Darin Patterson of Make: Automation Must Come Before AI Implementation
I'm back with another contribution today from Darin Patterson, VP of Market Strategy at Make. After the great feedback we had from Darin's Q&A discussing their newly release agent functionality, I asked if he'd be keen to write an opinion piece on the wider topic of AI implementations. So here we go!
Over to you Darin:
AI agents present some of the greatest business opportunities of this century, so it's no surprise that many executives are eager to explore them. Too often though, organisations rush to implement AI without considering how it will integrate with their existing technology, making it easy to make mistakes. In fact, Gartner forecast that 40% of agentic AI projects will be cancelled by 2027, due to misapplication or poor integration with existing systems. It's crucial organisations take a clear-eyed look at their existing technology foundations before making the leap
Taking a reality check
Most organisations operate within complex systems which support different parts of the business. Data is siloed and governance structures vary across the business therefore simply plugging in AI is unrealistic. For AI to be truly effective, it needs to connect and make sense of all the data across these environments. This means AI alone isn't the starting point, but rather part of a broader integration effort.
People spend their days clicking buttons, dragging files and copying the strings of text. Today, these tasks are fundamental to keeping businesses running smoothly, but executing them manually is not the best use of your team's intellect or energy.
Businesses need to think not just about what AI can make possible but how they can enable it in their current systems. In order to be effective, AI needs good input as well as structured and timely data which is a challenge for businesses with disconnected systems, unclear data flows and limited governance.
Automation as the starting point
Automation empowers teams to cut down the time they waste on repetitive administration, and focus on innovating and making the most of their talents and skills. It also offers the perfect launch pad for businesses embracing the power of AI.
To really unleash the potential of automation, no-code or low-code approaches help to put power in the hands of business users. No-code automation means users can automate and streamline workflows without using code: it is visual, simple to understand and allows people to automate processes without using programming languages. For organisations, that means that business users can build technology to suit their day-to-day needs.
From operations to marketing, those that innovate and drive change are empowered to do so. Automation often provides the efficiencies and improved business agility that business leaders hope to achieve through AI, and it forms the perfect bedrock for layering in AI going forward.
"The difference no-code makes here is that someone who is closer to the business can design, implement and maintain automated solutions"
Reaping Automation ROI
In many cases, traditional automation alone delivers massive return on investment (ROI) through saving time, improving business productivity and adding scalability.
The difference no-code makes here is that someone who is closer to the business can design, implement and maintain automated solutions, rather than handing the work to a business analyst who hands it to developers who then miss out on half the issues the business was trying to solve in the first place. As a result, we have seen many businesses who initially research technologies such as AI agents in hope that it can help their business, but then they learn about the benefits of automation and choose just to automate instead.
"The HR department were able to 'skill up' to create their own automations within a few days"
Bolt, an Uber competitor working in the ride-hailing and e-scooter space, aimed to automate their HR processes, and opted for a no-code approach. The HR team previously had to rely on their IT team to create software for them: this meant that any changes to workflows could take up to half a year. Using no-code, people within the HR department were able to 'skill up' to create their own automations within a few days, and then change processes and drive automations to make the department more efficient.
Unlocking the AI level
By automating routine and structured tasks, organisations unlock the ability to layer in AI-driven capabilities. Once information and actions are able to move from system to system, AI can be introduced and can be used for simple tasks like content generation where it excels. It can also be used for more complex tasks where decisions can be made, customer interactions are handled or contextual data is analysed in real time.
Over time, businesses will be able to build true AI agents. AI that does not just generate insights but takes meaningful action. These agents collect data, make judgements and access internal systems to drive outcomes, all of which are built on the foundation of automation which connects everything together. In other words, automation sets the stage; AI brings adaptability, scalability, and real-world impact. Without the automation layer you will just be prompting, rather than orchestrating your processes in a meaningful way.
Building for success
AI is already moving beyond supporting business processes to actively driving them, but it's key to put in place a solid base of automation rather than impulsively implementing AI without thinking about how it will fit in. Automation streamlines processes across the business, making the organisation more responsive and able to deal with change, including technological change.
Once automation has streamlined workflows, businesses will be ready to reap the benefits of artificial intelligence, creating responsive, intelligent organisations fit for tomorrow.
Thank you very much, Darin. Absolutely brilliant – and I totally agree. I am continually seeing many organisations rushing to do "the AI bit" when actually, the crucial step before is to automate as much as possible. Indeed, I often see this adding more value (in the short term) than any AI activities.
You can read more and connect with Darin via his LinkedIn profile and find out more about Make at www.make.com. I'm a massive user of Make myself – and much of Conversational AI News is produced in part by a whole set of automations I've built. If you haven't checked it out, create an account today and test it out.