9 Skills Marketers Need in the Age of AI

The robots are taking over. They’re going to be better than us, and then steal our jobs, and eventually turn on us and conquer the world. That’s the kind of hyperbolic language you’re probably hearing everywhere today. But some of it is true. Will the skills you’re learning today be obsolete by tomorrow? Is there a […]

The post 9 Skills Marketers Need in the Age of AI appeared first on CXL.

The robots are taking over. They’re going to be better than us, and then steal our jobs, and eventually turn on us and conquer the world. That’s the kind of hyperbolic language you’re probably hearing everywhere today. But some of it is true.

Will the skills you’re learning today be obsolete by tomorrow? Is there a point to learning new things if artificial intelligence (AI) is going to do it better? Is marketing going to become completely automated, and no longer require the human touch?

What Exactly Is Artificial Intelligence (AI)?

AI lets computers learn from experience and complete human-like tasks. Computers can take large data sets, catch onto patterns within the data, and carry out critical tasks.

AI has the ability to recognize sounds, faces, emotions, and objects, solve problems, understand languages, and plan. At the core of some AI is machine learning, or ML, which involves creating systems that can change their minds given the available data in order to execute a goal.

On a general level, AI can learn patterns and automate a variety of tasks, become more accurate with its results the more it’s used, personalize content and promotions for consumers, and chat with humans to help them achieve goals.

How Marketers Are Already Using AI

Some examples of AI uses in the marketing world include the following.

  • Stella & Dot empowers women to start businesses. They target three types of women: those who want to sell jewelry, those who want to buy online, and those who host sales events. The company wanted to optimize performances for these three separate audiences, according to Guy Yalif, CEO at Intellimize. They used AI to run more than 700 versions of their pages, including their shopping bag and product detail pages, to see which ones worked for each audience.They determined that changing the headline on the cart page to include emotionally compelling language contributed to a 52% lift from cart to checkout. Additionally, making the call-to-action stay on the screen as visitors looked through product photos contributed to an 8% lift in adding items to a cart. To drive overall engagement, Stella & Dot tested 25 different versions of headline text on a site-wide banner, and found that the headlines contributed to a more than 400% lift in engagement.
  • Epson created Rachel, a sales assistant/chatbot, to engage with visitors to the company’s site. With the help of AI, Epson discovered that visitors need to be contacted between six to eight times before they respond, and figured out what times of day visitors respond to best, according to Founder of the Marketing Evolution Experience & Digital Analytics Association Jim Sterne.Thanks to chatbot Rachel, Epson experienced a 240% increase in response rate as well as a 75% increase in “hot leads.” This resulted in a $2 million increase in revenue in three months.
  • HR GO, an HR recruitment firm, used Sentient Ascend, AI-powered conversion optimization technology, to test 1,080 designs on their site. They experienced over 153% more conversions, according to VP of Marketing for Sentient Technologies Jeremy Miller.

AI can make a huge difference, and we’re only getting started. AI usage is not going to slow down anytime soon.

In 2018, revenues from the AI market worldwide were $7.35 billion U.S. It’s expected to reach about $17 billion in 2020 and jump to $89.8 billion in 2025.

According to Salesforce’s 2017 State of Marketing report, 51% of marketing leaders are already utilizing AI. Seventy-two percent of high-performing marketers use AI for predictive lead scoring and product recommendations.

So if marketers aren’t needed to complete these tasks, then what are they left with? Can they compete with AI? How will they fit into this ever-changing landscape? What skills should they be focusing on to secure their roles now and in the future?

Building Your AI Skill Set

You shouldn’t worry about AI replacing you. Instead, you need to refine the skills that AI does not have and the tasks it cannot perform.

“Just as Photoshop has not replaced artists, Word has not replaced writers, and Excel has not replaced mathematicians, machine learning enabled tools will not replace people but will take over tedious, repetitive functions,” says Sterne. “It will augment the work that people do rather than obviate the need for people.”

