Let’s face it, AI is a confusing space with lots of buzzwords. AI stands for “artificial intelligence”. At the heart of it, AI uses machine learning to make predictions about future outcomes based on historical data. Machine learning gets smarter as new data becomes available. It’s why Google can re-route you due to traffic or construction, with the objective to getting to a destination most quickly. Machines are inherently dumb, but data combined with predictive logic programming, based on an objective, makes machines smarter like humans.
AI in our life right now
Artificial intelligence is commonly used in our life right now, although we likely give it little thought. When Netflix recommends a show, Google re-routes you, Amazon predicts what you’d like to buy or corrective text helps you write an efficient email – these are all examples of how AI is already personalizing an experience and making your life easier or more convenient.
Why AI in marketing?
We give little thought to maps which re-route us in Google. It just does, and it makes our life easier. The same thing will happen in marketing and sales as AI is put to work. Software will get smarter and make things more personable. It will make predictions about future outcomes based on historical data, and as new data becomes available, machine learning will get smarter and better. While AI may seem abstract to anyone who lives outside the world of coding and programing, we appreciate the outcome. If AI can help make better decisions as marketers, or if AI can help make the customer experience better, it will be welcomed. Consumers are basically already giving up data and privacy in exchange for personalization. AI will give brands the ability to personalize at scale, not humanly possible currently. Consumers will then start to expect a certain level of personalization as competitors use it. We’ve already been conditioned to expect a frictionless experience by Amazon and Netflix, and that level of ease will become the new standard.
SIX Practical applications of AI in marketing
Chat bots splashed on the scene a couple years ago, and have become more widespread, user friendly and engaging. Chat bots enable 24/7 responsiveness, ensure prompt answers, and record the entire buying history. They are used for product recommendations, customer service, sales Q&A, and e-commerce transactions. But contrary to popular belief, according to my research, most chatbots currently in the market are not AI driven, but instead use human powered branching logic to come up with an exhaustive list of responses through an if/then/else pattern. In the future AI will get rid of the need for humans to set up the branching logic, and natural language processing for voice will also be introduced. If you want a fun example of Chat bots used well, check out “Meet Rose – an AI chat bot with personality” – an article I wrote last year.
2. Content strategy
How might a machine help with content strategy you ask? Some companies like Hubspot already use content clustering to recommend topics related to search words for your website. But as AI goes deeper in this area it could recommend topics to write about, keywords to use, how many times to publish, and even timing of posts based on what time of year people search for things. AI powered image enhancement will also become more widespread. Can you imagine getting an alert to write a post today, for this reason, and here are the keywords to insert?
3. AI for social listening
Social media monitoring is very important, and although there are tools out there to make it easier, AI has the ability to take it to the next level. Monitoring competitor content, reviews and what content to respond to would be helpful. There will be lots of opportunity in reputation management and crisis monitoring using AI.
4. Influencer identification & auditing
As influencer marketing grows and matures (see trend #1 post for more on that) AI will be used to identify micro influencers in certain categories, measure real influencers by interaction across platforms, and be able to audit out fake follows or engagement.
5. Automatic ad placements, ad creation, audience selection
This is where AI holds great promise for marketers, but also harbors some dark possibilities when in the hands of those with questionable ambitions. Facebook already generates AI mock up ads and asks if you’d like to run them based on past posts. Machine learning can be used to decide which image or headline will perform better in ads. In the future split A/B testing and tabulating your own results will seem so old school. Some companies are already using AI for this as well as deciding which audiences will be most receptive. Automated audience creation, given objectives is a perfect task for AI, as is campaign tracking. AI content creation, automated ad buying and audience targeting will dramatically change the competitiveness of social media creative, targeting and buying. Once more advertisers use the system this way, there will be no going back, since efficiency and ROI will prevail as ad costs climb. In the end, using AI will be the only way to remain competitive.
6. Using AI to predict trends
AI can be used to process millions of social media posts and find subtle similarities in the text, images, and hashtags used. A subtle similarity could indicate an emerging trend. We’d still need human powered research to interpret what the similarities mean in the context of the industry, but the promise is fascinating.
Here’s a run-down on some specific AI tools for marketing you might want to consider:
AI Tools for marketing
1. CRM platforms
Your existing customer relationship management platform may have features you haven’t activated. That might be a good place to start. Salesforce, Hubspot and Mailchimp CRM are just a few examples.
2. Chat bots
3. Social listening
4. Email platforms
Drip and Mailchimp are two email platforms that provide list segmentation and automated sequencing. In the future we can expect them to use AI to predict what time to send (machine learns when you personally open emails, versus time zone) and personalize time to send 1000s of emails. Automated personalized content will become available – learning what links you click on and customize the content each person gets. Right now email funnels use human powered branching logic but AI is coming to these platforms.
5. Creative performance
Pattern89 helps predict winning creative performance. Their tools can be used to optimize social ads, creative, budget management, and channel management. It lets you know what is working, so you can allocate budget there, or quit using certain creative. The AI makes a prediction based on data, and makes a recommendation to you, then you decide if you want to execute on it. If Facebook and Google are smart, they will offer this type of service within their platform.
Persado leverages comprehensive marketing language knowledge based on over 1 million words and phrases, powered by AI and data science. It unlocks the power of words to engage customers, moment by moment across every marketing channel. It can be used to create content for social ads.
Phrasee is another tool that uses AI powered copywriting to get more opens, clicks and conversions on emails. They claim that their AI optimized language is always in your brands voice. They just need 100,000 emails for the machine to learn, and then it can be trained to write email subject lines that get 57% click uplift and opens.
Focus on the job to get done, not the tool. If AI has been shown to get something done faster, cheaper or better, you could benefit by being an early adopter.
Where to learn more?
The Marketing AI Institute is a great resource. Sign up for their newsletter and check out their blog. They publish 5 articles a week in really accessible language to educate marketers about AI. Paul Roetzer, the founder, is also launching The Marketing AI Podcast in Q1 of 2020. Link here to be notified when it goes live.
And if you want to be the first to see videos when they’re published that support my weekly insights, be sure to subscribe to my YOUTUBE CHANNEL.