Most contractors do not leave QuickBooks because it failed. They leave because the business changed shape around it: more jobs, longer jobs, prevailing wage work, a second entity, a bonding company that wants a WIP schedule ten days after close instead of thirty.
This is written for general contractors, specialty subcontractors, design-build firms, and developers who are running on QuickBooks now and trying to work out whether what they are feeling is a software problem or a process problem.
It is a diagnostic, not a pitch. Three of the seven signs below and you are probably fine. Five or six, and the spreadsheets have already become the system of record, whether or not anyone has said so out loud. There is also a section on when you have not outgrown QuickBooks, because that is the honest answer more often than most software vendors will tell you.
How to tell if you’ve outgrown QuickBooks
The short version. A construction company has outgrown QuickBooks when job costing, WIP reporting, and compliance reporting have moved out of the accounting system and into spreadsheets. The seven signs are:
- Job costing happens in spreadsheets rather than in QuickBooks, so the job cost report and the general ledger never tie without manual work.
- The WIP schedule is rebuilt by hand every month-end and takes days rather than hours to produce.
- Retainage is tracked manually on both the receivable and the payable side, because QuickBooks has no retainage sub-ledger.
- Certified payroll reports take days to assemble because QuickBooks does not produce WH-347 output natively.
- Each legal entity sits in a separate QuickBooks company file, so consolidation and intercompany allocations are a spreadsheet exercise.
- Nobody in the business trusts the cost-to-complete number, so estimates at completion are argued rather than reported.
- The project management system and the accounting system report different costs for the same job.
What QuickBooks actually does well for contractors
Is QuickBooks worth it for a construction company? Yes — up to a point, and that point is further out than most vendors admit. For a contractor running fifteen or twenty jobs a year with one entity, QuickBooks handles general ledger, AP, AR, bank reconciliation, and sales tax as well as anything at its price.
Desktop Premier Contractor edition adds customer: job tracking, estimates versus actuals, job profitability, and a committed costs by job report if your purchase orders are clean. QuickBooks Online has Projects for income and cost by project. Class tracking separates a service division from new construction, and third-party apps fill the AIA billing and certified payroll gaps.
The limitations of QuickBooks for construction are not defects. They are the boundary of what a general ledger package can carry before the structure moves into Excel — and the people move with it.
Where QuickBooks stops being native. The table below is what a contractor actually needs versus what QuickBooks does without help.
| Task | QuickBooks (native) | What it takes at scale |
|---|---|---|
| General ledger, AP, AR, bank rec | Full native support | Nothing more needed |
| Job-level cost tracking | Customer: job and Projects | Cost code structure with budget revisions |
| Cost code budgeting and committed cost | Not native — POs are non-posting | Commitments carried into job cost reporting |
| WIP schedule | Not native | Contract value, cost to date, cost at completion, and billings as of one close date |
| Percentage-of-completion revenue | Not native | Automated earned revenue and over/under billing |
| Retainage receivable and payable | No retainage field or sub-ledger | Native retention tracking with release status |
| AIA G702 / G703 billing | Not native — add-on required | Pay application generation from the schedule of values |
| Certified payroll (WH-347) | Not native — add-on required | Per-job, per-week output by work classification |
| Multi-entity consolidation | Separate company files only | One database, intercompany posting |
| Labor burden | Blended, applied at period end | Burden on each hour as it hits the job |
Sign 1: Job costing lives in spreadsheets, not in QuickBooks
If your real job cost report is a spreadsheet rebuilt each month from QuickBooks exports, you have outgrown QuickBooks job costing. QuickBooks tracks cost against a customer:job or a project, but it does not hold budget, committed cost, cost to date and units in place against a cost code structure. That structure ends up in Excel, and Excel becomes the system of record.
Job costing is the practice of tracking every labor hour, material purchase, subcontract and equipment charge against a specific cost code on a specific job, so that profitability is visible per job rather than only across the company.
The tell is the lag. A PM asks how much is left in 26 05 00 and the answer is that the file is current through the 31st. A change order approved in the field three weeks ago is not in the budget column because nobody retyped it. Labor burden goes on as a blended percentage at month-end rather than onto each hour as it hits the job, so every mid-month margin is wrong by the same unknown amount.
