Keeley Gloster
From helping people write emails to generating code, analysing data and answering questions, artificial intelligence is finding its way into almost every corner of business. But beyond the headlines and bold predictions, what does AI actually mean for businesses today?
It’s also a topic we’re discussing more and more with our clients. Businesses are increasingly asking how AI can support them, from customer facing tools to improving internal processes.
At Barques, we’ve been exploring where AI can add value, where it falls short, and what businesses should be thinking about before jumping on the bandwagon.
For years, businesses have focused on getting found through search engines such as Google. Increasingly, however, people are turning to tools like ChatGPT, Claude and Gemini to answer questions directly.
It’s easy to assume this signals the end of traditional search, but the reality is more complicated. Many AI tools still rely on search engines to find and verify current information. If you’ve ever seen “Searching the web…” appear in an AI response, you’ve already seen this happening in practice.
As Barques’ Developer Elliot Pettingale explains:
“Traditional SEO is still just as important because many AI tools use search engines to locate and verify information.”
In many ways, the same principles still apply. Well-structured websites, useful content and strong SEO remain important because they’re often the source material AI systems draw upon.
The difference is that businesses now need to think about visibility in both traditional search results and AI-generated answers.
Barques Developer Ross Whitehouse also highlights another challenge emerging alongside AI search:
“One of the biggest challenges is that we still have very little visibility into how often AI recommends our websites—or why.”
Unlike traditional SEO, where rankings and performance can be measured using established tools, AI search is still evolving. Understanding how websites are surfaced by large language models remains one of the biggest unknowns for marketers.
There are a few new techniques emerging, including adding files that help AI tools understand and navigate your website more easily.
At the moment, these approaches remain fairly niche. Most major AI platforms aren’t relying on them heavily, but they’re inexpensive to implement and could become more relevant as the technology develops.
As Elliot says:
“Adding an llms.txt file (a simple file that gives AI tools a summary of your website) is such a low-effort measure that it’s worth testing, even if the standards are still evolving.”
What’s arguably more useful right now is measurement. Businesses can already track traffic arriving from AI platforms such as ChatGPT, helping build a clearer picture of how AI-driven discovery is influencing website performance.
One of the most visible uses of AI has been software development.
Tasks that once involved hours of searching through forums and documentation can now be completed with a simple prompt. Ask a question in plain English and you’ll often receive working code within seconds.
For experienced developers, that’s often a productivity boost. For less experienced users, it can sometimes create solutions that work initially but become difficult to maintain later.
Ross explains:
“AI lets developers ask questions in their own words instead of hunting through documentation—but you still need the expertise to know whether the answer is actually right.”
AI has made development more accessible, but it hasn’t removed the need for technical judgement. Developers still need to understand trade-offs, evaluate solutions and ensure code is appropriate for a project’s long-term needs.
AI is excellent at producing code quickly. The challenge is that it often prioritises getting to an answer over finding the cleanest answer.
It’s not uncommon for AI-generated code to include multiple layers of redundancy, unnecessary complexity or solutions that technically work but aren’t particularly elegant.
As Elliot explains:
“AI is fantastic for rapid prototyping and tightly controlled tasks, but it’s most effective when it’s assisting experienced developers rather than replacing them.”
He also points out that AI-generated code often solves problems differently from experienced developers.
“AI often gets to an answer by adding layers of redundancy rather than understanding the best solution, which can make code harder to maintain.”
As more businesses adopt AI-assisted development, we may well see growing demand for experienced developers who can review, refine and maintain what AI produces.
When people think about AI, they often picture chatbots.
In reality, some of the most valuable applications are entirely internal.
Most businesses sit on vast amounts of information spread across documents, systems, emails and shared drives. Finding the right answer can take far longer than anyone would like.
Ross believes this is where AI offers some of its greatest potential:
“AI really shines when it helps people find answers hidden across huge amounts of information.”
Imagine asking:
Rather than searching multiple systems, AI can help surface answers quickly by connecting existing knowledge sources together.
For businesses managing complex digital estates, the time savings can be significant.
Just because something can be automated doesn’t necessarily mean it should be.
Customers generally accept automation when they expect it. Improving website search, helping users navigate information or handling straightforward support requests are all examples where AI can improve the experience.
Where businesses need to be more careful is replacing interactions that people expect to be human.
As Elliot puts it:
“Businesses should focus AI on the areas where customers already expect to interact with a system—not replace the moments where people expect a person.”
Most of us have experienced the frustration of fighting with an automated phone system when all we really wanted was to speak to someone.
The best AI implementations tend to be the ones users barely notice. They remove friction, save time and improve experiences without trying to replace human relationships entirely.
Public attitudes towards AI remain divided.
Some people embrace it enthusiastically. Others are concerned about its impact on jobs, creativity, privacy or customer service.
As Elliot explains:
“Customers are far more likely to embrace AI when it improves their experience than when it looks like a cost-cutting exercise.”
Businesses therefore need to think carefully about how AI is introduced and communicated.
The conversation shouldn’t begin with asking how AI can replace people. It should begin with asking how it can create better experiences for customers and support teams alike.
AI is developing at an extraordinary pace, but many of the big questions remain unanswered.
Issues around copyright, regulation, security, data ownership and environmental impact are still evolving. Meanwhile, headlines often move faster than the technology itself. That’s why a measured approach is important.
Ross sums it up well:
“The businesses that get the most from AI will be the ones using it where it genuinely solves problems, not simply because it’s fashionable.”
Rather than viewing AI as a magic solution or an existential threat, businesses should see it for what it currently is: a powerful tool with clear strengths, clear limitations and significant potential when applied thoughtfully.
AI is moving quickly, but successful digital strategies are still built around people. The right technology should solve problems, strengthen your brand and improve the experience for your customers.
We’re increasingly being asked how businesses can successfully implement AI and make the most of it within their business. From AI-powered chatbots to improving how your website is understood by tools such as ChatGPT, we’re helping clients explore practical applications that deliver value.
Whether you’re looking for a proven off-the-shelf chatbot that integrates quickly with your existing systems or a bespoke AI solution designed around your organisation, processes and customers, we can help you find the right approach.
We’re also helping businesses prepare for the future of AI search by optimising websites for large language models (LLMs). This can include making your website easier for AI tools to understand, helping them provide more accurate information about your business, products and services.