We are often asked: Should you use AI in your Fire & Security Marketing?
And the simple answer is yes. BUT ALWAYS with a human editor.
Having been using AI since late 2022, we know that just letting a bot loose on your marketing on its own is going to lead to disappointing, unsuccessful and potentially damaging results!
So here’s a quick run through of the highlights and low points of our experience with using AI for Fire & Security Marketing so far… with all the cussing, footstamping and general frustrations taken away so you don’t have to deal with them.
It was a Eureka moment!
One morning, Jon had a brainwave about a client’s message whilst in the bath. And I was trying to create a LinkedIn status update about it.
I was struggling. I knew there was something interesting here but I couldn’t find the right angle.
So I asked the AI to help.
It was the bot that came up with the Eureka moment.
Of course, I knew what Archimedes had shouted in the bath… I had just forgotten.
And that is one of the things that the AI is great at.
It reminds you of things you know but which have just slipped your memory.
It can inspire you to create better copy.
Sadly, there is a flip side of the coin… the AI likes to add in generic sales tropes alongside a few emojis that we have not seen before.
Things that don’t fit with my personality or Lollipop’s ethos…
Published without a human edit, anyone who knows me would be able to tell that AI had written the update… not me.
But does that matter? Do people care whether you’re using the AI?
Authenticity and de-skilling…
All the LinkedIn gurus will tell you that you need to be ‘authentic’ if you want to achieve success.
Which brings me on to LinkedIn Comments and AI options which suggest replies and comments.
Underneath many posts, you can see LinkedIn’s two or three word suggestions for how you should reply. And then you can see several comments that have all chosen the same reply.
And I’m not a fan of the third party tools that you can use to create replies and comments to a particular post. I don’t think there are many people who can’t recognise an AI comment when they see one.
It will look nothing like any of the other comments on the thread.
And, if you know the person concerned, it probably won’t sound like them either.
Does that matter? Well yes it does. LinkedIn (and selling on LinkedIn) is all about trust and authenticity. My customers buy from me because they know they can trust me.
Which brings me onto a conversation I had recently with another marketer who was staying away from AI because they didn’t want to be ‘deskilled’
De-skilling: Is AI supplanting humans?
Deskilling is the process where skilled labor is replaced by technologies that require semi- or unskilled workers, resulting in cost savings.
We are starting to see reports of potential and actual job cuts across a range of industries. Companies are embracing the possibilities of using the AI.
But is replacing a lot of your humans a good idea?
Business coach Perry Marshall told me:
“AI is good at answering questions – it’s not good at asking them”
The AI regurgitates based on what it has been fed.
So think of the comprehension tests we used to have to do at school. Take these three texts and produce your own version.
Well AI can do that based on loads of texts… but it can’t necessarily write it in the best way because it doesn’t do emotion.
Remember the old adage… garbage in, garbage out… you cannot always be sure the AI has chosen the best texts to work from. So the more AI there is, the more human we need to be.
What the AI does do well is to help employees to be more efficient and effective. Some jobs that used to take 3 days can now be done in two hours because the AI can take the grunt work out of it.
And help you to make really complex decisions by giving you better information.
Perry Marshall also told me:
The best way to think about it is: “That job you think that AI is stealing from you? That’s the job you didn’t want to do anyway.”
And yet…
Lies, damned lies, assumptions and hallucinations
ChatGPT, Claude, Grok, Gemini, Co-Pilot, Perplexity – the generative AIs are proliferating like tribbles on the Starship Enterprise! Everyone seems to be developing their own AI.
In theory, they’re like a super-competent personal assistant on Warp Speed 9. You ask them to help you with a task and they spit your answer right back at you.
But…
When we were planning our trip to the Security Event in April, I wanted to find out which of the BSIA members also had a stand at the event.
I had the URL of the website page for the BSIA Members Directory. And I had the URL of the website page for the list of exhibitors at the Security Event.
I could have asked my assistant to compare the two lists manually but I hoped that this was a ‘grunt task’ that the AI would be able to do more quickly than a human.
I was wrong. After an hour of patient questioning and prompting to try to improve the answers that I was getting which didn’t seem right, Claude finally admitted the truth.
It told me it had been making assumptions because it couldn’t actually read URLs on the internet.
Basically, it had been telling me fibs and making out that they were facts.
The experts have given this behaviour a name. They say the AI is ‘hallucinating’
Although Claude apologised profusely for making assumptions, it had wasted an hour of my time… sigh…
I decided to ask a different AI that I believe does have the ability to read URLs on the internet.
Sadly, ChatGPT was not very helpful. It told me to go and do it manually myself!
