You already know what you don’t need. You don’t need another platform that makes onboarding easy for the business owner while hiding the chart of accounts three menus deep. You don’t need a slick dashboard that summarizes revenue nicely but can’t produce a trial balance without exporting to Excel. And you definitely don’t need a system where your client’s categorization of a $14,000 vendor payment as “Miscellaneous” gets locked into the ledger before you ever see it.
Most client accounting software is designed to impress the person writing the check, not the person doing the books. That’s understandable from a sales perspective, but it creates real friction for the accountant or bookkeeper who has to work inside these systems daily, often across a dozen or more client files.
If you’ve spent years navigating that friction, you’ve probably developed a mental checklist, the things you actually test before recommending software to a client. What follows is that checklist, made explicit. Not a feature comparison grid. A set of questions that separate tools you can work with from tools that will slowly drive you out of your mind.
Can you get to the general ledger without a treasure hunt?
This sounds basic. It is basic. And yet, a surprising number of platforms bury the GL behind layers of navigation designed for people who don’t know what a general ledger is. The interface prioritizes invoicing, expense tracking, and cash flow widgets. The actual books are an afterthought.
For an accountant, the general ledger is the work. Everything else is data entry that feeds into it. If you can’t get to the GL, review journal entries, trace a transaction back to its source, and make adjustments without navigating through five screens, the software is fighting you instead of helping you.
The test is simple: open the software for the first time and try to post a manual journal entry. If that action requires you to leave the main interface, search a help article, or activate an “advanced” feature, the platform wasn’t built with you in mind.
How does the chart of accounts actually work?
A chart of accounts isn’t a list. It’s the structural framework of the entire financial record. And the way a platform handles it tells you almost everything about whether the software was designed by people who understand accounting or by people who think accounting is just categorizing transactions.
What matters: Can you create parent and sub-accounts with flexible numbering? Can you merge accounts without losing transaction history? Can you lock accounts to prevent client-side changes? Can you set up accounts that mirror the way professional services firms actually operate, with proper separation between labor costs, subcontractor expenses, direct project costs, and overhead?
QuickBooks Online handles the basics here, but customization gets clunky fast. Xero gives you more structural flexibility but limits you on the number of accounts in certain plans. FreshBooks barely qualifies as having a chart of accounts in the traditional sense.
For professional services clients specifically, you need a chart of accounts that can distinguish between revenue by service line, direct labor costs by role or department, and overhead allocated across projects. If the chart of accounts can’t support that granularity, you’ll never produce meaningful profitability reporting for your client, which limits your ability to deliver advisory work.
Is the bank feed actually reliable?
Bank feed quality varies enormously between platforms, and this is the area where marketing claims diverge most sharply from daily experience. Every platform will tell you they connect to thousands of institutions. Fewer will tell you how frequently the feed syncs, how it handles pending transactions, what happens when the connection drops, or how categorization rules perform over time.
The questions that matter from a practitioner’s perspective: Does the feed sync at least daily? Does it distinguish between posted and pending transactions, or does it import pending transactions that later change, creating reconciliation headaches? When the connection breaks, and it will, how easy is the reconnection process? Does the client need to re-authenticate, or can you do it from the accountant portal? Does the platform learn from categorization corrections, or do you re-categorize the same recurring charge from the same vendor every single month?
Bank feed reliability is the kind of thing you can’t evaluate in a 14-day trial. It only reveals itself over months of real use. If you’re evaluating software for a new client, ask other practitioners about feed stability with your client’s specific bank. That saves more headaches than any feature demo.
What does the audit trail actually capture?
An audit trail that records “Transaction modified by User on Date” is marginally useful. An audit trail that shows you the previous value, the new value, who made the change, and when, and that can’t be edited or deleted, is essential.
For accountants, the audit trail serves two purposes. First, it protects you. When a client’s numbers don’t match what you saw last month, you need to determine whether someone reclassified a transaction, deleted an entry, or changed an amount. If the audit trail doesn’t preserve the before-and-after state of every modification, you’re reconstructing history from memory and guesswork.
