Guides
The same job gets entered twice, once where the work is scheduled and once where the money is counted, and at month end the two sets of figures disagree by an amount nobody can explain. This is the most common operational complaint in small business and it is almost never solved by buying a third piece of software. It is solved by deciding which system owns what, and connecting them in one direction only.
This applies whether the other system is Jobber, Housecall Pro, Jane, Acuity, Buildium, Guesty or a spreadsheet. The shape of the problem does not change.
Four causes, and they need different fixes. Most owners assume it is the first one and it usually is not.
A customer is Dave Wilson in one system and D. Wilson Construction in the other. A job type is called one thing on the schedule and another on the invoice. Nothing is wrong with either record and no report will ever reconcile them. This is the most common cause and it is a naming problem, not a software problem.
The job software counts the work when it is done. QuickBooks counts it when the invoice is raised or when the money lands. Those are different dates, so the two systems are correct and different at the same time. Businesses waste weeks hunting a discrepancy that is a calendar rather than a fault.
Card payments through the job software, a payout from a booking platform, a deposit taken at the door. The payout that reaches the bank is net of fees, so it never matches the invoice total, and the fee is a real expense that has to be recorded somewhere or your margin is quietly overstated all year.
Materials bought on a card and never attached to a job. Labour recorded on the schedule but not costed. Work done and never invoiced. This is the expensive one, because it does not show up as a mismatch. It shows up as a business that is less profitable than the owner believes and cannot see why.
One system owns each piece of information, and it is the system where that information is created by the person doing the work.
The job or booking software owns the customer, the schedule, the job, the labour and materials on it, and the invoice that comes out of it. That is where a real person is standing when those things become true. QuickBooks owns the bank feed, the reconciliation, payroll, sales tax, and the reports your accountant works from. It is a ledger and it is very good at being a ledger.
Information then moves in one direction, from where it was created to where it is counted. Customers, invoices and payments go across. Nothing comes back. The moment you turn on a two-way sync, you have created a system that needs rules for what happens when the same invoice is edited in both places on the same afternoon, and nobody in a small business has time to be the referee for that.
Most job and booking platforms have an official QuickBooks connection. It is free or close to it, it is supported by both vendors, and it handles customers, invoices and payments, which is most of what you need. The limit: it syncs what its authors decided to sync. If your business depends on job costing, per crew profitability, or a field the platform does not map, it will not carry it and no setting will make it.
A third party that sits between the two and gives you control over the mapping. Right when the built in connection is close but misses something specific, or when you have more than two systems in the picture. The limit: another subscription, another vendor, and one more thing that breaks silently when either side changes an API. It needs an owner.
Genuinely the right answer for some businesses. A solo operator with thirty invoices a month may find that fifteen minutes on a Friday is cheaper and more reliable than any integration. The limit: it stops working the moment volume grows or someone else has to do it, because the process lives in one person's habits.
Which one is right is a question about volume, about how much of your margin depends on job costing, and about whether anyone will maintain a connector. It is not a question about which software is best.
Before any connection is switched on, four things need to agree across both systems, and getting them wrong is what makes integrations produce nonsense rather than time savings.
Merge the duplicates first. Decide which system is the master list. Every later report depends on this and nothing downstream can repair it.
Every service or job type on the schedule needs a matching item in QuickBooks pointing at the right income account. Unmapped items land in a catch-all and your revenue reporting becomes one large undifferentiated number.
Where does a deposit sit before the work is done, and where does the card processing fee get recorded. Decide once, write it down, and let it happen the same way every time.
A short monthly comparison of the two systems, with a written rule for what to do when they differ. Not a reconciliation project. Fifteen minutes, and the point is catching drift early rather than discovering it in January.
This is the part being oversold hardest right now, so the limits matter more than the capabilities.
Reading both systems, listing what does not match, and grouping it by likely cause so a person looks at four categories instead of two hundred rows. The limit: it reports and sorts. Deciding what a mismatch means, and fixing it, stays with a person.
Work marked complete on the schedule with no invoice against it after a set number of days. This is usually the first thing that pays for itself, because it finds money that was already earned. The limit: it can only see jobs that were recorded. Work that never made it onto the schedule is invisible to it and always will be.
Rules-based coding of routine transactions is fine and QuickBooks has done it for years. Letting something categorise unfamiliar expenses unsupervised is a different proposition, and it is your return that is wrong at the end of it.
Every business we have looked at that tried this ended up paying someone to unpick a year of it. The bookkeeper who knows your business is not the expensive part of your accounting.
In most businesses the double entry is not the first thing to fix. It is the symptom that makes the underlying problem visible, and the underlying problem is usually that the job software was set up quickly and never revisited.
That is what the free plan works out: which of the four causes is actually costing you, whether the fix is a connection or a naming exercise, and what it is worth in hours before you spend anything.
Almost always both, and the real question is the seam between them. QuickBooks is a ledger and it is very good at being one: bank feeds, reconciliation, payroll, sales tax, and the reports your accountant needs. It is not built to schedule crews or track a job. Trade software is built for the work and is a poor general ledger. Businesses that try to run everything in one of them end up doing the other half by hand, which is the situation that produced the double entry in the first place.
Usually one of four reasons, and they need different fixes. The same customer or service exists under two different names, so nothing reconciles. Or it is timing: the job software counts work when it is done, QuickBooks when it is invoiced or paid, so both are right and different. Or money is arriving net of card fees, so the payout never matches the invoice. Or one system knows something the other does not, such as materials never attached to a job. The last one is the expensive one, because it shows up as thinner margin rather than as a mismatch.
By hand is a legitimate answer at low volume and we will say so. Thirty invoices a month and fifteen minutes on a Friday can beat any integration, and it never breaks. It stops working when volume grows or when someone else has to do it, because the process lives in one person's habits rather than in a system. A connector earns its subscription when the manual step is either taking real time each week or being skipped.
Not necessarily, and it is worth resisting the pressure to decide quickly. Desktop still connects to plenty of systems, and moving is a real migration with a real cost, not a settings change. It is worth moving when something you actually need requires it, such as multiple people working in the file at once from different places. It is not worth moving because a connector's marketing page only mentions Online. Check whether your specific setup is supported before you accept that the move is forced.