“Because our systems didn’t talk to each other, intercompany data was unbelievably difficult to maintain. Missing an entry in one company became a humongous headache for us.” – Joel Sisto, CFO, IOC Construction, Read the Full Customer Story
Sign 2: WIP reporting is a month-end fire drill
A WIP schedule that takes several days to produce means the accounting system is not holding what percentage-of-completion accounting needs. That is contract value including approved change orders, cost to date, estimated cost at completion, and billings to date, as of the same close date, for every open job. QuickBooks holds two or three of those reliably. The rest arrive by email.
A WIP schedule is the report that reconciles what each job has earned against what it has been billed, using percentage-of-completion to recognise revenue in proportion to cost incurred against total estimated cost.
So the controller assembles it in Excel, computes earned revenue, falls out over-billing and under-billing, and posts the adjusting entries afterward. The schedule is correct, and only one person can produce it. You also cannot look at WIP mid-month, so a fade shows up in the last week of the quarter.
The cost of this is not only internal. The same reconstruction happens every time a surety or a lender asks for a financial package: export from QuickBooks, reconcile against the spreadsheets, rebuild the WIP and backlog schedules, and get it all consistent before the deadline. A period restated for a stale cost-at-completion estimate becomes a conversation with your surety, and bonding capacity is priced partly on how much confidence they have in numbers you produce. If “which jobs are in trouble?” and “what is our backlog margin?” take days to answer rather than minutes, that is the reporting cost of the same underlying gap.
Sign 3: Retainage is tracked by hand
If retainage lives on a spreadsheet of open retention by job, you are carrying a receivable and a payable your accounting system cannot see properly. QuickBooks has no native retainage field on a customer invoice or a subcontractor bill. The workaround is a negative line on the pay application plus an other current asset account, with releases tracked manually.
Retainage is the percentage of each progress payment — commonly 5% or 10% — that an owner withholds until the job is substantially complete, and that a general contractor in turn withholds from its subcontractors.
That works on four jobs. On thirty jobs with two years of open retention, it produces the same two problems every time. Retention released on a job that closed in March is not invoiced until somebody notices in July. And retainage payable to subs is not aged, so you cannot see who is owed release and who is still missing a final lien waiver.
Sign 4: Certified payroll takes days instead of minutes
Certified payroll that occupies two people for two days every week means the payroll system is doing prevailing wage work it was not designed for. QuickBooks records the wages. Assembling the reports is where the days go.
Certified payroll is the weekly report required on federally funded and many state-funded projects, showing each worker’s classification, hours, gross wages, fringe benefits and deductions, submitted on Form WH-347 or a state equivalent.
A laborer who spends Tuesday on a federally funded school at one classification and Wednesday on a private fit-out at another has to be split correctly. Fringe paid in cash reports differently from fringe paid into a plan. Multi-state contractors deal with several formats and agency portals, and the penalty for getting it wrong is not a rejected form — it is a withheld payment.
Sign 5: Each entity is a separate QuickBooks file
Running each entity in its own QuickBooks company file means every consolidated number in your business is produced outside your accounting system. Contractors accumulate entities for good reasons: a separate LLC for a state license, an equipment company, a holding entity, a joint venture. QuickBooks does not consolidate across company files or post intercompany transactions automatically.
So the equipment company bills hourly rates to jobs through invoices keyed manually on both sides, and the consolidation workbook has a mapping tab one person understands. Class tracking is the usual substitute, but it cannot produce a clean per-entity balance sheet without discipline nobody sustains, and QuickBooks Online plans cap combined classes and locations.
This is also where the limitations of QuickBooks Online bite hardest, because the desktop workaround of simply opening a second file is clumsier in the browser.
Sign 6: Nobody trusts the cost-to-complete number
When the CFO and the project managers walk into the same meeting with different cost-to-complete numbers, the problem is structural, not personal. A defensible cost-to-complete needs committed cost — issued purchase orders and executed subcontracts not yet invoiced — sitting alongside actual cost and budget. If committed cost is not in the accounting system, finance works from invoices received and the field works from a number it cannot support.