Second, it makes year-end and tax prep faster. A clean audit trail means fewer questions for your client, fewer reconciliation mysteries, and fewer hours spent figuring out why the balance sheet doesn’t tie. Over a 12-month engagement, the quality of the audit trail directly affects your realization rate on that client.
Can your client use it without destroying the books?
This is the tension at the center of every accountant-client software relationship. Your client needs to interact with the system: sending invoices, recording expenses, maybe logging time. But every interaction is a potential data quality problem. A miscategorized transaction. A duplicated entry from a manual and an automated import. A deleted invoice that throws off the AR aging.
The solution isn’t to lock the client out. It’s role-based access that gives them the tools they need and nothing else. Can you configure permissions so a client can create invoices and record expenses but can’t modify the chart of accounts or delete reconciled transactions? Can a team member who only needs to submit timesheets access just the time tracking module without seeing financial data?
This is particularly important for professional services clients with employees. The person logging billable hours shouldn’t see the firm’s P&L. The office manager processing vendor invoices shouldn’t be able to modify payroll accounts. The business owner should see reports and approve invoices, not edit journal entries. When the permission model is this granular, your cleanup work drops significantly.
How painful is the multi-client workflow?
If you manage fifteen clients, you log in and out of fifteen separate company files every day. Some platforms mitigate this with an accountant dashboard that provides a unified view across clients. Others don’t. The difference in daily workflow efficiency is substantial.
The accountant dashboard question goes beyond convenience. It affects what you notice. When you can see overdue bank reconciliations, uncategorized transactions, and anomalous balances across all your clients in a single view, you catch problems earlier. When each client lives in a silo that you have to individually open and review, small issues accumulate into big ones because they’re invisible until you actively go looking.
What separates a genuinely useful multi-client view from a cosmetic one: Can you see reconciliation status across all clients without opening each file? Can you jump directly from the dashboard to a specific transaction in a specific client’s books? Do notifications surface issues you need to act on, or do they just tell you things happened? Can you perform batch operations like running standard reports or checking outstanding items across multiple clients at once?
Two-thirds of accountants report feeling overwhelmed at least weekly by the complexity of their tech stacks. A significant part of that overwhelm comes from context-switching between client files in systems that weren’t designed for multi-client workflows.
Does the reporting actually help you advise?
Standard financial statements, the balance sheet, income statement, and cash flow statement, are table stakes. Every platform produces them. The question is what else you can get.
For professional services clients, the reports that drive advisory conversations aren’t the standard financials. They’re project profitability by client, utilization rates by team member, effective billing rates over time, aging analysis segmented by client type, and work-in-progress reports that show unbilled time. If the accounting software can’t produce these natively, you’re building them manually, which means they either don’t get built at all or they cost you unbillable hours every month.
The advisory services opportunity is real. Nearly 80% of accountants expect advisory work to grow over the coming year. But advisory work requires data the client can’t get from a basic P&L. When your software produces that data as a natural output of well-structured books, advisory becomes a conversation you can have during a regular client meeting. When it requires a separate analysis built in a spreadsheet, it becomes a project that gets deferred indefinitely.
The real evaluation happens after the trial
A 14-day trial tells you whether the interface is tolerable. It doesn’t tell you whether the bank feed will hold up through month-end close, whether the audit trail will save you during tax season, or whether the permission model will prevent your client from accidentally uncategorizing three months of reconciled transactions on a Saturday afternoon.
The meaningful evaluation criteria for accountants aren’t about features you can test in a demo. They’re about structural decisions the software made that you’ll live with for years: how the GL is organized, how the chart of accounts behaves under customization, how reliably data flows from external sources, and how much control you have over what your client can and can’t touch.
If you’re evaluating software for professional services clients specifically, add one more criterion: does the platform treat time as a first-class financial input? In a services firm, time drives revenue, costs, and profitability. If time tracking is an add-on, an integration, or a separate subscription, the software has already told you that it doesn’t consider time central to the financial model. For a firm that sells hours, that’s a fundamental architectural mismatch.
CarteFi gives accountants a dedicated portal with a unified view across client organizations, granular role-based access control, native project accounting, and an audit trail designed for practitioners. Two accountant seats per client organization, included at no additional cost. Learn more about CarteFi for accountants.