A few years ago, it felt like the narrative was settled: “Accounting is going to be automated.”
To be fair, a lot of it should be.
Modern tools are incredible. Bank feeds, rules, auto-categorization, OCR, workflow automations, real-time dashboards - all of that has changed what's possible. Personally, we've embraced it. We're early adopters by nature, and we've seen first-hand how the right tech removes friction and makes financial operations feel lighter.
However, here's the stance I'll take pretty clearly:
In a tech-heavy world, accounting still needs to be human-led. The most consequential financial decisions depend on judgment, not software.
The “accounting will be automated” story isn't necessarily wrong; it is, however, incomplete.
“Tech will replace accountants”
If your books were simply:
- a predictable set of transactions
- clean integrations
- consistent timing
- a stable chart of accounts
- and zero edge cases
…then yes. Software could do most of the work. For certain parts of the process, it already does. However, that's not what real businesses look like, most especially once you're growing.
Where the automation story falls short
Automation works best in stable, predictable environments. A consistent flow of transactions. Clean integrations. Clear rules that don't change very often. In those conditions, software can do a lot of the heavy lifting, and it already does.
The problem is that most real businesses don't stay in that state for long.
What usually causes accounting to break is change. The business evolves faster than the systems do. Teams use tools differently. Revenue models shift. Pricing becomes more layered. Entities multiply. Cash tightens. Suddenly, leadership is asking questions that don't sit neatly in a dashboard. That's where the automation-first narrative starts to break.
What technology does well and where it stops
Tech is very good at speed and consistency - that's a good thing. Automations handle repeatable, rules-based work extremely well: pulling transactions, matching payments, generating drafts, routing approvals, flagging obvious anomalies. That baseline efficiency is exactly what modern accounting should have.
Where it stops working on its own is the moment judgment is required.
As soon as the questions shift to things like whether something should be capitalized or expensed, whether a cost truly belongs in COGS or operating, whether revenue is being recognized correctly under a specific contract, why gross margin moved, whether cash is actually improving or just being delayed, or what's really driving a variance, you're now asking for interpretation, and interpretation is where experience shows up.
“Human-led” doesn't mean anti-tech
When we talk about being human-led, we're not talking about doing things manually or resisting technology. We use plenty of it.
What we mean is accountability.
Someone owns the outcome, not just the task or the workflow. Someone understands the business context and knows that context changes how numbers should be categorized, reported, and interpreted. Someone catches the moments where the numbers technically “work” but don't reflect what's really happening. Someone helps turn financial output into decisions because clean books are just the baseline. Decision-ready books are what move a business forward. That's the real distinction.
The hidden risk of automation-first accounting
One of the more subtle risks I've seen over the years is that you can automate a mess.
You can automate incorrect mappings.
You can automate inconsistent processes.
You can automate a chart of accounts that doesn't match how leadership thinks.
You can automate delayed reconciliations.
You can automate “good enough” reporting.
When that happens, the system still produces output (and it often looks polished). This is where founders get blindsided later - they ignored accounting, because they assumed the system was giving them the truth when it was really just giving them results.
Automation scales what you already have. If what you already have is shaky, automation just helps it travel faster.
What human-led accounting looks like
Human-led accounting isn't exactly complicated. For most of the companies we work with, “winning” looks like a close that's predictable and doesn't take over the month, reports that stay consistent from one period to the next, a balance sheet that actually ties out, margins that are meaningful, and cash flow that becomes clearer before it turns into a problem. Decisions get easier because the data is usable.
Technology plays a big role in making that happen. Getting there usually requires people who can design systems around how the business actually operates, enforce discipline as things change, ask the uncomfortable questions early, and translate financial output into decisions leaders can act on.
Why this matters as businesses grow
This usually starts to matter once a business moves beyond the early, simple stages. What used to be straightforward becomes layered, and revenue stops being just “money in.” Costs don't sit neatly in one bucket. Cash timing starts to matter as much as profitability. Different teams interact with systems in different ways, and small inconsistencies begin to compound.
On the surface, the tools might still show everything running smoothly. Underneath, the questions leaders need answered become more nuanced: what's really driving performance, what's sustainable, and where risk is quietly building. Technology is excellent at surfacing activity. Making sense of that activity and understanding what it means for the next decision still requires human judgment.
Looking ahead
The future isn't “tech versus humans,” ‘it's tech, guided by people who know what to look for and what questions to ask.
If you're building a serious company, you don't want accounting that's merely automated. You want accounting that's reliable, decision-ready, and owned by people who understand the stakes.
That's why we continue to lean into being human-led. We've seen what happens when companies grow faster than their financial foundation, and we know how to build it properly.
Curious to hear from you: where has automation helped your finance function the most, and where has it created blind spots you didn't expect?
