I’ll be honest, my AI learning curve is slow. I don’t have the benefit of working with other marketers everyday who can share their own experiences, nor do I have someone in my business dedicated to establishing best practice and informing me of what I should be doing. It’s just me, learning day by day. I read books, I watch webinars, I discuss AI with other people in the industry, and I experiment just as I would with any marketing tactic or platform. And whilst I’ve been open to using generative AI and LLMs for some marketing tasks, I’ve been stubbornly reluctant to use them for others. Specifically, writing marketing strategy, because that’s my job!

One of the key selling points of using AI in marketing is ‘efficiency’. I personally thought that writing the House of Comms strategy in under two hours was pretty efficient, but that efficiency came from years of experience in marketing, as well as in-depth knowledge of my business and target audience. Not everyone who is expected to write a marketing strategy is in such a privileged position and so maybe I have an unfair advantage. Would it be possible for someone with less experience or expertise to achieve the same outcome? Maybe Claude would be a better solution after all. There was only one way to find out.

So, here’s the challenge. Kate versus Claude. Can AI write my marketing strategy faster? Probably. But can AI write it faster AND better? Let’s see!

Sadly, Claude doesn’t need to experience nature to generate strategy quickly. A good, detailed prompt that gives context and sets parameters, should save my legs from another 5km walk. Ten minutes later and I had a “House of comms lead gen marketing strategy”. Points already deducted for not capitalising Comms.

 

Where Claude wins

• Claude had a good understanding of how my target audience buys strategic services, namely that it is “on trust, referral and proof of thinking” which I had not necessarily considered. I know my audience and know that this line is accurate, but I’d potentially overlooked this specificity in my strategy.
• Building on this, Claude laid out a number of “buying trigger patterns” for my target audience, around which I should design content and outreach. This I definitely hadn’t thought about in my own strategy, so that’s a great addition.
• Claude also had a good grasp of the role of marketing in my business, given I’m a solo consultant, and framed it as “narrower and more deliberate than typical B2B lead gen” showing an awareness that my resources are finite and therefore my marketing remit needs to be really focused.
• I particularly liked how Claude created a 6-month phasing plan from foundation to amplification to compounding. Whilst I’m nearly three years into running my business, my proposition has recently been refreshed, so it’s worthwhile pressing the pause button on my marketing and building in time to see what’s working.

 

Where Kate wins

• I naturally have a much better understanding of my target audience and its nuance. I expected Claude to infer from my services that my target audience would be interested in walking as well as requiring strategy. I consider this a key self-qualifying attribute of my prospects and clients and feel it was a missed (and obvious) note to make.
• Whilst Claude’s strategy was comprehensive and in some areas more detailed than my own, the channel strategy was merely a playback of the tactics I already use, which Claude had gleaned from my website and social platforms. Claude failed to suggest any new channels and also failed to note my upcoming event in September, which is specifically designed for that crossover in interests for my target audience.
• On the whole, Claude was highly knowledgeable, but much of the strategy could have been copied for any similar B2B services business. Considering how niche and novel my method of delivery is (i.e. using nature to stimulate create thinking), there was nothing truly unique or creative about Claude’s strategy.

 

And that last point is why I’m still not convinced that AI can replace original, divergent thinking done by people. It isn’t a thinking partner, it’s a machine that relays patterns of data back to us that relate to the prompt. I’m fully aware that no idea today is truly original, we all build on creative executions that have come before us across all media, however, AI doesn’t seem to be able to replicate that same level of ideation despite having access to so much more information than any single person or group of people has. The originality, in my humble opinion, just isn’t there.

 

So, what’s the outcome of Kate vs Claude? I’ll definitely be stealing some of Claude’s notes for my own strategy, but I won’t be replacing my own thinking with AI any time soon. I stand by the strategy I wrote in just two hours (thanks to the power of movement in nature) and will continue to encourage you to embrace human thinking when you tackle your own. As always, if House of Comms can help you start that journey, just ask!