Do you want to ensure you’ll be able to keep up in this new era of AI? Then do the following:

  1. Sharpen your soft skills. This includes your emotional intelligence and communication skills so you can deepen your relationships with customers and clients.
  2. Understand your customers through the use of qualitative research.
  3. Keep learning how to use and analyze data. You need to know which data sets should be analyzed and tune out any data that doesn’t.
  4. Go towards math and analytics – don’t shy away from them, so you know what to do with the data when you see your results that AI delivers.
  5. Create content (written, video, audio, etc) that your audience will love and will drive them into the sales funnel.
  6. Weigh and handle the privacy concerns, since AI is privacy-blind and consumers are increasingly concerned with privacy.
  7. Look at how AI will serve all your business functions, and form a bigger picture for how it will fit in.
  8. Come to business conclusions with the help of AI, since AI cannot do that for you.
  9. Think about new ways you can use AI to push your business forward and grow. What new applications does it have in store that will boost your business?

Let’s deep dive into each of these skills and how you will use them in the age of AI.

#1 Refining the Soft Skills

AI can take incredibly large data sets and analyze it better than any human ever could. It can come up with valuable predictions and become more efficient than your average marketer at certain tasks. But it can’t personally connect with a customer or interpret a client’s emotions. And isn’t that a huge part of what marketing is all about?

As Raviv Turner, CEO of CaliberMind puts it:

“All the technology in the world probably won’t help you as a marketer if you don’t have the soft skills such as empathy, communication, and accountability, as well as the creative mindset which is required not only to understand your customer, but also to communicate and explain marketing to the c-suite, starting with your CFO & CEO, that often see marketing as a cost center vs. a revenue center.”

Rebecca Horan, a brand strategist, says that because of the human element, marketers will always be necessary, even as the technology becomes more advanced.

“They may be needed more than ever to bridge the gap between automation/AI and human connection. Great brands are built not just on efficacy, but on connection with the consumer. We align ourselves most passionately with the brands that represent who we are, and who we’d like to be. That connection is sometimes nebulous and hard to pin down, but it almost always begins with emotion. I’m of the opinion that a human is still needed to assist in forging that bond between brand and buyer.”

Strengthening communication skills is key, as is learning how to do segmentation, automation, data, and analytics, and pairing them with a customer-centric mindset, says Turner. “A machine can’t provide the soft skills such as empathy, curiosity, and personal communication that successful marketers posses.”

#2 Understanding your customers through qualitative research

AI can be part of how you understand your customers.Y

For instance, you can use AI to collect tons of data on how customers interact with various parts of your website, including your headlines, product pages, and shopping cart experiences. ou can collect data on thousands of calls between your sales team and your clients, or your customer service representatives and your customers, and see what pain points they have.

AI can show you how customers get through different points of your sales funnel and what interactions they have along the way.

But AI and raw data alone aren’t going to help you understand your customers. You have to look at the data it gives you, and then decide what elements and strategies to test.

Yalif says that marketers need to strive to understand their customers, no matter how advanced AI becomes:

“Spend time speaking with and studying the needs of your customers. Look for chances to walk in your customers’ and prospects’ shoes. Experience what they experience with your brand, your product, and your website. Think through what it is like for your customer to go from prospect to interested buyer to customer to repeat customer.”

#3 Choosing the correct data sets to analyze

Your customers are going to generate a ton of data. But it’s a waste of your time to use AI to analyze all of it. You must be able to identify which data sets should be considered and can give you the outcomes you’re searching for.

“Too little data results in highly confident answers that are wrong,” says Sterne. “If you flip a coin three times and it comes up heads each time, the machine will predict heads on the next flip with 100% confidence. An insufficient variety of data will leave blind spots. The machine will recommend reaching out to customers who haven’t been in touch for years, without knowing that they have moved or died. Too many types of data will simply be noise and cause the machine to waver on any result. It will predict that all outcomes have the same chance of success.”

#4 Getting up to speed with analytics, coding, and math

AI is excellent at identifying patterns, and it may eventually be able to perform all the hard skills of today. But marketers will still need to figure out how to use the data to propel their businesses forward.

According to Turner, you should stay curious about your data and technology. That means going on marketing forums and looking at code samples and video tutorials.

