Why “Having a System” Isn’t the Same as Having an Advantage

Here’s a pattern we keep running into with new clients. Two businesses, same size, same industry, same amount of hustle. One of them is automated. The other isn’t. From the outside, they look about the same. Six months later, they don’t. That gap has a name now, and it’s not just a feeling, in fact the data backs it up. 

But first, a quick definition, because “automation” means something different depending on who you ask. For a marketing team, it might mean emails that they send themselves. In operations, it’s a workflow that routes itself without anyone forwarding anything. For us, it’s simpler than the two above: automation is any system that does a repeatable task for you, correctly, without someone having to remember to do it manually every time. That’s it. It doesn’t have to be AI and most certainly it doesn’t have to be complicated. This month we want to talk plainly about what that actually does for a business, what it doesn’t do, and what happens if you sit this one out. 

First, let’s start with the boring but important part: money and time.

Deloitte’s global survey of companies running automation past the pilot stage found an average cost reduction of around 16–32%. In fact that’s not a marketing number, that’s what happens when repetitive work stops eating hours that could go toward the work that actually grows the business.

Speaking of hours: Zapier’s survey of small-business knowledge workers found marketers save an average of 25 hours a week when their repetitive tasks are automated. IT and customer service teams save close to that too. That’s not a slightly lighter workload. That’s most of a second work week, every week, given back.

And it isn’t just about doing more, faster. It’s about doing it right. For example IBM’s Cost of a Data Breach research found companies using automation extensively in their security operations paid $1.88 million less per breach on average, and caught problems nearly 100 days faster, than companies that didn’t. Speed isn’t just a convenience. Sometimes it’s the difference between a bad day and a very bad year.

There’s a people side to this too, and it’s easy to underestimate. Zapier’s “How We Work” report found that only 14% of employees who use automation at work had considered quitting, compared to 42% of those stuck doing everything manually. Nobody enjoys spending their week on the parts of the job a system should be handling.

We’d be doing you a disservice if we stopped there, because automation isn’t magic, and pretending otherwise is exactly the kind of overselling we try not to do.

Ernest & Young has said publicly that it sees 30–50% of first RPA (robotic process automation) projects fail. Likewise, Gartner is predicting that more than 40% of “agentic AI” projects will be cancelled by 2027, mostly because they were rushed, poorly scoped, or never had a clear business case to begin with.

And over-automating is a real risk, not a hypothetical one. Tesla famously over-automated its Model 3 production line in 2018, and Elon Musk publicly admitted it was a mistake — the robots slowed things down that a person could have done better. McDonald’s spent three years testing AI-powered drive-thru ordering and pulled the plug in 2024 after it kept mishearing orders in the noisy, unpredictable reality of an actual drive-thru.

The lesson isn’t “don’t automate.” It’s “know which parts of your business are stable and repetitive enough to hand off, and which parts genuinely need a human paying attention.” That distinction is most of what separates automation that pays for itself from automation that becomes an expensive lesson.

Gartner expects that by 2028, at least 15% of everyday business decisions will be made autonomously by AI systems, up from essentially zero in 2024. The broader automation and AI-agent market is projected to more than double in size over the next several years, depending on whose estimate you read.

But here’s the number that matters more than any market-size projection: McKinsey found that 88% of companies are now using AI somewhere in the business, and yet only about 6% are actually seeing meaningful results from it. That means almost everyone is in and yet, almost nobody has figured out how to make it count yet.

The gap between adopting a tool and actually benefiting from it, is where most of the real opportunity sits right now. We’ve passed the point of who gets the new model first, but who is being one of the few that uses it properly.

This is the part that doesn’t get said enough: not automating isn’t a neutral choice. It’s a decision with its own cost, it’s just one that shows up quietly.

Salesforce found that 75% of small-business leaders already feel like they’re falling behind their competitors on technology. That’s not a handful of worried owners, it’s three out of every four business owners you know.

