With 21 locations, I’m pretty much always hiring. Unfortunately, the employment market these days can be chaotic, as candidates send out applications across dozens of job boards with a single click. For managers like me, this results in more time spent sorting through signals and static.
I love numbers. Give me a profit-and-loss statement, a sell-through report or an idle inventory analysis showing me which toys and lubes are moving and which ones are gathering dust, and I’m in my element. What I don’t love? Copy and paste, download, upload, log in, sync, fix the error, repeat tomorrow.
It’s not about outsourcing your work to a computer; it’s about removing the friction between the work that matters and the results it delivers.
Unfortunately, that’s often the reality of accounting, in adult retail just as in other sectors. The insights are valuable, but obtaining them requires data entry and reconciliation routines repetitive enough to make a CPA cry.
We use Lightspeed for our POS, QuickBooks Online for accounting and iSolved for payroll. Each has integrations with the others — in theory. In practice, I download payroll summary reports, upload them to the QuickBooks integration, verify that invoices match what the store received, and monitor account settings whenever someone changes a configuration. This does the job, but it’s slow and definitely not an efficient use of time.
That’s why I adjusted my approach — and the results have been eye-opening.
‘Skills’ Versus ‘Tasks’
There’s a lot of handwringing out there about AI taking over jobs, and I get it. We don’t want machines doing work that requires human judgment, insight and discretion. Those are some of the skills that AI has yet to master. Naturally, the same is true with skills unique to adult retail, such as helping customers feel comfortable enough to engage on highly personal topics.
Let’s make a distinction, however, between “skill” and “task.” Understanding which numbers matter, noticing when something seems off and asking the right questions about your data are skills where human judgment proves its worth. Copying and pasting, on the other hand, isn’t a skill. Neither is downloading and uploading.
Last week, I led a major project analyzing idle inventory across all 21 locations. The essential intelligence work was customizing the reports, identifying what data I actually needed and setting thresholds to flag. But the rest of it? Formatting the output, merging files and building the workbook? If I spend half a day doing that manually when AI can do it perfectly in minutes, I’ve wasted company money. Simple as that.
That realization unlocked something: I started using Claude Code to automate the tedious handoffs between my systems. I gave it access to my browser, pointed it at the iSolved portal and the QuickBooks integration and told it to download the payroll files, rename them to our naming convention, move them into the correct project folder and upload them. Then I went and did something actually useful with my time. The task runs automatically. I don’t think about it.
That’s not about outsourcing your work to a computer. It’s about removing the friction between the work that matters and the results it delivers.
How AI Is Streamlining Retail Accounting
For single-location owners and small adult chains alike, the AI accounting transformation is unfolding in a number of very practical areas. Implementing these fixes does not demand a finance degree or a large enterprise budget. Here are some examples.
- Receipt and invoice capture. Tools like Dext, QuickBooks’ built-in receipt capture and Xero now use optical character recognition to read vendor invoices — including amounts, dates and vendor names — and automatically assign them to the correct accounts. Over time, the system learns to recognize recurring vendors. Instead of manually entering invoices, you can photograph them or forward an email, and the system takes care of the rest. For a multi-location retailer managing vendor invoices across different corporate entities, this can save hours each week.
- Bank reconciliation that actually learns. AI-driven bank reconciliation tools suggest transaction matches and learn from your corrections. Every time you override a suggestion, the system becomes smarter. Over time, it stops asking about the vendor you pay regularly and instead focuses on transactions that truly need a human review. QuickBooks, Xero and Zoho Books all have versions of this feature built in now.
- The exception-only workflow. This is the major shift, and it’s where I see the most potential for retailers who reconcile everything manually. AI lets you move from checking every transaction to reviewing only the ones the system can’t explain. Think of it as an AI “auditor” quietly working in the background. For Lightspeed and QuickBooks users, this acts as a layer that compares your POS daily totals with what actually posts — highlighting discrepancies above a set threshold, rather than requiring you to verify each line. Tax mismatches, unusual voids, and register totals that don’t add up are flagged. Otherwise, you don’t need to intervene. Tools like Webgility already serve this purpose for retailers by creating reconciliation dashboards between POS and accounting platforms, rather than relying on silent one-way syncs that you have to verify afterward.
