There has been an explosion of web design and marketing companies in the last couple of years that use AI tools to generate much of their content. As a freelance designer working with numerous companies, I am often required to work with a marketing person or a third party marketing company who are also working with a specific client. And after a few weeks you get a good idea if they are genuinely knowledgeable about marketing, or not!
Experience still counts.
We’ve all read about the glaringly obvious mistakes that AI systems make, such as images of hands containing 6 fingers, or news alerts with completely false information. What looks like an obvious error to a human is often missed by AI. Even if it’s only a 1% chance, AI cannot be fully trusted at the highest level.
If a marketing company relies 100% on AI tools, without the experience to notice an obvious error, how can they pinpoint an error or prevent an obvious error slip through the net?
AI in paid advertising.
If you haven’t done a lot of PPC advertising, it’s easy to fall into the trap of believing the AI recommendations. Meta and Google will basically create your advert for you using AI, but if you follow all of their AI advice, you will end up spending way more than is necessary. It’s like having Count Dracula in charge of a blood bank… “I would recommend that you donate 24 pints of blood today Sir”
Go back 5 or 10 years, none of the AI recommendations existed, you had to use your own brain to generate an advert. Which is why a marketing person with a few years experience will defeat AI every time! Read more about ignoring Google Ads recommendations.
For example…
When an inexperienced marketing person runs a Facebook Ads campaign, they will usually have to prepare a report to explain the results. What they often say is…
The metrics illustrate that our Facebook advert was well received by the audience, gaining 5000 individual clicks, costing 15p per click, generating a substantial stream of traffic to the website, gaining a large number of potential customers.
But what they should have said was….
I spent £750 of your money to run a Facebook Ads campaign, which received over 5000 clicks, but we actually received zero sales or customer enquiries. So it was basically money thrown down the drain.
Overuse of technical jargon.
Even before the rise of AI, I have worked with various marketing companies, and it’s usually a red flag to me when people overuse technical “marketing speak” to explain a situation, where normal English could have done a better job. The insertion of unnecessary technical phrases is usually designed to convey an air of expertise, but they usually shoot themselves in the foot by basically saying something that makes very little sense.
Highlighting irrelevant statistics.
Another common misconception that you find with inexperienced marketing people is that increasing traffic is the number one goal. And they often proudly show you their reports, illustrating how they have brought 1000s of people to your website. But if this traffic does not turn into actual sales or customer enquiries, it was clearly the incorrect audience.
Similar to the way AI Ads recommendations work – getting traffic is not the main goal. Getting the CORRECT traffic is the hard part! Anyone can throw money into a Google / Facebook Ads campaign and get 100s of clicks on an advert. But if you then get zero enquiries or sales from the advert, it’s a total waste of time and money.
The main flaw in AI content.
Going back to AI in general, it’s my understanding that AI generates content based on information that has been gathered from multiple platforms. Most AI companies have “trained” their systems using existing content. So this suggests that any content created by AI is not truly original, it’s merely a regurgitation of content that has been generated before. But reworded and shuffled around with other information sources, using the correct conjunctive words to present the information as perfect English. It’s copied content disguised as original content.
Truly original content will always win!
Much of my work involves managing large websites, and I am often asked to generate a blog article for a specific subject. Of course I have experimented with AI tools to generate content, such as Chat GTP (Open AI), Google Gemini or DALL.E and it is highly impressive. But I always edit the text content slightly, to ensure it is not simply copied & pasted.
I suspect that truly original content will always rise to the top.
When the dust settles, and Google has filtered through the content on billions of websites, content generated by AI will be quite easy to detect because it is not truly original. And any company that relies on AI content for their blog articles or general page information will eventually find themselves below the companies that invest in original content.
AI tools for web design.
For many years now we have had DIY website building platforms like Wix or Squarespace, so anyone who is a little “tech savvy” can create a nice looking website. And it’s a clear, natural progression that these platforms should use AI to help their customers produce a web design with a healthy amount of relevant content.
And this can be a successful method of creating a decent website, for a small, locally based company. But for a large, national company with the desire to outrank their competitors, a truly bespoke website will give them a better chance.
Of course it depends on your budget whether you invest in a professional web design, or try the DIY route. But if you really want your website to be successful and actually generate sales, I would recommend a WordPress web design. The yearly running costs will be far lower than the standard Wix / Squarespace prices, so after 2 or 3 years you will already be better off.
AI tools for SEO.
There are a number of tools that can scan a website and give you a list of improvements, such as Screpy, SEO Optimer or SEM Rush. But after managing 100s of websites over long periods of time, you start seeing the flaws in these systems. They basically give you the “cookie cutter” advice on how to improve SEO, with lots of valid tips, but some tips that are highly debatable. Even the long-standing SEO advice about META descriptions is simply not correct any more. I’ve put it into practice by testing different websites, using META descriptions with more or less text than is widely accepted, or even not using META descriptions at all. It makes very little difference to your search engine rankings – on both Google and Bing.
The number one factor in creating a high ranking website is to generate original content. Keep adding content to show search engines that the website is currently updated. Keep producing information that ties in to your business or services, so you are giving a search engine various phrases to pick up on, depending on whatever someone is searching for. If you can closely match what someone is searching for, you will appear high in the results.
Unique content is one of the most important factors in search engine rankings READ THE GOOGLE SEO STARTER GUIDE
Conclusion…
Much like the AI recommendations you get for PPC campaigns, I believe that AI content is highly impressive and can be extremely useful. But it’s simply a tool to generate “regurgitated” content. For a small company that is just starting out, with a local audience, using AI tools can produce a decent website and content which can get you off to a good start.
But for a national company with strong competition, AI generated content is not good enough – just yet! Companies that really strive to be above their competition will always benefit from truly original content. And after the tidal wave of AI content is eventually sifted through by the search engine spiders, the cream will rise to the top.