Every software company says their product is accurate. We would rather not be believed; we would rather be tested. So here is a named customer with her numbers, and two ways to put AllDone up against your firm's own work, on your own files, judged by your own reviewers. None of it requires you to take a single claim on trust.
Kim Lanighan runs Optics Accounting. What follows is her account, in her words, from one of her first runs on AllDone, with no training.
"One client took us thirty hours last year. Plenty are like that.
When a job takes too long I worry about it, because if a client has a big corporation tax bill they need to know as soon as possible. It has cost me sleepless nights, especially with five or six jobs piled up. I was working evenings and weekends.
There were points this summer when I thought I didn't want to do this anymore.
I was worried about trying AllDone, because I assumed it would take up more time. It's not until you've done two or three year ends that you get used to it. Then it's really clear what you need to do.
I took the finished AllDone accounts and corporation tax comp into Iris to file, and everything matched what AllDone was saying. Add-backs, capital allowances, the lot.
That thirty-hour client: I've spent two hours on it today, from loading to now, and it's basically done.
You can see it's saving huge hours, but the main thing is it has brought back control. I know exactly what's happening in these accounts within an hour. It's brought the excitement back too, the feeling that I can make this business work, and the passion for accounting and for helping my clients. And once it's up and running I can take time off, a complete shutdown to reset. It's been life-changing."
Kim Lanighan, Optics Accounting
Put ten jobs you've already signed off to the test. We'll find something. The engine works through each one the way your coldest reviewer would: the schedules, the reconciliations, the tax treatment, the disclosures, the judgement calls. What comes back is a review of each file, item by item, with anything that could have been done differently set out and referenced.
In beta, the Challenge found a fixed asset register that didn't agree to the accounts, and a director appointment missing from the file, both in jobs the preparers had rated perfect. Maybe your ten come back clean, in which case you have learned something valuable about your team. Maybe they don't, in which case you have learned something more valuable still. Either way, the evidence came from your work, not our marketing.
Nothing to prepare, nothing retained, no client names reach any AI model, and you sign a short data processing agreement before anything moves.
Book a call to set up your Challenge. Then choose a typical ten, in whatever state they naturally exist: no preparation, no cleansing, no cherry-picking. Include your best, we would rather be tested.
Pick a live job. One team member runs it through AllDone; another prepares it the way your firm works today. Same records, same deadline, blind to each other. When both are done, put the two files side by side in front of whoever reviews most fiercely, and compare four things: the working papers, the granularity of the schedules, the audit trail behind the decisions, and the hours on the clock.
Seriously. Choose the person most convinced this will get things wrong, the one who wants the lead schedules just so and asks why about everything, and put them in charge of the comparison. If the file satisfies them, you have your answer for the whole firm. If it does not, we want to hear exactly why, because that is how the product got this far: one of the firms in our beta was chosen for precisely that temperament.
If the comparison goes our way, you will know what you want to do next. If it goes yours, you will have ten cold-reviewed files and a well-tested job for your trouble, and we will have learned something. Both outcomes are fine with us.