Selling AI Before It Was Cool: What Ten Years in Natural Language Generation Taught Us About Survival

When my company’s co-founder completed his PhD in AI back in 2006, he felt practically unemployable. This was no fault of his – but was simply because companies weren’t hiring people with backgrounds in artificial intelligence at the time. For them it wasn’t strategic; it wasn’t a priority. For most businesses, AI was something from the future – not an operational reality – and was something that many people had never heard of.  

Fast-forward to 2015, and he’d co-founded a business leveraging AI technology. There was a simple pitch behind it: using natural language generation to write better email subject lines. At the time, that was all but unheard of. While investors were intrigued, most weren’t convinced. They were skeptical and dismissive. They’d say that whilst it was no doubt interesting, they just couldn’t see a future in it. As a commercial proposition, it was clear that AI wasn’t on anyone’s radar yet – amongst the tidal waves of VC funding we’re seeing in AI currently, it might seem hard to believe that this was ever the case. But it was.  

As the company grew, the team attended conferences on computational linguistics and language generation. These were overwhelmingly academic affairs. There were, at most, a handful of other companies like us in the space. We were one of very few, and got used to it.  

Then, almost overnight, ChatGPT launched. Suddenly, companies of all sizes were able to generate competent copy in seconds. What we’d spent so long building went from something the public saw as ambitious and futuristic to a given. A known quantity. In response, dozens of startups emerged – AI copywriting companies appearing left, right, and centre. The market flipped, moving from “what’s AI?”, to implementing AI at every stage of the business in a heartbeat. More competitors came out of the woodwork, many founded by people who’d never worked in marketing or natural language processing. It was a gold rush. Most of those companies are long gone. Jacquard, however, is still here.  

There were several reasons we survived while others fell by the wayside. Our user base evolved alongside the market. What once required education now requires differentiation. As standard LLM-powered copywriting flooded the market, it became more and more important to find a way to stand out. We found prospective clients pivoted from asking “what can AI do?”, to “what makes your AI different?”.  

We encountered AI fatigue – something we’d never faced early on. Brand managers were exhausted; they’d spent too much time on introductory calls with companies promising generative AI tools would revolutionise their workflow, only to find out that they were just well-disguised ChatGPT wrappers. The output was generic, off-brand, and failed to capture the nuances and intricacies of the voice these brand managers had spent so long cultivating. The honeymoon period with AI was over.  

We grew when many others struggled because we hadn’t built around the promise of AI. We were built around a problem that needed solving – and AI was simply the most effective tool to solve it. The companies that failed were the ones founded on the premise that AI itself was the innovation. They pitched the technology, not the outcome. And so, when that technology became ubiquitous, they had nothing to differentiate themselves.  

The problem we were solving as a company existed before ChatGPT, and it still exists now. Our AI has evolved dramatically over the years – it’s had to in order to keep up – but the core value proposition hasn’t changed. 

The LLM gold rush kept us grounded. We learnt to focus on proving ROI when investors were skeptical. When potential clients needed an education, we built resources and case studies. When the market exploded, we deepened our specialisation in brand language and enterprise messaging. Now with AI fatigue settling in, we can point to a decade of real results – and not just gesture to empty promises.  

The LLM gold rush was definitely a scary moment. Things are different now; the pace of change has accelerated. What took years in 2015 now takes months – but our focus on customer value and practical outcomes hasn’t changed. If anything, it’s only grown more important as the noise around AI grows louder. Technology is a tool, not a strategy. Companies that survive solve real problems; they don’t just jump onto the latest tech hype cycle. Today, that’s LLMs and agents, but tomorrow it could be something else. The problems your clients need solving, however, will remain.  

It’s ironic that our co-founder felt unemployable with his AI expertise in 2006. Today, that’s almost impossible to believe. But the gold rush taught us to focus on what matters: not the technology itself, but what you build with it.