niche and the amount of things that they know and can do is very limited. However, they still use NLP techniques to understand human language. The more sophisticated ones also use sentiment analysis to understand the emotion behind the user’s words.

Chatbots, as they are today, are still a very useful tool to help automate certain parts of the sales and marketing process.

For example, chatbots can:

  • increase engagement through personalized conversations with users
  • handle customer inquiries on your website
  • improve targeting by collecting useful insights about users

Now, for companies that already use chatbots on their website, there are tools that can help them understand how well they’re performing.

Liveperson.com analyzes chatbot conversations in real-time to assess when customers are having a poor customer experience. Companies can then take action based on this.

Live person chat

Not sure if we’ll use this…may come up with a new diagram.

AI and Email Marketing

Email marketing is one area that could benefit tremendously from AI.

Just think about it – an AI tool could help you determine which type of content you need to send and when you need to send it to increase your chances of converting an individual prospect.

Given the fact that AI can process enormous amounts of data in no time, you’d be running smarter and more efficient campaigns with a better ROI. Not the mention the time you’d save on A/B testing!

An email marketing tool powered by AI could also help with another challenging area for marketers – sending highly personalized emails at scale.

AI can take into account a customer’s history with your company and determine the type of messaging and offers that work best.

For example, Phrasee is an email marketing tool that uses AI to generate subject lines, body copy, and CTAs to encourage higher click-through rates and engagement on email marketing campaigns.

AI and SEO

Artificial intelligence has the potential to make search more human.

It means that search engines now look more at the meaning and the context of the searcher’s query to deliver more meaningful results.

The era of keyword stuffing is over. Search algorithms are now focusing on the user’s context and search intent.

And this is a good thing.

Marketers can also leverage AI tools to improve the ranking of their content.

Now you can use AI to improve your SEO efforts in a variety of ways, including:

  • Identifying content opportunities
  • Performing keyword research
  • Identifying opportunities for content optimization
  • Content personalization, and more.

AI and Social Media

Every time you log into Facebook and view the news feed you are seeing AI in action.

Facebook is continuously monitoring who you follow, what you interact with, how you consume content and more.

These algorithms learn over time to produce better news feed results.

Facebook is all about engagement.

If you spend more time on the platform they can show you more ads and they make more money.

It’s that simple!

It makes total sense to track what you interact and don’t interact with.

If you follow a Facebook page and never interact with the posts they publish, that is a sure sign that you have no interest in that page’s content.

Here’s another example of AI for social media.

Persado provides “machine-generated marketing copy to drive maximum performance in any channel.”

It picks out the best words, phrases, visuals and emotions to drive more engagement.

And social media is all about engagement.

With this social media module they will automatically create the text and find the best images that will drive the most engagement.

AI and Conversion Rate Optimization (CRO)

Conversion rate optimization is all about improving conversion.

For example, out of 100 visitors to your website you convert 2%, and then you make changes to your website and increase your conversion to 3%.

There are many ways to increase conversion:

  • Improve your ads so that you get a higher click-through rate and lower cost
  • Improve ads so you are sending a better audience to your offer
  • Build a different sales funnel, for example, add an up-sell option after someone buys
  • Change the pages that are part of the funnel e.g. colors, text, images, video, etc.

This is a very time consuming and manual process and this is where AI can help.

Unbounce is a landing page tool.

They recently built a pilot project around AI and included 34 customers over a 6 week period.

The AI analyzed the performance of the landing pages on real campaigns and instructed conversion specialists on what to change.

On average, the increase in conversion on the pages was 19.8% with one page achieving over 100%.

This is certainly a higher performance increase than you’d expect to get from working with a conversion specialist.

AI and Listening / Monitoring

Every company out there wants to be able to capture as much of the conversations around their brand as possible.

The goal is to understand not only what people are saying about their brand, products or services, but also how they feel about them.

This helps marketers to analyze their brand presence and use those insights to improve communication with their audience and target their campaigns better.

NLP and Sentiment Analysis can really help in this area.

Companies can use AI to understand conversations around their products so they can spot potential issues and act on them, as well as to uncover purchase intent.

AI and Image Recognition

We all know how important visual content is for marketing.

Now we can use AI and image recognition tools to analyze trends and uncover the type of visuals that would bring the best results on social media and other channels.

Image recognition allows marketers to ‘listen’ to what their audience is saying through images so they can deliver visual content that fits the interests of that audience.

AI can help analyze millions of social media posts and filter through the images that people share and engage with.

Without image recognition tools, it would be impossible for marketers to analyze this amount of visual material!

One example of this is the Image Insights platform from Brandwatch. This tool is focused on helping companies uncover how people are using images that contain their brand across social media.

It basically analyzes visual mentions of a brand’s logo across millions of social media posts.

AI and Influencer Marketing

Influencer marketing is a very powerful form of marketing but brands find it difficult to identify the right influencers.

With AI technology there are now smarter ways of analyzing and finding influencers.

For example:

  • Image recognition – AI can analyze thousands of properties of an image to find out what the image is really about.
  • Content analysis – AI can analyze influencer content to find out what exactly the influencer is passionate about and gets engagement for.
  • Assess engagement – AI tools can distinguish between fake and real engagement and analyze this level of engagement.
  • Influencer – Through the analysis above and other analysis it can work out how influential someone is and in what areas.

Demand for useful content from trusted experts is taking the marketing world by storm in the form of influencer collaboration and AI is playing multiple roles.

From AI powered virtual influencers on Instagram like @lilmiquela with 1.5 million followers to sophisticated AI systems used in influencer marketing platforms, the impact and implications of artificial intelligence on influencer marketing are just beginning.

Future applications of AI and influencer marketing include the ability to predict potential impact of certain influencers, content types and channel combinations as well as more advanced filtering of influencers with fake followers.

Lee Odden – Founder TopRank Marketing

C H A P T E R – 4

Security Concerns about AI

In 2018, the EU brought in a regulation called GDPR (global data protection regulation).

Its goal is to regulate the collection, storage, and use of personal data by companies without permission.

As consumers get more and more concerned with the use their personal data, I expect that similar regulation will be implemented in other parts of the world.

As AI is all about collecting and processing data this has serious repercussions.

Let’s say you walked into a supermarket and the supermarket used facial recognition to identify you and then tailored your experience based on the available data. Do they have the permission to do this? Not in Europe.

So, although AI is extremely powerful, some of it’s use will need to be approved.

Summary

There is a bright future ahead of us for AI.

It will have a huge impact on marketing for many years to come.

It will change marketing roles, it will remove some of them entirely, and it will provide a whole new level of sophistication which was never possible before.

Should you be concerned as a Marketer?

Of course.

You need to stay on top of developments in AI and see how you can incorporate it into your marketing.

You need to think about your role as a Marketer and how your role will evolve or be replaced in the future.