If there’s one word to sum up the business sentiment towards AI going into 2025, it’s FOMO.
Fear of missing out on innovations that competitors have already worked out. Fear of wasting money on activities that others can now automate. Fear of being too slow to act.
We’re all aware of AI’s productivity potential. Now we’re itching to see some results.
The good news is that brands are ready to spend. In 2025, the number of US companies investing over $10 million in AI is set to double.
They aren’t just throwing money at the wall either.
According to a KPMG study, 73% of leaders are planning to recruit new board members to help guide generative AI within the business. And after a few years of half-hearted exploration, this feels like exactly what’s needed to drive meaningful adoption that boosts productivity.
The question, then, is where to spend that money. But first, let’s look at where not to.
Forget about AI for creative tasks
The CEO of OpenAI, Sam Altman, reckons that “95% of what marketers use agencies, strategists, and creative professionals for today will easily, nearly instantly, and at almost no cost be handled by the AI”.
I call bullshit.
The only people getting excited about the stuff AI creates are geeks and suits – geeks because of the tech behind it and suits because they don’t know what great creative work looks like.
As Freelance Content Marketing Writer, Kris Hicks explains, “No agency worth hiring will use it for 95% of their work (or 15% for that matter). It’s time for the hype around this technology to cool. It has some limited uses it’s genuinely good for, but far less than its proponents have been promising. And ‘creativity’ is not on the list.”
The big problem is that generative AI is uncreative by design.
The only reason AI looks and sounds half-convincing is because the results contain elements of things you’ve seen a thousand times before. So when you write a prompt for a piece of copy or design, it deliberately gives you something that blends in with stuff that’s already out there.
And that’s why AI fails on the three core principles of great marketing:
- Stand out – it can only ever give an amalgamation of existing ideas
- Say something new – it doesn’t have anything original to say or communicate
- Make an emotional connection – it doesn’t understand people
“I believe that AI will never be able to produce original ideas,” says Nikola Baldikov, CEO and Founder at Inbound Blogging. “AI primarily operates by analyzing and remixing existing data rather than generating something of its own, which means it lacks the capacity for genuine innovation or intuition.”
Four processes to optimize with AI in 2025 and how much time you’ll save
The key to successfully investing in AI is identifying existing processes that are slow, ineffective, and unsatisfying. Then you can explore ways to optimize those processes with AI.
Let’s dive into four specific examples – plus how much time you can expect to save.
1. Searching for information – 205 hours/year per employee
Employees spend 1.8 hours per day searching and gathering information. That’s a whopping 468 hours when you add up all the working days in a year!
This covers a wide range of tasks like:
- Carrying out market research
- Investigating competitors
- Looking for company policy documents
- Tracking down old messages
- Finding the right version of a file
In recent years, tools like Slack, Asana, and even Filestage have rolled out advanced search features to help you find what you need faster. But if we estimate that around half our searching time is spent looking for information on the web (which feels pretty conservative to me), there’s still a lot of potential to boost productivity.
On top of that, research shows that it can take up to eight searches to find what you’re looking for. So if AI can help cut that to just one search every time, employees will save a huge chunk of time.
Let’s crunch some numbers to see what that could look like:
- 468 hours – total time spent searching for all types of information per year
- 234 hours – an estimate for how much of that time is spent searching for information online
- 205 hours – the total time saved by finding what you need in one search instead of eight
AI search tool, Perplexity, accelerates the research process and includes links to the original sources.
Steve Sacona, Founder and Legal Writer at Top 10 Lawyers has experienced this time saving first hand.
“Writing legal content was slow and costly,” explains Sacona. “It took hours for people to research laws, check cases, and then write articles. The process often had mistakes or missed details, which hurt trust with readers.”
But with AI search tools like Perplexity, employees can rapidly accelerate this process.
“We now use AI to quickly search legal databases, find key information, and create article drafts. Writers step in to check, edit, and add a human touch. This saves over 60% of the time and keeps the quality high. AI does the heavy work, and humans focus on what matters most.”
“Some might think relying on AI is risky,” says Sacona, “but it’s actually very accurate. AI doesn’t get tired or miss things. Humans still check everything – so it’s a strong mix of speed and trust. The process is not just faster – it is smarter. This shows that the best results come from using both AI and people together.”
2. Staff scheduling and resource management – 170 hours/year for managers
The next inefficient process we’re going to talk about is resource management.