And it shows up in ways you can measure directly. A well-known study out of MIT found that a business that responds to a new lead within 5 minutes is roughly 100 times more likely to make contact, and 21 times more likely to qualify that lead, than one that waits 30 minutes. Harvard Business Review found the average company takes 42 hours to respond to a new lead, and 23% never respond at all. If your competitor’s system replies in seconds and yours replies whenever someone gets to it, you’re not just a little behind. You’re losing customers you never even knew you had. 

If you’re reading this and thinking “okay, but where do I even begin,” here’s the honest answer: not with the flashiest tool. Start small and boring. Something like your inbox auto-sorting itself into folders by client or urgency, instead of you triaging it every morning, is automation. So is connecting a tool like ChatGPT or Claude to draft first responses to routine emails, or having a form on your website automatically create a task instead of sitting in someone’s inbox until they remember to act on it. Get one of those working, notice what it frees up, and only then look at the next repetitive thing eating your week.

The bigger judgment call is knowing which tasks are safe to hand off that way, and which ones still need a person paying attention, a routine email reply is one thing, a decision about pricing or a client relationship is another. Automate the first kind. Protect the second. Measure what changes before you automate the next thing.

And here’s the part worth being blunt about: don’t build a system just because everyone else seems to have one. A lot of what gets sold as “automation” right now is noise:  impressive-looking dashboards nobody asked for, solving problems nobody actually has. The real question was never “should we have a system.” It’s “what does this business actually need a system to do,” and that answer is different for every company. A system should support a way to move faster on the judgment calls that are already yours, not take over the decision making process.

We didn’t just write about this. We built CARIM, our own internal system, around exactly this idea: no forced workflow, no feature we couldn’t point to a real need for, built to fit how our team actually works rather than the other way around. It’s also, not coincidentally, the difference between the 6% McKinsey found actually benefiting from AI and the 88% who are just using it.

If you’ve been putting this off, or you tried it once and it didn’t stick, we’re always happy to talk through where the real leverage is in your specific business. No pitch required to have that conversation.

If this resonated, share it with someone building something of their own. That’s usually where the best conversations start.

– FutureLab Team
Invent the future with friends.

Sources
Asana. (2022). Anatomy of work index 2022: Work about work hampering organizational agility. Asana. https://www.nasdaq.com/press-release/asana-anatomy-of-work-index-2022:-work-about-work-hampering-organizational-agility

Deloitte. (2022). Automation with intelligence 2022 survey results. Deloitte Insights. https://www.deloitte.com/us/en/insights/topics/talent/intelligent-automation-2022-survey-results.html 

Ernst & Young. (2016). Get ready for robots: Why planning makes the difference between success and disappointment. EY. https://eyfinancialservicesthoughtgallery.ie/wp-content/uploads/2016/11/ey-get-ready-for-robots.pdf 

Gartner. (2025, June 25). Gartner predicts over 40% of agentic AI projects will be canceled by end of 2027 [Press release]. https://www.gartner.com/en/newsroom/press-releases/2025-06-25-gartner-predicts-over-40-percent-of-agentic-ai-projects-will-be-canceled-by-end-of-2027

Harvard Business Review. (2011). The short life of online sales leads. Harvard Business Review.

IBM Security, & Ponemon Institute. (2024). Cost of a data breach report 2024. IBM.

McKinsey & Company. (2023, June). The economic potential of generative AI: The next productivity frontier. McKinsey Global Institute.

McKinsey & Company. (2025). The state of AI in 2025. McKinsey Global Survey on AI.

Oldroyd, J. B., Elkington, D., & the InsideSales.com research team. (2007). Lead response management study. MIT Sloan School of Management / InsideSales.com. https://www.onecavo.com/wp-content/uploads/2015/11/MIT-InsideSales.com_Lead-Response-Management.pdf 

Salesforce. (2024). Small & medium business trends report (6th ed.). Salesforce.

Zapier. (2021, June). Automation and the American workforce: How SMB employees use automation. Zapier.

Zapier. (2021, August). How we work report. Zapier. https://zapier.com/blog/how-we-work-report/