Artificial intelligence (AI) is everywhere. We don’t mean in the scary, apocalyptic ‘rise of the robots’ way that is often depicted in films and books. No, we’re talking about the dozens of useful, seamless interactions we have with AI every week as we go about our daily lives.
Recommendations on Amazon, Netflix and Google? AI. Personalized feeds on Twitter and Facebook? AI again. Talking to Siri or Alexa? You’ve guessed it. It’s all driven by conversational AI. Everything from facial recognition and plagiarism-spotting software to spam filters and smart gadgets in the home. So much is underpinned by artificial intelligence.
In marketing, it’s no different. AI is everywhere.
Let’s explore some awesome examples of where artificial intelligence and machine learning are a driving force behind a brand’s marketing strategy. Some of the bleeding-edge and breakthrough uses of AI for marketing purposes you may not yet be aware of. And what marketers are able to do with AI today that they couldn’t before.
And let’s also start at the beginning: What is the current state of play of AI for marketers? And where are we heading?
How many marketers use AI?
Many marketers will be using AI indirectly. Almost all of them, in fact, when you take into account the algorithms built into Facebook, HubSpot and other everyday tools. But how many have a comprehensive, fully defined AI strategy? According to Salesforce, around 60% in 2021 – up from 57% the year before.
Notably, this figure climbs even further to 64% when focusing on high-performing organizations. Meanwhile, just 45% of underperforming brands have a clear strategy. This suggests that a proactive approach to artificial intelligence is helping organizations to perform better.
Note that it’s not a silver bullet. Not all high-performing teams use AI, and using AI doesn’t make a high-performing team. However, there is a correlation. Those who use AI for marketing are more likely to see higher levels of performance. It likely comes down to how well it’s deployed, used creatively, and bought-into within the team and business.
How do marketing teams use AI?
How are brands using AI at the moment, we hear you ask. In truth, there are countless ways – and we’ll come on to some of the more creative approaches next. But, as per Salesforce, here are the top five use cases:
- To drive next best actions.
- To personalize the overall customer journey.
- To automate customer interactions.
- To personalize individual channel experiences.
- For process automation.
We’re not surprised to see such a strong showing from marketing personalization and automation. These are great benefits from AI, helping marketers deliver what they couldn’t do previously. And, most importantly, both lead to a better user experience.
But, as it’s often said, the proof is in the pudding. It’s one thing to talk about marketing AI, and another to look at it in action. So let’s look at some fascinating real-life examples of AI driving better personalization, automation and other marketing efforts.
Five amazing use case of AI in marketing
1. Powerful personalization
People expect personalized experiences, and those expectations are growing year on year. Last year, 45% of consumers in Twilio’s State of Personalization Report said that a brand would lose their loyalty if it delivered an un-personalized – i.e. generic – experience. By 2022, this figure had rocketed to 62%.
Businesses benefit from more than just stopping customer churn too; 80% agree that personalized experiences lead to consumers spending more money – an average of 34% more. Brands that fail to provide tailored experiences are clearly leaving a lot of money on the table.
Enough of the stats though (for now); we promised examples, and Starbucks has been a trailblazer in this area. Through its stores, loyalty cards and app, it’s been gathering over 100 million transactions a week – and that information has been put to good use by Starbucks’ dedicated team of data scientists.
For example, the app – which has 17 million users – continually collects data on what their customers drink, where they drink it and when. This is done through the company’s Digital Flywheel Program, a cloud-based AI engine that can provide accurate recommendations based on previous purchases.
Visiting a different Starbucks than usual? Doesn’t matter. The store’s point-of-sale system will recognize you’re a member from your phone and let the barista know what you typically order. AI has also helped Starbucks use climate data (combined with customer preferences) to provide weather-related suggestions.
Freezing cold outside? It might offer you a hot chocolate. Sweltering hot? An iced tea. It’s personalization, in liquid form.
2. Fast automation with chatbots
Make-up and skincare retailer Sephora knows how to get the most out of its AI-powered chatbots. Since 2017, the brand has been teaming up with social media giants like Facebook and Kik to automate its customer interactions online.
Sephora’s Ora chatbot is a virtual beauty coach, which is available 24/7 to answer frequently asked questions, help customers with their orders and offer personalized content, such as tutorial videos and beauty tips. The upshot? Within just a few weeks, this chatbot alone was successfully handling 20% of all client requests, according to Viseo, the tech company that created the chatbot.
Where does AI come in? The chatbot uses AI building blocks and plugs into Sephora’s back-end information system, helping it to create interactions and conversations that are more personalized, natural and situation-specific.
Fast-forward to 2022, and Sephora isn’t the only brand that uses conversational AI and chatbots. And for those that do, the results are encouraging. A recent Drift survey found that 82% of companies that use intelligent chatbots find them to be a very valuable asset to their market strategy.
And while there have sometimes been mixed results as to whether consumers prefer interacting with a chatbot or a real human, people do appear to be warming up to chatbots. The Drift research states that use of AI chatbots jumped 45% year on year among consumers between 2020 and 2021.
That’s not to say chatbots get everything right of course – it’s important for brands to know how to use them properly and when other options might be better.
3. Showing AI personality with digital humans
Speaking of other options brings us to digital humans. Chatbots are great for specific tasks, such as giving customers information on products, directing them around a website or soliciting feedback. But they have their limits, even when they’re powered by AI.
It’s unlikely you’ll be able to build a chatbot that portrays warmth or empathy, for example. Most brands don’t design their chatbots with anything more than a rudimentary ‘personality’, and nearly a quarter of people prefer not to interact with them because they lack friendliness and authenticity.
