AI in B2B Marketing: What It’s Good At, Where It Falls Short, and How to Stay Original

Maelien:
Welcome back, I’m Maelien Halton.

Louis:
And I’m Louis Halton Davies. 

Today we’re talking about how AI is reshaping the B2B marketing landscape. AI is everywhere and to be honest, I’m surprised it’s taken us til episode 15 to dedicate an episode to it.

Part 1 – Content Shock revisited

Maelien
Now I’m sure you’ve noticed the fact that since the rise of AI, there’s been a massive increase in the amount of content being published. 

And I’m also sure that by now you can smell AI generated content and comments from a mile away. The long em dashes and emoji lists are a dead giveaway.

In 2014, so, over 10 years ago, Mark Schaefer coined the term ‘Content shock’ in a blog post that rippled through the marketing world.

Back then, he noticed that the amount of content being published was growing faster than the demand to consume it – and he warned that it would get harder and harder to cut through the noise.

And that was before the impact of AI. 

Today, anyone – from a junior marketer to a business owner with no marketing background – can churn out content with no barrier to entry.

It kinda feels like before AI became the norm, we were in a golden age of content – where the standard was really high. 

But with AI, we haven’t seen more high quality content that cuts through the noise. 

For the most part, we’ve just got even more noise. More content with much less engagement and a rise in empty soul-less comments.

Part 2 – Where AI really sits today

Louis
What’s really interesting is that everyone, myself included, predicted that AI would disrupt lines of work that were more logical first. 

Where there were very clear rules of play, procedures and parameters – so things like admin and accounting. 

And while we have seen that happen, the thing that nobody expected was how quickly it disrupted the creative industries too. With copywriting, image generation and even video and animation.

But I’m not too sure it’s going to take our jobs just yet. From a marketing perspective, the comparison we make is that with the support of AI you get the equivalent of a team of a very competent junior level. 

Now that’s not to say it’s not going to evolve. But at the moment AI struggles to consider the bigger picture, to factor in nuances, and to zoom out and contextualise.

So looking at the market. the commercial season. the business model,  the customer – and a whole load of other things.

Sure, you can get a very powerful assistant. But it’s definitely not at the level yet where it’s going to be replacing a marketing manager, a consultant or a domain specialist.

But there are definitely questions around how long that will take – how to nurture entry level talent and also how the middle of the career ladder and above will evolve over time.

Maelien:

A point that I think gets missed a lot, is that the human in the loop is absolutely critical. 

If you expect AI to take entire jobs off your plate from start to finish, right now, there’s a good chance you’ll end up being disappointed.

I also think we need to be really careful not to outsource our brain to AI.
We need to be outsourcing our time instead. 

Using it to speed up the parts of tasks where we provide little value. 

So the formatting, the cleansing, the structuring, the summarising. 

And then doubling down on the high impact 20% where human input is still way more powerful.

Part 3 – The race to the mean

Louis
In Episode 13 I talked about when I went to the HeroConf event in Brighton earlier this year, one of the speakers used the phrase ‘AI is a race to the mean’. Which made so much sense. 

If you think about, it takes it’s knowledge base – which is the internet – and forms it’s output based on an average of what’s already out there. 

Now obviously it’s a LOT more advanced than that and it’s coloured by what your LLM knows about you and how you use it – but the output of AI by default is an average. 

So if you’re asking it to “write a blog on B2B marketing strategy,” what you’ll get is a polished version of what’s already been said a hundred times before. 

And in a market that’s getting more and more flooded with average content – the bump in awareness from just simply putting content out there is a lot less than it was before.

Now obviously on the flip side – if we’re not using AI we’re falling behind. So we have to use it, but we have to do it right. 

We have to train it on what works and for the moment, we also have to take over at a certain point and take it from average to unique.

This means layering in things that AI can’t: lived experience, our own hot takes, your own data, even a unique tone of voice.

And this has a huge implication for performance marketing. If your ads, your content, and your landing pages all sound like and say the same thing as everyone else – they won’t stick.