“It’s fairly easy to learn the basics of machine learning using Coursera or other online courses,” he says. “I see marketers failing with predictive marketing tools because they don’t even have a basic understanding of how machine learning works or what data they need to be successful.”

You also have to be excellent at math so you can make key business decisions after reviewing your data.

According to Mike Moran, former IBM Distinguished Engineer and Senior Strategist to AI marketing technology startups Converseon and SoloSegment, “If you are going into the marketing profession as a refuge from math, that’s likely a mistake. You at least need to be comfortable making decisions based on data, even if you’re not the one collecting or calculating the numbers yourself.”

#5 Creating engaging content for your audience

Content marketing is highly important today: 60 percent of marketers are putting out at least one piece of content per day. It has proven its value, since it costs 62 percent less than traditional marketing and can generate around three times as many leads.

In the future, content marketing will continue to be critical: By 2020, content marketing spend in Europe alone is going to reach 2.12 billion Euros, up from 740 million Euros in 2014.

AI will work hand in hand with content marketing. It can deliver personalized paid ads that surround content, and even curate content itself. But it can’t write blog posts, make videos, or take photos that are going to entertain, inform, and connect with customers.

Director of Digital Marketing at Solodev and DigitalUs Wes Marsh says that in 2020, 2025, and beyond, marketers will still need to know how to be great at content.

“Content is king is a motto that’s been around for decades, but it won’t go away any time soon. The ability to create relevant content for unique audiences will be just as important in the 2020s as it was in the early years of the millennium.”

#6 Understanding the ethical concerns

AI may be highly “intelligent”, and it can collect a huge amount of consumer data, but it doesn’t have a conscience. As marketers have experienced with GDPR, ethical concerns are going to be key moving forward. Marketers have to be able to communicate how they are protecting consumer data and make their customers feel comfortable.

“Ethical considerations will also be paramount as we navigate privacy and automation issues,” says Courtney Herda, CEO of Smarter Searches. “This will demand a certain ethical understanding and moral compass from the marketing industry in previously uncharted waters.”

#7 Looking at the big picture

AI is only one part of running a successful marketing firm. Marketers also need to tie all their findings together and see how they fit into the big picture.

“The data hounds are the ones that will be able to drive the most value not just by crunching numbers, but by identifying the key trends, casualties, and relationships that businesses can leverage to improve results,” says Marsh.

When you combine the number of tools in the MarTech stack with AI and ML, you can become better at marketing, according to Marsh. “However, being able to see how various marketing activities impact each other and overall results is still a relatively rare skill, and tremendously valuable.”

#8 Codifying analytical thinking skills

AI is going to outperform humans when it comes to tasks like reviewing campaigns and finding the low performers, as well as analyzing marketing segments and picking the ones to target, says Moran. But it can’t come to business conclusions for you.

“What you should be doing instead is codifying the thinking that leads you to your conclusions, because the machines can do these jobs better and faster than you. Instead, you should be aiming at higher-level thinking around building the right teams, thinking up the new ideas, and coming up with that better, simpler marketing message.”

Moran continues, “AI does not automate jobs, it automates tasks. So fill your job with more tasks that are hard to automate. If you do, I think you will always find more of them as time goes on.”

#9 Adopting a growth mindset

AI can’t suggest new ways to use AI to propel your business forward. Only you can think creatively about how to leverage it for your business needs.

“Having a growth mindset means believing you can improve your skills and value by continuously learning and improving over time,” says Yalif. “This usually means fostering a spirit of curiosity and continually testing, learning, and iterating. In the context of this discussion, a growth mindset may mean developing new skills so that you can better leverage AI to do rote work while you ideate more and deliver better results for your company.”

Instead of manually fine tuning email and ad campaigns, constantly testing different strategies, and using a plethora of platforms to run analytics, marketers will need to go old school in this new landscape.

Miller says, “I predict the future of marketing for the professional to look a lot like marketing of the past. [This means] less time spent on pulling the levers of different analytics tools, DSPs, and marketing automation platforms and more time spent on strategy and planning, ideation, and creativity.