Apparently it was going through a ‘teenager’ phase where it wasn’t being as useful as it could have been… because it had so many demands on its time…
After two wasted hours, I asked my human assistant to manually produce a list of the companies that were both members of the BSIA AND exhibiting at The Security Event.
It took 30 minutes. I had wasted two hours trying to get the AI to do this task.
More facts, figures and fabrications
And let’s not forget the time I asked the AI about paying VAT in European countries. The answers it spat back as gospel truth caused some consternation when I showed them to my accountant.
This experienced professional had never heard of any of the facts and figures that Claude had assured me were correct.
What did I learn? That you actually have to ask the AI if it has the ability to do a specific task before you ask it to carry out that task. And that you always need to cross reference the information you have been given to fact check it.
So is it time to sack your Marketing Manager?
The consensus seems to be that marketing agencies, content creation, call centres, subject matter experts and low wage admin work are all in the firing line to lose their jobs to AI…
Obviously, as marketers, we are very focused on that.
You have tools that seem to be able to create content and tools that can create images. Surely it’s time to fire your marketing manager and just give the work to AI?
Not so fast my friend… you would quickly come to regret that decision.
You can’t just use it to replace a human by saying “write me an article about x” and think you can just publish it straightaway.
Recently, on the Steven Bartlett podcast, Diary of a CEO, he asked comedian, Jimmy Carr, whether he was worried about the impact of AI?
Jimmy replied:
It’s Artificial Intelligence not Artificial Consciousness. If you tell it to write a joke, it can spit back stuff you’ve already written and reorder it slightly. It cannot create original jokes. It’s a covers band: that is a genre and I can do something similar. But there is something about human creativity that I don’t think it’s getting close to.”
AI has all the technical skills to write content but it’s not reliable… or original.
And the graphics…
At first glance, they look really good. But on closer inspection, not so much
The AI is famous for missing fingers or adding fingers
And huge CCTV cameras fastened to the sky watching over the roofs of buildings.
It’s not quite there yet!
Images or words… You have to check and double check anything that AI produces for you from scratch.
But it can be really helpful at editing and improving copy that a human has written.
AI Assistants that work alongside humans…?
Marketing guru Dennis Yu says:
“The biggest mistake in using LLMs or any form of AI is to generate content instead of advising you and helping you edit the content. The AI is more like a smart team member than some magic software that creates stuff for you while you sit back and do nothing”
When you use it to help you… to augment what you as a human can write, it can be really effective.
What does Google think about using AI for Fire & Security Marketing?
Last year when AI first started to come onto the scene, some ‘marketers’ were rubbing their hands together with glee.
They created websites with thousands of pages. In a matter of just a few hours.
These poor quality websites were stuffed with ads that made the websites’ owners money if they were clicked.
Because they were big websites optimised for SEO, they started ranking at #1 on Google for buying keywords.
These ‘marketers’ were raking it in from the ad money and affiliate commissions.
Of course. it didn’t take too long for Google to smell a rat.
So Google developed the ‘helpful content’ algorithm update to fight back.
Their algorithm started to de-index pages that didn’t fit their definition of ‘helpful’.
Pages that are not indexed do not rank. So people don’t visit them and they don’t click on the ads.
Those ‘marketers’ have been crying all over Twitter ever since.
Google issued some guidance on the subject. https://developers.google.com/search/blog/2023/02/google-search-and-ai-content
Using automation – including AI – to generate content with the primary purpose of manipulating ranking in search results is a violation of our spam policies.
Appropriate use of AI or automation is not against our guidelines. This means that it is not used to generate content primarily to manipulate the search rankings, which is against our spam policies
Google wants E E A T
What does Google want? She wants to EEAT… Experience, Expertise, Authoritativeness and Trustworthiness
This is what Google is hungry for. And where the human can beat out the AI every time. We are the ones who have the stories and the real life experiences to illustrate what we are trying to say in our blog posts and social media updates.
Should You Use AI in Your Fire & Security Marketing?
YES BUT use it to Augment your human side…
Marketing guru, Neil Patel, has some great data about the average traffic per post for an AI Generated article vs a Human generated article
The human-created content wipes the floor with the AI generated content in terms of traffic.
And visitors spend longer on human content than AI-generated…
But use AI content that has been tweaked by a human and you get the best results.
I’ll leave it to Neil to conclude this post with this:
“AI is becoming essential in marketing – from creating content to understanding customers and making ads.
But it’s the human touch that keeps things real and engaging.
To stay ahead, CMOs must adopt and adapt to AI, mixing its power with human insight.
Embrace AI and you’ll be the rock star of marketing.”
Blending AI with your human Fire & Security insight
Looking for innovative ways to embrace AI AND the human touch in your Security Marketing?
We can help you produce memorable social media posts that get your company noticed.
Get in touch today.