Source: YouTube
Research from Asana shows that around 60% of employees’ time is spent on work coordination. That’s stuff like scheduling tasks, lining up resources, and facilitating collaboration between team members.
This is another area where AI is changing the game.
“Before using AI, our scheduling process relied on manual coordination,” says Gal Cohen, Field Area Manager at JDM Sliding Doors. “Customers would call or email to book an appointment, and our team would check calendars, confirm availability, and assign technicians based on proximity and skill. This process was slow, prone to errors, and usually left gaps in our schedules.”
“Now, with Jobber, an AI-driven scheduling and dispatch tool, the process is almost completely automated. Jobber instantly matches incoming appointments to the best technician by analyzing location, skill set, and availability.”
This type of resource planning improvement is predicted to save 170 hours every year for managers. And, as Cohen explains, the quality is up to scratch too.
“I’m confident in the quality of this AI-enhanced process because it eliminates the repetitive and error-prone tasks that humans struggle with, like dealing with multiple variables at once. The AI hasn’t replaced our human touch. In fact, it complements it, freeing us to focus on what we do best: delivering excellent service.”
3. Non-creative writing tasks – 400 hours/year per employee
Studies show that US professionals spend around 19 hours per week on writing tasks – or 988 hours per year. This includes everything from writing internal emails to producing reports and presentations.
People agonize over these kinds of communications. But why? Nobody reads them as carefully as your employees write them. All your team needs to worry about is getting the main points across as clearly as possible.
That’s something AI is great at. And by using tools like ChatGPT, employees can expect to cut their time spent on writing tasks by 40%.
This is how SpeakNotes turned my waffly voice clip into a clear, concise message.
Here are a few examples of how AI can save your organization time on writing tasks:
- Voice-to-text – Use a tool like SpeakNotes to convert a voice recording into a short summary, cutting a 30-minute task down to three minutes.
- Automated meeting notes – Invite an AI notetaking app like Fireflies to your calls. So meeting owners can share the notes instantly without writing them up themselves.
- Content summarization – Use ChatGPT to summarize a long document or describe a piece of design. (Fun fact: I asked it to describe a holiday photo I uploaded and it correctly guessed that it was taken in Porto, Portugal!)
Speaking of describing images, Jay Egger, Senior SEO Manager at SBG Funding, has found a great way to save time on writing tasks with AI.
“Before the use of LLMs, I would handwrite all alt-text on images,” says Egger. “Although they don’t take too much time, generally anywhere from 30 seconds to one minute per image, when you have a lot of images, it can really be hard to scale.”
“I have created a custom tool that makes use of LLMs and GPT Vision. This allows me to batch upload a set of images and it will generate the alt text for me at scale. This has sped up my process significantly and allows me to focus more on my core tasks.”
And in this case, faster really is better.
“To be honest, I’ve found that responses are probably higher quality than humans,” says Egger, “mainly because the bot doesn’t get stuck on decisions and the task is purely based on description; it’s intentionally non-creative.”
4. Reviewing creative assets – 300 hours/year per employee
The fourth and final process we’re going to talk about is one we know a lot about at Filestage.
A few years ago, our study found that marketers spend three hours per day giving and receiving feedback on creative assets. That’s 780 hours per year! And it was the most senior (and therefore expensive) staff who were impacted the most.
Luckily, there are ways to cut feedback costs with AI. Here’s how.
At Filestage, we estimate that 50% of feedback time is spent reviewing files against brand guidelines and quality standards. This covers very basic and repetitive things like checking:
- The correct logos and safety badges are present
- The ingredient volumes add up
- The colors match your brand guidelines
- The font sizes are consistent
- The URL points to the right website (we all remember that Wicked packaging fail)
But, often, it also requires in-depth knowledge of industry regulations. In other words, rules. And if there’s one thing AI is great at, it’s following rules!
By using AI assistants as the first line of defense in the approval process, employees could get a huge chunk of this time back. This would boost productivity dramatically and create space in the workweek for more meaningful tasks.
Filestage’s AI assistants can add comments and annotations to designs based on FDA regulations.
Final thoughts
Between economic uncertainty and an ever-more competitive business landscape, this is the year that companies need to start seeing the productivity gains AI has promised.
My tip? Unless you want your brand to blend in with your competitors, steer clear of using AI for creative applications like copywriting and design.
Instead, try switching from Google to Perplexity to find information faster. Or, next time you’re about to write an email to your teammates or clients, give SpeakNotes a go. I know I will.