- Cash flow forecasting. QuickBooks, Xero Analytics and newer AI-native platforms like Digits now produce short- and medium-term cash flow forecasts directly within your accounting dashboard. Instead of exporting data and building models in Excel, you get a continuous view of your projected cash position, highlighted based on your actual revenue patterns and upcoming payables. For a small retailer managing tight margins — or a chain coordinating cash across multiple locations — that type of forward visibility used to require a controller. Now, it’s built into software you’re probably already paying for.
- Anomaly detection and fraud flags. AI not only classifies transactions but also detects anomalies like duplicate payments, unusual refunds, or spending that conflicts with policies or seems out of character for a vendor. These internal control checks have traditionally required a dedicated accounting team or expensive software. Zoho Books, QuickBooks’ intelligent features and stand-alone anomaly detection tools now provide this at prices accessible to small and midsize businesses.
Your choices for taking advantage of these options can be divided into three levels, based on how much you wish to invest and how involved you want to be.
Level 1: DIY with General AI Tools
If you’re using Lightspeed and QuickBooks and aren’t ready to add another subscription, you can achieve a lot with tools you already have. Upload your sales reports, reconciliation exports or payroll summaries to an AI assistant like Claude or ChatGPT and ask it to analyze variances, identify anomalies or format outputs for review.
The trade-off is manual effort — you’re still managing the download and upload process, just with a smarter tool on the receiving end. This is where I handle many tasks right now. It’s not perfectly seamless, but it’s much faster than doing the analysis manually.
For retailers with only some technical comfort, tools like Claude Code or Cowork, designed for less technical users, can automate repetitive tasks such as system handoffs, file renaming according to conventions, and data transfers between platforms, all without needing to touch a keyboard. This is free aside from your subscription costs, and the only limit is how well you can describe what you need. If you’re really ambitious, you can even start building your own software.
Level 2: Mid-Market Accounting and Integration Tools
For retailers seeking a more organized approach, several platforms are worth considering.
- Webgility connects Lightspeed to QuickBooks with reconciliation dashboards, exception reporting and sync analytics, so you see what posted, what didn’t and why. Starting around $59/month, it’s purpose-built for the exact multi-store, multi-account headache many of us are managing.
- Zoho Books starts at $15/month and offers features like smart transaction matching, anomaly detection and vendor insights out of the box. If you’re not already committed to QuickBooks, it’s worth considering.
- Digits is an AI-native accounting platform that offers real-time financials and around-the-clock automated bookkeeping. The free tier is usable; paid plans start at $100/month. The main selling point is that your books are always up to date, not just updated monthly.
- Botkeeper and Truewind use a hybrid approach: AI for categorization and reconciliation, with a human-review layer for edge cases. Pricing is customized, but they target larger SMBs. These sit on top of your existing accounting software.
Level 3: Integrated HCM and Enterprise Accounting Stacks
If you’re already using a platform like iSolved for payroll, review its features before adding more subscriptions. iSolved now includes AI-powered job ad authoring, candidate matching and payroll processing features that connect to your accounting stack — some of which you may already be paying for but not utilizing.
Honest Reality Check
I want to remind you of something that doesn’t always make it into AI articles: These tools are only as good as your data and your workflows.
If your chart of accounts is disorganized, AI categorization will still be disorganized — just more quickly. If your Lightspeed-to-QuickBooks mapping isn’t configured correctly, an AI auditor will flag variances it cannot explain because you haven’t clarified them to the system. Just as in inventory management, the adage “Garbage in, garbage out” applies here.
The retailers seeing real gains aren’t just adding AI to their existing workflows — they’re questioning why they do certain tasks before automating them. They start by standardizing their chart of accounts. They clean up vendor records. They document their reconciliation rules so the system can actually learn them. Bolting AI onto a broken process just makes it faster and still broken. First, redesign the process, then let AI handle the execution while you focus on judgment calls.
If you’re unsure where to start, identify the most painful recurring task in your accounting process — that weekly task you dread — and check to see if there’s a tool that can handle exactly that. There almost certainly is one, and it’s likely cheaper than the time you’re currently spending on it.
Zondre Watson is the general manager of technology and analytics for adult retail chain Ero-Tech. With a background in finance, chocolate and controlled chaos, he blends retail know-how with AI tools to keep 17,000 products moving smoothly.