This is where digital humans come in. A lot of personality is conveyed through facial expressions, tone of voice, eye contact and a sense of humor. Digital humans incorporate all of these, powered by AI and making them great brand ambassadors to build rapport with customers.
Last year, we worked in partnership with the Hebrew University of Jerusalem to build a virtual version of Albert Einstein on our digital human platform. Digital Einstein embodies everything that made the real man so captivating – it doesn’t just look and sound like him, we worked hard to recreate many of his unique quirks and mannerisms.
Via any browser, people can interact with the great physicist’s digital clone by either speaking to him ‘face to face’ or using text inputs. This version of Einstein is able to answer questions about his life and even host a daily quiz.
The results weren’t too shabby from a marketing perspective either. We saw a massive 350% increase in total web traffic after Digital Einstein launched, with a 140% jump in marketing-qualified leads. We packaged more of our findings on what happens when you give AI a personality in the guide below, for those interested.
4. Getting creative with content
Would it surprise you to know that this entire article was written by an AI? Well, actually, it wasn’t. But it soon could be! (Much to the existential dismay of this particular writer…).
The truth is that artificial intelligence is helping firms to produce copy faster than ever before, and the quality is coming on leaps and bounds. How does it work? Algorithms underpinned by machine learning scour millions of websites to understand patterns in language, including commonly used words, syntax and grammar.
You can then set some parameters, provide a topic and generate content from scratch. And get this – some firms are claiming AI can add more humanity to their copywriting. More humanity than a human, apparently.
JP MorganChase partnered with technology-driven marketing company Persado to rewrite headlines and copy for the bank’s audiences using AI. The computer-written copy saw clickthrough rates on ads jump as much as 450% during a pilot exercise, whereas content produced by regular agencies topped out at 200% increases.
“It rewrote copy and headlines that a marketer, using subjective judgment and their experience, likely wouldn’t have. And they worked,” said Kristin Lemkau, CMO of JPMorgan Chase.
And it’s not just copywriting where AI is shaking up content creativity. Recently, Heinz launched a fascinating new campaign where it asked DALL-E 2 – an AI system that creates realistic images from text inputs – to draw “ketchup”.
The program came up with a lot of quirky images, including, for some unfathomable reason, a ketchup bottle floating in a swimming pool. But many of the images had one thing in common: they looked A LOT like Heinz’s tomato ketchup, all the way down to the historic glass bottle and distinctive logo.
Understandably, Heinz has claimed this as a big win for its branding.
“We’re thrilled to see even the most unbiased source recognizes that when it comes to ketchup, it has to be Heinz,” said Jacqueline Chao, the company’s senior brand manager of brand communications.
5. Charming people with AI voice tech
Continuing the content theme, many marketers would argue that video content is king (especially on social media).
A 2022 Wyzowl survey revealed that 88% of consumers had admitted they’d been convinced to buy a product or service after watching a brand’s video. When asked how they’d like to learn about a new product or service, 73% chose a short video explainer. The next highest result was 11% for text-based articles (#sadface).
Creating quality video content isn’t easy, though. Even recording voiceovers can be a long and arduous process. It would usually go something like this:
- Find talent
- Audition talent
- Choose talent
- Record voiceover
- Cross your fingers and hope there are no issues (but, in all likelihood, do some retakes)
We’re sure you could add your own steps in this already overly long list. Well, save your breath: it turns out AI voice technology can take a lot of the pain out of the process. In fact, the final five steps can be done in a matter of hours or even minutes using the latest voiceover tech.
The results are beginning to sound more human than ever, too. Breakthroughs in deep learning mean that AI voiceovers aren’t just saying the words, they’re adding pauses and breaths in all the right places. They can even convey emotion and empathy.
What’s more, companies like Veritone can (ethically) clone celebrity voices for that extra wow factor. Which means your brand ambassadors don’t need to spend long, costly hours in a studio learning, reciting and redoing lines.
We’re particularly close to the advantages of AI voice tech. We need our digital humans to be able to say anything, and it’s unrealistic (to say the least) to require celebrities, ambassadors or VO artists to record every possible utterance they might need to say. Instead, a couple of hours with a microphone and some clever AI trickery is all it takes in 2022.
As with Digital Einstein, it’s also possible to bring some of history’s most famous people back to life by working with their estate or IP owner to digitally recreate their voice. Who wouldn’t love to hear Elvis’s dulcet tones advertising Brylcreem? Or the late Sean Connery promoting James Bond’s favorite tipple, Martini?
AI in marketing: the road to ROI
In McKinsey’s latest State of AI report, ‘marketing and sales’ is one of the top three business areas where AI adoption is highest (along with ‘service operations’ and ‘product or service development’).
If you’re not using it, you’re at risk of falling behind. Not only this, but the biggest increase in the use of AI has been in marketing budgets. So while you don’t have to spend big on AI, companies are seeing it as a good time to spend with ROI in mind.
AI in marketing is no longer a new, bleeding-edge innovation. Nor does it require a PhD in computer science to deploy. The process is democratized through user-friendly software and platforms, and the impact is often easily measurable.
We’re sure Starbucks knows the increase in average customer spend based on its personalized messages. Heinz will be surely salivating over the click-through rate of its AI-generated content. In our digital humans, too, we’ve seen online conversion rates jump by almost 100%, while other metrics like cart abandonment rate fall significantly.
Some successes are harder to quantify, like how clever use of AI might help you build your brand and show your position as leaders in innovation.
However you determine success and ROI, marketing tools like AI are only just getting started. Brands continue to find ways of making their marketing more personalized, better targeted and more human – even if, somewhat ironically, it’s being shaped by computers.