I’ve thought for a while that as technology evolves and AI becomes more commonplace, it’s actually the foundational marketing skills that get more and more important.

Speaking to the right people with the right message at the right time. 

Even before AI, I had a real problem with the fact that the vast majority of marketing was written almost as if the business was its own ideal customer. And now that’s getting amplified.

So there’s never been a better time to really get to know your audience, what they care about, what they’re struggling with, what they’re trying to achieve – and speaking to that. I’d say that doing that, and doing it originally and as a human is a real differentiator right now.

Part 4 – Practical ways we use AI

Maelien
I guess it would be good to talk a little bit about how we use AI at Web Marketer.

AI is great at carrying out basic ad account audits and we use it to analyse Google Ads Editor exports 

Everything has to be double checked and there’s always more to add – but this gets us off to a great start.

It’s also great for Advanced Google Sheets formulas – I remember when Louis blocked out his calendar for two weeks to create our reporting back end and now we’re regularly doing what would take days in minutes.

It’s made our Google Tag Manager setups better

while it does like to lead you down a rabbit hole sometimes, it’s taking us a lot less time to get to really robust solutions.

It’s also saved us hours as a troubleshooting assistant.

It helps us with review mining – so summarising customer reviews into themes to capture the “voice of the customer” and what they care about. 

Also mining competitors’ negative reviews to target things they do badly that our clients do well.

We also use it to summarise forums and Reddit threads – 

This helps us find challenges and pain points that prospects are openly discussing. 

So we end up with a basic read on the market and the perception of what’s being done well and done poorly.

And we’ve just started using it to build internal tools too – 

we’ve just finished building a new pricing calculator in Replit. 

It looks so much better and presents our costs way more professionally than it did before in Google Sheets.

So hopefully you can see we’re not asking AI to create strategies, or to tell us where to put budget, or to write ads that we’d run unedited. 

Those are all jobs that need input from an expert. 

AI accelerates the data gathering and handles the low value work – taking it to 70% done – 

ready for us to jump in and add the finishing touches you need a specialist for.

Part 5 – Integration and the B2B data challenge

Louis
I also feel like AI is moving much faster in ecommerce than it is in B2B.

I mean, Product Studio was very recently added to Google Merchant Centre for AI generated product photography and improving images.

And for me, this goes back to what we talked about in one of our earliest episodes. Where ecommerce is much more of a closed loop system.

Products, sales and customer support all live in one system that also powers the website. It also doesn’t need much manual intervention because it’s all digital and led by the customer – so the data is pretty clean by default. All of this together makes it ripe for AI to disrupt.

But B2B is different. It’s a lot more fragmented. Tech stacks are messy, bespoke, they change from business to business and, if we’re being honest, often poorly maintained. 

As a result it’s a lot more difficult for AI to cover as much ground in B2B and so it’s grown in a much less all-encompassing way.

Takeaway & Outro

Maelien:

As always, let’s end with a cheeky takeaway:

• Content shock was talked about in 2014 and AI has taken this to a whole new level.
  
  We needed to work hard to cut through the noise before, but now that noise is much louder.
• AI today is like a very competent junior. Brilliant at speeding up simpler work, but it still needs an expert human brain at the end.
  
• By default AI is “average.” If you want to stand out, you’ve got to add the human back in. So lived experience, tone, and perspective – things that it can’t replicate.
  
• Outsource your time, don’t outsource your brain.
  
  Use AI to accelerate things like quick insights, mining reviews, or building simple tools.
  
  The stuff that you can do, but you don’t really add much value when you do it.
  
• And finally, Ecommerce is further ahead because it’s a clean, closed-loop system.
  
  Whereas B2B is much more fragmented.
  
  So, It needs more careful integration, because it’s use cases are more siloed.

Louis:
I guess it’s encouraging that even though AI is touching all aspects of B2B marketing and operations – it’s never been more important to be human.

If you found this useful, please share it with a colleague and leave us a review. And if you’ve got a challenge you’d like us to tackle, head to webmarketeruk.com/topic and drop us a line.

Thanks for listening.