The Four Jobs To Be Done in Accounting

When people talk about AI in accounting, the conversation often drifts into broad promises and vague headlines. But accounting isn't vague. It's structured, outcome-driven work, built on repeatable processes that firms execute every single day.

That's why it's more useful to frame this conversation around jobs to be done: the foundational tasks that underpin how firms operate and deliver value.

Across disciplines and firm sizes, four show up again and again: Getting the data. Understanding the tax or compliance implications. Producing outputs (like reports or files). Advising clients.

Whether you're closing the books, filing taxes, preparing for audit, or offering outsourced finance, some version of these four jobs shows up in every engagement. They cut across service lines, tech stacks, and org charts. They are the backbone of accounting work.

And they all have the potential to be meaningfully impacted by AI.

1. Data Collection & Ingestion

This is the unglamorous but essential backbone of accounting work. Pulling in raw data from disparate sources like bank feeds, GLs, payroll systems, AR/AP platforms, ERPs, point-of-sale systems, even emailed spreadsheets, and structuring it so the rest of the process can begin.

Matching files. Chasing formats. Cleaning inconsistencies.

Why it matters: Every workflow depends on accurate, timely data. If the inputs are wrong, the outputs collapse. This job is about creating the foundation, one clean, connected source of financial truth. For most firms, and rightfully so, this is where the time goes.

Why AI is a good teammate here: Because data collection tends to be repetitive, pattern-based, and largely rules-driven. It's also work few enjoy. AI (especially when paired with traditional automation) excels at parsing formats, mapping fields, and normalizing inputs across messy systems.

What's the impact: Internally, it frees up junior capacity, reduces manual errors, and speeds up the monthly close. Externally, it shortens client wait times. Faster answers. Happier clients.

Is it feasible today? This is already within reach. AI can match vendor names, identify duplicates, propose journal entries, and extract transaction-level data from unstructured documents. Soon we'll see smarter ingestion that learns client-specific patterns and flags anomalies automatically.

2. Research & Retrieval

This is the detective work behind accounting. Whether it's understanding a niche tax treatment, referencing audit standards, or looking up how a similar transaction was handled, research supports compliance and judgment.

Why it matters: Research is how accountants apply context. It's also how they avoid costly mistakes. No one wants to make a treatment decision in a vacuum.

Why AI shines as a co-pilot: AI can surface relevant guidance fast, flag treatment anomalies, and cut down the time spent digging through docs. Because research is mostly about finding the right thing fast, large language models are excellent at pulling the right rule, summarizing guidance, or retrieving prior context.

What's the impact: Internally, faster onboarding for juniors, smarter reviews, and more consistent decisions across teams. Externally, clients get confident answers faster.

Is it feasible today? Retrieval-augmented generation works well now for static guidance and precedent search. Custom-trained copilots are already in play in leading firms.

3. Report Generation & Filing

The formal output. Whether it's an audit report, tax return, board-ready financials, or a month-end close package, this job is about structuring, summarizing, and presenting financial information. It's high-stakes, deadline-driven, and detail-heavy.

Why it matters: This is what clients, regulators, and stakeholders see. It's the "finished product" of your work. Accuracy, timeliness, and consistency are everything.

Why it's primed for AI: Once data is structured, generating reports becomes a formatting problem. AI can draft, populate, and even narrate reporting while following firm or industry templates.

What's the impact: Internally, it reduces back-and-forth and gives reviewers a head start. Externally, clients get faster reports with more room for narrative and strategic insight, not just numbers.

Is it feasible today? Basic population of templates, variance analysis, and simple commentary is doable. Output formatting is improving, though Excel manipulation still trips up AI.

4. Client Service & Strategic Advice

The most human part of the job and the most irreplaceable. Whether it's answering client questions, flagging risks, or advising on margin improvement, this is where firms move from vendor to partner. This is about making sense of the data, providing interpretation, and offering forward-looking guidance.

Why it matters: This is where firms prove value. It's what keeps clients coming back. It's also what turns accounting from reactive to strategic.

What's AI's role here? Keep accountants informed, proactive, and responsive. AI can spot patterns, surface insights, and even suggest next steps. But judgment, empathy, and trust? Still human territory.

What's the impact: Clients experience more proactive communication. They feel guided, not just informed.

Is it feasible today? AI can already assist with insight surfacing, benchmarking, and basic what-if analysis. But nuanced recommendations still require human input.

TL;DR: Start With the Job That's Slowing You Down

Every firm has a bottleneck. A job that's eating capacity, delaying delivery, or draining energy. Instead of trying to "AI your firm," pick the job that's costing you the most and fix that one.

Which of these four jobs is eroding your team's time today?