Above all, learn how to love AI and work with it. It’s here to stay, and it’s here to help. Learn how it can complement you and you marketing activities. Determine how it can boost your current efforts and amplify your marketing spend. When you combine AI with human skills, you can achieve fantastic results and boost your bottom line.

The post 9 Skills Marketers Need in the Age of AI appeared first on CXL.

How to Build a Top 1% Growth Team 3x Faster than Your Competition

Discover insider secrets on how to successfully build a high-performing growth team. How can you leverage CXL Institute, Kolbe tests and other online resources to genuinely impact your bottom line? Which skills truly matter for fast growth, and which are the nice-to-haves? In this conference talk (from Elite Camp 2018) by Chris Out from Rockboost […]

The post How to Build a Top 1% Growth Team 3x Faster than Your Competition appeared first on CXL.

Discover insider secrets on how to successfully build a high-performing growth team. How can you leverage CXL Institute, Kolbe tests and other online resources to genuinely impact your bottom line? Which skills truly matter for fast growth, and which are the nice-to-haves?

In this conference talk (from Elite Camp 2018) by Chris Out from Rockboost will show you a behind-the-curtain look on how they operate at RockBoost, and how you can leverage this knowledge for your agency or in-house.

But more importantly, you will see how this translates into running more growth experiments; at a faster pace, and that win more often.

Play the video here:

[This post contains video, click to play]

The post How to Build a Top 1% Growth Team 3x Faster than Your Competition appeared first on CXL.

Top 5 Insights from Every Speaker of Digital Elite Camp 2018

Elite Camp 2018 (10th anniversary!) brough together 180 marketing and optimization people all over Europe. It was 3 days in a secluded beach resort with the best speakers and parties. Here are 5 (or so) thoughts from every speaker from this year’s lineup. Peep Laja: Repeatable Patterns in CRO We cannot predict what will work. […]

The post Top 5 Insights from Every Speaker of Digital Elite Camp 2018 appeared first on CXL.

Elite Camp 2018 (10th anniversary!) brough together 180 marketing and optimization people all over Europe. It was 3 days in a secluded beach resort with the best speakers and parties.

Here are 5 (or so) thoughts from every speaker from this year’s lineup.

Peep Laja: Repeatable Patterns in CRO

  • We cannot predict what will work. Our intuition is terrible at it.
  • You can’t copy market leaders or competitors to get ahead.
  • Can we skip conversion research if we know the repeatable patterns? No.
  • Basic things like “less unnecessary forms fields” are best practices, it’s a stretch to call them repetable patterns
  • The only repeatable pattern in CRO is doing the hard work of conversion research – it gets results every single time.

Hana Abaza: Thriving on Change, Driving Growth and Lessons Learned at Shopify

  • Positioning: Is it really the product you have to differentiate or is it the experience?
  • Your customers are ready for the future of commerce. Are you?
  • Marketers are often talking about unicorns, ice-creams and rainbows. But actually there is no basis to positioning. Use plain language. No jargon. No ice-cream, unicorns, rainbows.
  •  Balance short and long term impact. Low hanging fruit seems tempting, but can lead you astray.
  • It’s not about more leads, but better leads.

Chris Out: How to Build a Top 1% Growth Team 3 times faster than your Competition?

  • The growth system is broken: product, marketing and sales are siloed. That’s a problem.
  • Growth hacking is not only top of the funnel. You need high tempo testing and experimentation throughout the whole customer journey.
  • A high impact teams needs top skills. Rockboost process of building up a skill set:
    • Personal T-shape plans.
    • What do your clients need?
    • List all the hard and soft skills.
    • Learning plan per person
    • Plan dedicated learning time into your day!
    • Monthly check-ins.
  • Look at your growth team score and ask yourself: are you working in a multi-disciplinary team where you focus on high tempo testing on the entire customer journey for bottom line and valuation impact?
  • Use CXL Institute to train your team members on growth

