AI in marketing: How to balance automation and the human touch

Marketing technology is progressing at such an accelerated pace that it’s hard to keep up and easy to let the tech do the driving. We now have machine learning and AI driving much of our marketing platforms. 

Whether it’s letting Google or Facebook make our ad decisions for us or asking ChatGPT to write all our content (a very bad idea), AI is making our job easier. Or is it?

It depends. Are you getting the results you want? Are you safe using AI-generated content? Are you in danger of losing your voice and jeopardizing your client engagement online? Are you being honest?

In this case, honesty means being human, genuine and transparent. And listening to the most important computer in marketing today — the one between your ears. 

The machinery of modern marketing

In the early days of the web, we were mostly experimenting and trying to figure things out. Even today, with all these tools, many marketers are still not getting the clarity and visibility they need. 

We have a saying here at Wellspring Digital, “Rule number one, don’t trust the tools.” Distrust but verify. For example:

Marketing automation

Platforms like HubSpot, Constant Contact, Marketo and others have opened all sorts of possibilities for marketers. We can now automate most of our marketing efforts.

We can automate our email campaigns and set up automated drip campaigns using workflows that send messages based on a user’s activity, amount of time elapsed or other factors. We can use AI-driven chatbots to streamline customer service and online sales. 

But this isn’t a “set it and forget it” situation. We still need to use our brains. Look at the user’s journey as they interact with your campaigns using a certain amount of empathy.  Use this insight to assess whether these campaigns will produce as you expect them to — not just as the machines expect them to. 

In other words, an automated drip campaign is only effective if:

• You’ve done your homework and truly understand who your audience is, what their needs/wants are and how you can get them to act. 
• You’ve developed content that encourages them to convert. It is on brand using your unique voice, keeping a human feel to your messaging. Your audience will be able to tell if AI cranked out your content without any of the above considerations and your campaign will suffer. 
• The path you want them to take is clear and you’ve motivated them sufficiently to want to take it. 

Marketing automation platforms have revolutionized marketing. But like the computer nerds always say, “garbage in, garbage out.” These platforms only work if you take the time to plan and execute based on your complete understanding of the convergence of your marketing goals and your user’s needs. 

Dig deeper: AI marketing automation: How it works and why marketers should care

PPC: Paid search and social

AI is also used in automated bidding for paid campaigns. Or auto-targeting and AI ad generation. It’s not wrong to use these tools. That’s why they’re there. But, again, you don’t simply hand over the keys to Google, Meta or whoever. 

You must really know your audience because the machines don’t. Take the time to verify the words/phrases/themes that you should bid on.  

Take a more holistic approach. Develop your buyer personas. Engage in SEO so you know which terms you can rank for versus which terms you’ll need to bid on and which terms are a waste of time. 

Understand how your audience uses social media. Think through the ads, get a complete picture of what will get their attention and encourage them to act. 

Developing realistic user journeys will help you verify that AI is making the best decisions for your ad campaigns. If you set it and forget it, there is a good chance you might blow through your budget with nothing to show. 

There are certainly benefits to martech

Don’t get me wrong. There are definite benefits to bowing down before our robot marketing overlords. This modern marketing machinery can, if used properly, make us better marketers.

It can also be used to serve our clients better. And that’s my point. 

Efficiency and scalability

In today’s ocean of marketing channels and advertising opportunities, most teams struggle to keep up or know where to begin. AI is a great idea generator. If you are blanking on a blog topic, use a tool like ChatGPT as an idea generator. Or use it to develop an outline for your post. You still must do the work and develop real, meaningful content. 

AI and marketing automation platforms can let a team of one or two do the marketing that used to take a team of ten. Automated messaging, scheduling/monitoring of social media and automation in advertising can improve the efficiency of campaigns while allowing the team to scale, expand into more channels and try different messaging and ads. 

Data-driven decision-making

And all these tools are grabbing tons of useful data as they do your work for you. Data is dumb, and on its own, it rarely tells a complete story. You need to compare the data from these tools with the data in your head and the heads of your team members. 

Data-driven decision-making is only as good as the data you use. If you look at that data and make a bad call, you will have a bad outcome. Be thoughtful about what the data is telling you and strategic in how you act on it. Engage your humanity to fill in the blank spots. 

Personalization and targeting

AI and marketing technology allow marketers to target audiences with much greater specificity. And this tech is making personalization much easier than before. 

Using AI, you could analyze your subscriber list, for example and identify specific buying habits or needs. You can then segment your list and deliver more targeted messaging or offers to a smaller, more specific segment of your audience. 

The point is not to be creepy. The point is to deliver what each person wants more specifically. They will feel “seen” and be more likely to act. They’re starting to expect it. 

Dig deeper: 3 steps to make AI work for you

Your AI honesty checklist

You can automate to the point of being annoying if you’re not careful. Your audience will tune you out or mark you as spam.

And AI content can seem cold or formulaic, which will hamper your customer engagement. People don’t want to feel as though they are dealing with a machine. 

People want to work with other people and feel seen. There is a real risk of alienating your customers with each incremental loss of authenticity.

“Openhearted” marketing means engaging your audience as a human, with empathy. Here is a quick checklist to reference as you utilize AI and machine learning to develop and automate your marketing efforts.

Openhearted marketing checklist

• What audience is this for?
  • Have you checked this effort against your buyer personas?
  • Is this something that will resonate with them positively?
  • Will they act based on your understanding of them through your previous persona work and customer interactions?
• Is it you? 
  • Is it on brand? Does it feel like it’s coming from your brand?
  • Is it authentic? Does it feel human?
• Is it honest?
  • Did you or someone on your team write it?
  • If not, is it written or created in a way that feels like it came from your team?
  • Is it your content and not borrowed from someone without permission? (AI content as is can often be pulled from another source, which creates plagiarism issues.)
• Will they feel seen?
  • How empathetic is it? 
  • Will they read or see this content or ad and feel like it was written for them with their needs in mind?
• Will it motivate them to act?
  • Honestly, get in their heads. Will they care? Will this move the needle with them?
  • Think through their experience and ask yourself, “Will this solve their problem, scratch their itch?”

This should be a quick spot check based on feeling, experience and audience knowledge. In time, it will become shorthand for all your marketing reviews. 

Embrace authenticity, practice empathy and prioritize genuine human connection. These tools are meant to support, not replace, these core values.