Cost-to-complete is the estimate of remaining cost required to finish a job; added to cost incurred it produces the estimate at completion, which is what drives percentage-of-completion revenue.
Purchase orders in QuickBooks are non-posting documents. A $180,000 steel PO and a $420,000 subcontract with $150,000 billed do not appear as exposure anywhere in the financials, so a job looks healthy until the last three subs invoice in the same month.
A diagnostic you can run this week: track how far cost at completion moves in the final 20% of a job. If it always moves the same direction, you are discovering rather than forecasting.
Sign 7: Your PM software and your accounting system disagree
If your project management system and your accounting system disagree about what a job has cost, one of them is being re-keyed by a person. Procore, Autodesk Build, and similar platforms own the field record: change orders, subcontracts, daily logs, subcontractor compliance and lien waiver status. When the link to accounting is a CSV export, the two diverge within a week of every close.
Cost codes exist in two structures with different names, so a mapping table sits between them and nobody owns it. A change order is executed in the PM system while the budget in accounting still shows the original contract value. Finding construction software that integrates with QuickBooks is not usually the hard part — the hard part is that an integration synchronises records without resolving which system owns the budget.
What resolves it is one cost code structure, one system of record for budget and commitments, and an integration that writes both ways.
When you have not outgrown QuickBooks
Most content on this subject exists to tell you that you have. Often you have not, and switching would be an expensive way to avoid fixing something cheaper.
Stay where you are if: you run one entity, your job count is stable, your payroll is standard rather than prevailing wage, you bill time-and-materials or simple progress rather than AIA pay applications, and your month-end close takes days rather than weeks. In that situation, the return on an ERP project is poor and the same money spent on a rebuilt chart of accounts, a disciplined cost code list and a competent construction bookkeeper will do more.
Two further cases where the answer is “not yet”. If exactly one of the seven signs above applies, fix that one thing — a certified payroll add-on costs a fraction of an ERP. And if your problem is entity structure rather than job cost complexity, a mid-tier accounting platform may close the gap without a full construction ERP.
The threshold is cumulative, not any single sign. Three or four at once, with a growing backlog and a controller who is the only person able to produce the WIP schedule, is where the arithmetic changes.
What moving off QuickBooks actually involves
Moving off QuickBooks is a finance project with a software component, not the reverse. Your controller is the critical path, not the consultant.
Open jobs migrate with full cost history by cost code, because you need comparative cost at completion. Closed jobs generally do not. You also bring the trial balance, open AR and AP detail, open commitments and retention balances in both directions.
The sequence that works
- Write down what actually hurts. Active job count, monthly change order volume, days to close, hours spent rebuilding WIP, whether any field cost data reaches the office the same week it happens. Get accounting, PMs and at least one superintendent in the room. A requirements list written only by finance produces a system the field will not use.
- Clean the QuickBooks data first. Standardise the cost code list, job naming, vendor records and chart of accounts before anyone demos anything. This work pays for itself whichever system you choose, and it is the single best predictor of whether the migration goes well.
- Shortlist construction-specific systems. Acumatica Construction Edition, Sage Intacct Construction, Viewpoint Vista, Foundation, CMiC and Jonas Premier are the platforms contractors most often evaluate. A general-purpose ERP without a construction edition will cost more in configuration than it saves in licence.
- Demo with your own jobs. Take a real budget, a real change order, a real subcontract, and a real pay application into every demo, and ask the vendor to produce the WIP schedule you currently build by hand. Generic demo data tells you nothing about fit.
- Phase the rollout. Core accounting and job costing first. Field apps, subcontractor compliance, and equipment second. Trying to go live on everything at once is the most common reason construction implementations slip.
- Name internal owners before go-live. One in accounting, one in operations. They answer the day-to-day questions that otherwise go to the consultant at consultant rates, and their visible buy-in is what determines whether PMs adopt or work around.
- Run both systems through one or two month-end closes. Do not cut over cold. Closing twice in parallel is where configuration errors surface while you still have a working fallback, and it is the cheapest insurance in the whole project.