Ed Fry: Data-driven Growth – Lies, Lawyers & Outsized Results

  • Where is your customer journey captured? Go a little bit further, it is not always in Google Analytics. What does your team need to access that data? How to get access to the right data? It’s in website analytics and emails, but that is just a small piece of it.
  • What are the decision making moments in your growth process? Can you create rules for them? There is automation & there are processes, but for growth you also need rules.
    • Pricing is a rule.
    • Sales compensation is a rule.
    • Content modelling is a rule. (Booking.com: user reviews, location..)
    • Content modelling for a blog: use different elements
    • Design = rules.
    • Development = rules.
    • Content = rules.
    • Segments as rules: “Who to talk to”
    • Templates as rules: “What to say” and “What to say” internally
    • Workflows as rules: “When and where to say it”. Control the complexity in your workflows.
  •  Unify all user data in one place.

Alexa Hubley: The Agile Marketing Playbook

  • Create your agile process:
    • Map (start at the end)
    • Sketch (remix & improve)
    • Decide (Rumble, storyboard)
    • Prototype – test
  • Active campaigning works. If you want more product adoption, market to your user base. 
  • Solve at the micro level.
  • Show, don’t tell.

André Morys: Understanding Disruptive Growth – Why Most Optimizers Fail to Produce Great Results and How to Change it

  • Understand the real challenge. It is not statistics, tools, errors on websites.
  • It’s all about customer experience. We have to help companies provide a better experience. Make a connection between A/B testing and your boss/strategy. It’s not about technology, but customer centricity + agility, data drivenness.
  • You are not optimising websites but helping your boss and client get over ignorance and see the real problem. Connect what you are doing to their strategic challenge.
  • Prioritise impact over speed. Go for “High Impact Testing” – those tests need a triple amount of effort of an average A/B test, but are worth it. Challenge your prioritisation. Select tests that make a real impact.
  • Have a workshop with the management to agree on how to report real ROI in a way they understand and care about.

Karl Gilis: Why You Fail at Digital Marketing

  • Hope is not a strategy. You have to know the basics – why something works or doesn’t.
  • Offline marketing is about getting attention. Online marketing is about paying attention.
  • Video backgrounds are the new sliders.
  • Zoom in into the problems. Don’t make it about you. Make it about them.
  • If you don’t care about words, you are a decorator, not a designer.

Craig Sullivan: Tools and Techniques for Optimising Cross-device Experiences

  • If customers cannot read the content, because it is too small, then it’s a marketing whiffle.
  • We all have product defects, we just don’t know where they are and how much they cost us. But until you have tested it, they are just bad assumptions.
  • Why don’t we hear about these bugs? Even if nobody complains, it does not mean everything is working fine. Everyone needs a process for finding the defects. Customers will not call.
  • Most important thing in the checklist: audit of Google Analytics. Otherwise you might have bullshit data, bullshit boards, bullshit dashboards, bullshit executive boards.
  • Data-driven is a tricky concept. Information does not tell you what to do. You are the lens that needs to figure that out.

Annika Oorn: Optimizing High Converting Websites

  • Aggregate data is crap as it hides the gold inside the segments
  • Optimisation is more than just running tests. Find bugs. Get started and move on to automated solutions.
  • What if you don’t have heaps of traffic?
    • You can still do personalisation
    • Look at broad segments
    • Learn from segments – dig deep
    • Cross-sell/up-sell
    • Use micro-conversions
    • Increase motivation
    • Qualitative research
    • First impression tests, user testing before going live (UsabilityHub etc)
    • Lower the statistical significance
  • Focus on upsells, cross-sells and personalization when the conversion rate is already very high

Andy Carvell: Driving Impact on Mobile

  • Apple App Store and Google Play store have a lot of competition. You would probably find 6-7 functionally identical apps for every idea.
  • Mobilegrowthstack.com – A framework for strategic mobile growth.
    • Acquisition
    • Engagement & Retention
    • Monetization
    • Analytics & Insights
    • Tech
  • Push notifications. Pretty saturated. In-App messages. Definitely not saturated. Segmented targeted interaction with your user.
  • Use of in-app messaging to rapidly test segmented onboarding. Impact = Reach x Relevance x Frequency.
  • Optimise relevance – you can improve it with personalisation. Leverage the demographics, behavior. If the relevance is high enough, people are happy to get the notifications.

