Expert Q&A: Alex Bacon on using AI effectively in B2B marketing

Alex Bacon is the Founder of BrightKeel.AI and an AI and marketing advisor who helps business leaders use AI and automation in targeted ways to support growth. We sat down with him to learn what it takes to use AI effectively and responsibly in B2B marketing.

Radix: What’s the biggest pitfall marketers should be aware of when they’re using AI?

Alex: The key thing to avoid is taking an extreme, all-or-nothing approach to AI.

At one end of the scale, you have marketers who dabble with AI and then abandon it. So for example. They might tell it to write a press release and then get something back in five seconds that looks great on the surface, but on closer inspection is garbage, so they just stop using AI. At the other extreme, you have people who see all the AI headlines and think they can just automate everything and get rid of the marketing team.

The reality is somewhere in the middle. It’s about understanding that the task of creating a piece of content is actually many different steps and you can map AI tools and assistants to the steps where it makes sense to use them. That could be doing the research, coming up with a suggested structure, or assessing whether the content will appeal to your ideal customer profile. But if you think you can just flip a switch and AI will do all of it, you’re probably going to fail.

Radix: So how can marketers be sure they’re making the most effective use of AI?

Alex: The first thing is to have the fundamentals in place before you start. That includes having a clear overarching strategy and value proposition, along with a messaging framework, a well-defined ICP, and persona-specific messaging. Without these fundamentals, AI usage will be directionless, and you’ll have conflicting and confusing messages out in the market. But if you’ve done this groundwork, you can feed, instruct, and use AI with the right context consistently, so you get results that are aligned to your strategy.

You also need a strategy outlining what you’re comfortable letting AI do and where human skills and expertise will still be required. You can’t just jump in and create loads of AI agents to do X, Y, and Z; it needs to be done in a deliberate, controlled way.

Whatever AI strategy you decide on, its success will be based on three pillars: education, culture, and leadership.

AI is like any other tool; people need to understand how to use it to get the most out of it. They need to feel confident in their ability to use the tool to get the right outcome, which requires education. Knowing how to prompt properly, build an assistant, train it, and so on is one element of it. But there’s also knowing your craft; even if I use AI, an experienced writer will produce something far better than I ever could, because that’s their job.

Culture is also very important. People need to feel they have permission to try new things – in a controlled, safe way – so they can find new efficiencies. There needs to be a culture of experimentation, where people accept that things aren’t going to be right first time and might even be less efficient to begin with as they’re learning.

Fostering that culture depends on strong leadership. There needs to be transparent, two-way communication between leaders and teams about fundamentals like AI strategies and responsible usage, as well as more experimental initiatives that might uncover new opportunities.

Radix: What kinds of things should marketers be looking out for to ensure they’re using AI responsibly and safely?

Alex: Historically, marketing teams didn’t create much risk in terms of data governance and compliance, but with AI potentially accessing all sorts of data sources, marketing needs to be part of that data stewardship conversation.

It’s not marketing’s call, of course, but the business needs to agree on what it’s comfortable exposing to public and in-house AI tools. You’re applying AI to amplify your capabilities and accelerate workflows. But if you don’t have the governance structures in place, you can easily amplify and accelerate the wrong things.

Once you’ve got a clear strategy and governance framework, it’s back to that education, culture, and leadership – making sure people understand what the guidelines and guardrails are and that they’re staying within them.

Radix: There’s also the potential risk of damaging the brand by publishing low-quality AI-generated content. How can marketers get value from AI while protecting their brand?

Alex: If you want to avoid putting out AI slop, it comes back to establishing in advance what you’re comfortable using AI for. You need to look at the steps in the content creation process and decide where you want a human to be in the loop. That will vary depending on your market and your risk tolerance.

For example, you might decide you want all your writing to be done by humans and AI will take care of the steps around that, like research, structure, and targeting. Or you might decide to have AI write your drafts and then you’ll review and edit them.

If you’re all-in on AI, that review and editing process is absolutely vital to avoid putting out content that’s too generalised or contains bias, inaccuracies, or fabrications. For example, I came across a blog post recently that had loads of really valuable stats in it, and I thought, “This is gold!” But I also thought, “This sounds too good to be true.” So I backtracked to try to find the sources, and the stats were completely made up – it was all just nonsense that someone had prompted AI to generate.

Problems like these arise when people outsource the thinking to AI, which is not really what it’s best at. That’s when you end up with the slop. Making something engaging and meaningful is really a human-to-human activity, so effective content depends on ensuring humans retain control of the critical thinking behind it.

Radix: Finally, if you could give a marketer just one piece of advice about using AI effectively, what would it be?

Alex: Identify where human input is really valued and work around that. Figure out how you can remove, improve, or automate the other parts of the process. So whether it’s defining strategy, writing content, or talking to customers, determine what you want to ringfence for human input and expertise before using AI to fill in any appropriate gaps.

Get more expert insights in our B2B Content Marketer’s Handbook

If you’re keen to learn more about how to get great content into the world – with or without an artificial helping hand – get your copy of our handbook , featuring contributions from experts across the B2B marketing landscape.