Rebuilding the cost code structure
This is the real work, and it is not technical. It is getting your PMs and your controller to agree on one cost code list, one set of cost types, and one rule for what gets coded where, before anything is configured. Skip it, and you rebuild the same spreadsheets on a more expensive system.
The Desktop question is now on a clock
New US subscribers have been unable to buy QuickBooks Desktop Plus subscriptions since 30 September 2024, though existing subscribers can still renew, and the 2023 Desktop products — Pro Plus, Premier Plus and Mac Plus — are discontinued after 31 May 2026. Enterprise is the exception and is still sold.
QuickBooks is not going away, and anyone telling you otherwise is overselling. But if you are on Desktop Premier Contractor, the platform decision you assumed you could defer for two more budget cycles is being scheduled for you.
What construction ERP costs
Construction ERP costs more than QuickBooks in license and considerably more in implementation. The honest comparison includes what your present setup already costs: controller hours on month-end WIP and weekly certified payroll at fully burdened rates, the working capital cost of pay applications that go out late, and retention invoiced months after release.
There are three components to the new number. Subscription licensing — for Acumatica this is based on the resources and transaction volume you consume rather than named users, so adding PMs, field staff and payroll clerks does not raise it. Implementation services, which are usually the larger first-year number. And your own team’s time, the line most often left out of the business case.
For a company coming out of QuickBooks, implementation usually runs between $40,000 and $250,000, and takes anywhere from 12 weeks to a year.
Four things move a project inside those ranges: the complexity of the business, how much of your QuickBooks history you want migrated, the size of the company, and how much of your own staff’s time you can dedicate to the work. The last is the one you control, and it is largely what separates a twelve-week project from a twelve-month one.
The ranges are wide because scope genuinely varies, not because the numbers are soft. A single-entity contractor moving job cost, progress billing and payroll is a different project from a multi-entity group consolidating several companies, migrating years of job history and running certified payroll across several states.
The implementation figure is also not fixed. Where budget rather than fit is the obstacle, there are usually ways to bring it down, and that is a conversation worth having before a quote lands rather than after.
The stay-put option you will be offered
Intuit Enterprise Suite is Intuit’s higher tier for multi-entity businesses that have outgrown QuickBooks Online, and for some contractors it may well be the right answer — particularly if your complexity is entity structure rather than job cost.
Ask five questions before deciding:
- Does it job cost to cost code level across labor, material, subcontract, equipment and other cost types?
- Can it produce AIA G702 and G703 pay applications?
- Does it handle retainage receivable and payable as native fields with release tracking?
- Does it produce certified payroll for Davis-Bacon and state prevailing wage work?
- Does it carry committed cost from purchase orders and subcontracts into job cost reporting?
If the answers are yes for your mix of work, that is a shorter and cheaper project than an ERP. If four of the five are “with a partner app,” you are assembling an ERP out of parts and should price it that way.
The four objections, answered honestly
“We can’t change systems with jobs in the air.” You are not changing engines mid-flight; you are running two engines for a quarter. Pilot on new jobs that carry no legacy baggage, keep QuickBooks as the read-only record for closed work, and close in parallel twice before cutting over. Contractors who get burned on this are almost always the ones who tried to go live on 1 January with no parallel period.
“The price is hard to justify.” It is, against a QuickBooks subscription. It is not against what the current setup actually costs. Put a number on the controller hours spent on WIP and certified payroll at a fully burdened rate, the working capital sitting in pay applications that went out a week late, the retention nobody invoiced for four months, and the margin on the last two jobs where the overrun surfaced too late to price a change order. That is the comparison. If it still does not clear, you have your answer and it is a legitimate one.
“Our data is a mess.” Then the ERP will be a mess, which is why step 2 above exists. Dirty cost codes migrated faithfully produce unreliable reports from day one and destroy trust in the system before anyone has learned it. Budget real time for data hygiene as its own workstream, not as something the implementation absorbs.
“The field will not use it.” Sometimes true, and it is a change management failure rather than a software one. Field staff need different training from controllers, delivered in the format they actually use, and they need to see something in it for them — a superintendent who gets same-day labour cost against budget has a reason to submit time correctly. If the only beneficiary of the new system is head office, adoption will be exactly as good as you would expect.
