Expect to spend well beyond the licence in year one, once implementation, data migration, integration and training are counted. As a working rule I use two to three times the annual licence on top of the licence, so a $30,000 a year quote is a $90,000 to $120,000 first year. That multiplier is my own rule of thumb from doing this, not a published benchmark, and it should be tested against your own quote rather than trusted.
If you are holding a quote right now, the number on it is almost certainly the licence: what you pay the vendor each year for the right to use the software. Quotes will say licence or license depending on where the vendor is; it is the same line. It is the number the vendor leads with because it is the smallest one in the conversation.
Everything that turns that licence into a working system is quoted separately, quoted later, or not quoted at all.
What the benchmarks say, and who they describe
The most cited source is Panorama Consulting’s annual ERP Report. Its 2025 edition puts the median enterprise software project at $450,000.
Before that number is any use to you, read the sample. It is 172 respondents with a median annual revenue of $400.5 million, a median of 750 employees, and 55% of them multinational. That is not a business like yours, and the report contains no breakdown by company size at all. Anyone quoting it at a 40-person firm is quoting the wrong population.
Two findings from it do transfer, and the first one is the opposite of what most articles on this subject imply.
| Outcome | Share of projects |
|---|---|
| Came in under budget | 15.1% |
| Came in on budget | 53.5% |
| Slightly over budget | 23.8% |
| Significantly over budget | 7.6% |
More than two thirds of these projects came in on or under budget. Roughly 31% went over, and only 7.6% went over significantly. The median project also finished in 9 months, with more than three quarters completing inside their expected timeline. The disaster story is not the typical story.
The second finding is the useful one: of the projects that did go over, the most common reason was needing technology nobody had planned for, at 51.9%. Underestimated project staffing came second at 46.3%, then scope expansion, technical issues and data issues, all at 42.6%.
Source: Panorama Consulting 2025 ERP Report, pages 23 to 25, for the median project cost, the budget adherence split, the overrun causes and the timeline figures.
Read those two together and the picture is not “software projects always blow up”. It is narrower and more useful: most come in roughly where they were planned, and the ones that do not are usually undone by something that was never on the quote. Which is the whole subject of this article.
The costs that are real and rarely quoted
Here is where the multiplier goes. None of these are hidden exactly. They are simply not on the page you were handed.
Configuration and implementation
Someone has to set the system up to match how your business actually works: your pricing rules, your approval steps, your quirks. This is the largest single line and it is usually billed by a partner rather than the software vendor.
Data migration
Your existing data has to move. In practice this is where projects discover that the same customer exists four times under three spellings, that half the records have no owner, and that a field somebody repurposed in 2019 now means two different things. Cleaning that is work, and it is work nobody scoped.
Integration
The new system has to talk to accounting, to your website, to whatever runs scheduling or payroll. Each connection is a small project. A vendor calling an integration “native” is telling you a connection exists. It is not telling you what it costs to make it carry your data.
Training
Usually the smallest line on the quote and the one most likely to be cut when the budget tightens. It is also, in my experience, the one that determines whether any of the rest of it was worth spending. More on that below.
Your own people’s time
The line that never appears anywhere. Your operations lead will spend weeks on this. So will whoever knows how the old system really works. That time comes out of running the business, and if you do not budget for it you will simply absorb it as everything else running slower for a quarter.
The cost nobody itemises
The largest cost in a software project is frequently not on the quote at all. It is the months of half-use, partial data and workarounds that follow go-live.
I ran a CRM rollout across all 17 branches of a distribution business and under-budgeted the training badly. It took about six months to land properly, and that gap cost the business more than any single line on the invoice did. Why software rollouts fail, and the training budget I got wrong, is the full account, including what I would do differently.
For budgeting purposes the point is narrow. The cost of getting this wrong appears nowhere in the numbers above, and it is the one that decided my own project.
How to pressure-test a quote before you sign
Five questions worth asking
- What is the total first-year cost, not the licence? Ask for licence, implementation, migration, integration and training as separate lines. If a vendor will not itemise it, that itself is information.
- Who does the implementation, and is that price fixed? Software vendors frequently hand this to a partner. Find out who, what they charge, and whether the number can move.
- How many days of training, and is there a second visit? One pass at go-live is the default and it is not enough. Ask what happens in week six when people hit real problems.
- What state does our data need to be in? The honest answer is usually “much better than it is”. Ask who cleans it and whose hours that is.
- What does support actually look like after go-live? Response times, in writing. Support quality is invisible while you are buying and decisive the moment you have bought.
The question underneath the question
All of the above assumes the platform is the right answer. Often it is. Sometimes the quote on the desk is solving a problem that a process change would solve for nothing, and sometimes it is solving a real problem in the wrong order, ahead of something that is costing more.
That is a different question from “how much does this cost”, and it is worth answering first, because the cheapest implementation is the one you correctly decided not to do. If you want a straight read on which question you are actually facing, you can describe the situation in writing in about two minutes.
If there is a quote on your desk right now
Blue Harbour Navigator is a fixed-fee independent assessment for owner-led businesses. Seven to ten days, and you get a diagnosis of where the business is losing time and money, what technology would fix it, and what to do first.
I sell no software and I do not implement it, and I take no commissions from any vendor. $5,000 flat, in your own currency.
Book a 20-minute fit call Or tell me what you’re facing, no call needed
Two minutes to write, I read them myself. Or see how Navigator works first.
Sources. The Panorama Consulting 2025 ERP Report for the $450,000 median project cost, the budget adherence split, the overrun causes and the 9-month median timeline, all from pages 23 to 25. Note its sample: 172 organisations, median revenue $400.5 million, median 750 employees. The 2x to 3x multiplier is my own rule of thumb from running these projects, not a published benchmark, and it is labelled as such in the text. Vendor pricing changes; treat every figure here as a starting point for your own arithmetic rather than a quote.
Trevor Jamieson is the founder of Blue Harbour Solutions in Halifax, Nova Scotia. He has spent twenty years across the vendor and buyer sides of technology, at Dayforce, The Hercules Group, TELUS and Xerox. He currently holds a contract role in partner marketing at Amazon Web Services, and is co-founder and CEO of PhoneStack. Neither pays Blue Harbour anything inside a client assessment, and both are declared in writing on the site. Blue Harbour sells no software and takes no commissions from any vendor.