Jonathan Epstein: From Darwin to Digital Marketing: Can Evolutionary AI Create More Effective Customer Journeys?

  • In nature, natural selection has optimal designs. Each species is uniquely optimised for the niche it is in. Modelling evolution this way helps to bring the model to other areas.
  • Evolutionary principles
    • Fitness – The fittest web page, the fittest radio antenna, training system..
    • Combination – if you have 2 better than average design, then you climb a performance hill
    • Mutation – like in nature, we are looking all the angles of possible ideas.
  • Evolutionary Optimization: parallel designing – combine two good designs and get a better one. Each generation requires less traffic than a single A/B test. In 15 generations of 40 designs each. This approach allows you to test much more things. They run 6-8 generations to get highest increase of conversion over time.
  •  Combination of evolution and deep learning. Neural networks connect inputs  (variables: customer profile, device, day of week, time of day) and outputs (what are they going to see).

Lukas Vermeer: Democratizing online controlled experiments at Booking.com

  • The question with data always is: How did this data come about?
  • Evidence-based customer-centric product development. You need to have theories about your customer behavior and ask what this test wants to achieve?
  • Failure is learning. 9/10 tests fail.
  • Take the biggest small step so you can challenge your riskiest assumptions quickly.
  • Customer-centric evidence on what they care about. With this approach you will learn what matters to your customers.

Momoko Price: Data-driven copywriting for brand-spanking new products

  • Worst advice you get for converting copywriting – tweaking random words on pages.
  • Longstanding conversion-copywriting myth is that conversion copywriting is AB testing, copywriting formula. Actual four steps:
    1. Research the customer mindset
    2. Map out the sales narrative
    3. Leverage cognitive biases – framing, anchoring etc- how humans make decisions
    4. Measure the impact
  • Good copywriting = Exercising empathy. Listen. Listen at scale.
  • When you feel it in your gut, you know it must be right. No – that’s confirmation bias.
  • Great technique for copywriting – online review mining. Instead of writing your message from scratch, steal it directly from your prospects. Go to a review site and steal from there – Tripadvisor, Airbnb, Amazon

Ivan Bager: Storytelling with data

  • Storytelling is good for idea pitches, one-off analysis, board meetings, sales efforts, persuasion of stakeholders.
  • When narrative is coupled with data, it helps to explain to your audience what’s happening in the data and why a particular insight is important.
  • When visuals and graphs are applied to data, they can enlighten the audience to insights that they wouldn’t otherwise.
  • Connect the narrative to your story by linking it to events and conclusions in the data. Visuals doesn’t have to be graphs, use images of the people involved.
  • Analyse your audience member’s frame of mind to help them better “hear” you. Structure your story effectively. Instill customer empathy into your audience to increase your story’s memorability.

Robin Langfield Newnham: Optimising for Voice AI in the Post-Website Era

  • Post-website era: No app. No browser. No search. Voice-only shopping. Voice shopping estimated to hit 40 million dollars by 2022 in the UK and US.
  •  Use keywordtool.io to identify long-tail questions with voice intent. Look for long-term keywords.
  • Use SEMRUSH or Ahreds to identify which keywords contain a featured snippet result page.
  • Create in-depth, mobile-friendly guides that succinctly answer each question.
  • Use ‘organization’ schema.org markup to gain a knowledge box snippet. Allows Google Home to pull answers about your brand.

Els Aerts: Without Research There Is Nothing

  • User research is part of every project, may it be information architecture project, conversion optimisation project. You have to do the research for your product/service, website, because it always “depends”.
  • How much research should you do? Just enough.
  • Qualitative research: Why? How? In-depth input needed. Interviews. Moderated user testing. Surveys with open questions. Unstructured data. Small.
  • 80% of companies say they are customer-centric. Only 8% of customers agree.
  • Focus group is not a user test.


It’s a real fun event + you’ll learn a ton.

Elite Camp 2019 dates: June 13-15. Mark your calendars now.

The post Top 5 Insights from Every Speaker of Digital Elite Camp 2018 appeared first on CXL.