Almost nobody publishes what post-merger integration costs. Search the term and you get consulting firm capability pages, market-size reports, and an invitation to request a proposal.
That is not pure evasion. The number moves enormously with how much you intend to change. But “it depends” is useless when you are sitting in front of a board asking for an integration budget.
So this piece gives you the published benchmarks, explains where they stop applying, then builds a number from the bottom up with the arithmetic shown.
How much does post-merger integration cost on average? Post-merger integration costs most acquirers a low single-digit to low double-digit percentage of the acquired company's annual revenue, and the biggest driver is integration depth, not deal size. EY-Parthenon's analysis of 236 deals announced, closed or effective between January 2010 and December 2023, each valued at US$500m or more, found median integration costs by sector ranging from 3.5% of target revenue in energy and utilities up to 10.1% in health care and life sciences, with consumer at 7.5% and technology, media and telecoms at more than 5.6%. The same EY-Parthenon article puts overall M&A transaction costs, a separate figure measured against deal value rather than against revenue, at 1% to 4%. Those are large-deal numbers, so treat the percentages as a shape and build your own figure up from the work you plan to do.
Two things follow, and they matter more than the percentages.
The first is the denominator. Integration cost tracks the size and complexity of the business you bought far more closely than the price you paid for it, so target revenue is the more useful base than deal value.
The second is that the range is wide because integration depth is a choice. A light-touch bolt-on where you connect financial reporting and leave the operating systems alone lives at the bottom of that range. A full systems and process merge lives at the top.
The wider benchmark set sits in our post-merger integration statistics roundup. That page collects the numbers. This one shows the arithmetic that turns them into a budget you can defend.
Why don't published M&A integration cost benchmarks apply to your deal? Published M&A integration costs do not transfer to most deals because almost all of them are drawn from large-cap transactions. The one finding that does carry down is the one nobody wants to hear: smaller deals cost proportionally more, not less.
EY-Parthenon states it plainly. Deals valued at more than US$10 billion “often incur lower average integration costs as a percentage of the deal value compared with smaller deals”, and the reason is that a chunk of integration cost is fixed regardless of size. Their list of those fixed items includes regulatory filing requirements, IT-related fees, and management consulting fees.
Read that carefully, because the percentage does not transfer. The smallest deals in EY-Parthenon's dataset are still US$500m or more, so nothing published there describes a £4m bolt-on. What carries down is the mechanism: a fixed floor of work that a smaller revenue base has to absorb.
You can see why once you write out the work. A £4m bolt-on and a £400m acquisition both need a chart of accounts mapped , a payroll cutover planned, an email tenant migrated , systems access reviewed, and somebody senior to run the whole thing.
The £400m deal spreads that floor across a much larger revenue base. The £4m one does not.
BTD Consulting's analysis of typical M&A integration costs , from a London-based integration practice, puts the complexity half of that point in one line: “smaller deals do not equate to simpler integrations.” The task list is broadly the same. What changes is how many of them you decide to do.
This is the trap for acquisitive groups specifically. The instinct after a small deal is that a small deal deserves a small integration effort, and then the cost arrives anyway, spread across months of part-time attention from people who also have day jobs.
What are you actually paying for? An integration budget has four cost lines: internal time, external help, systems and licensing, and one-time integration work. Most first-pass budgets contain only the second of those four.
Most integration budgets I get sent are really just an external quote with nothing around it. The internal time, the software, and the one-off cleanup work turn up later as surprises, which is how a project quoted at one number ends up being defended at another in a board meeting.
Here is the full shape.
Cost line What sits in it What drives the number How to size it early Internal time Your finance lead, ops manager, IT contact and the acquired company's key people, pulled part-time onto integration work Number of entities, number of systems, how much of it lands on one person Count the named people, estimate days per month, cost them at a loaded day rate (salary plus employer costs). Usually the largest line and the one nobody books. External help Integration lead or programme management office, migration specialists, accountants, legal, change support Integration depth, how much internal capacity you genuinely have, whether you need judgment or just hands Get a scoped diagnostic quote first, before anyone prices a build. Systems and licensing New or expanded seats, tenant consolidation, ERP or CRM modules, reporting tooling, dual-running while both stacks are live Headcount added, how long you run parallel systems, how many platforms overlap Price 12 months of dual-running as the default, then argue it down. One-time integration work Data cleanup and migration, chart of accounts mapping, rebranding, retention or severance, decommissioning old systems Data quality in the acquired business, people decisions, how many legacy systems you actually switch off Run a data quality check before committing to a timeline. This is where overruns start.
That last line deserves a flag. In EY-Parthenon's dataset the most frequent cost drivers were severance and retention costs, followed by real estate and IT, and buyer-paid severance on its own accounted for more than half of integration costs in certain deals.
People decisions, in other words, are frequently the biggest number in an integration budget, which makes change and retention planning a budget question rather than a soft one. If your deal has no headcount changes, your costs should sit well below the published medians. If it has several, expect the opposite.
What does a mid-market integration budget look like? A worked example A light-touch integration of a £4m-revenue target works out at roughly £43,400, or about 1.1% of target revenue, on the illustrative assumptions set out below. Every input below is an assumption I have chosen so the arithmetic is visible, so replace all of them with your own numbers before you use the total.
The scenario: a £4m-revenue bolt-on, integrated light-touch. Connect financial reporting into the group, harmonise the chart of accounts, leave the operating systems alone. Assume a six-month window.
Internal time. Your finance lead at 4 days a month and an ops manager at 2 days a month, across six months, is 36 days. At a loaded day rate of £400, meaning salary plus employer costs rather than headline salary, that is £14,400.
External help. A scoped 20-day engagement covering the reporting integration and the mapping work, at £750 a day, is £15,000.
Systems and licensing. Additional seats plus the group reporting layer at £500 a month, which is £6,000. The project window is six months, but licensing is a run-rate cost rather than a project cost, so the example carries a full twelve months of it. There is no dual-running line at all, because this scenario leaves the operating systems in place, and that absence is a large part of why the total lands where it does.
One-time work. Ten days of specialist time on data cleanup and chart of accounts mapping at £600 a day, so £6,000, plus £2,000 to retire the standalone spreadsheets the target used to report from. Note what is not in this line: no severance, no retention payments, no property.
On those illustrative inputs the total is roughly £43,400, or about 1.1% of the target's revenue. Notice it lands well below EY-Parthenon's 3.5% to 10.1% sector medians, and that gap is doing exactly what it should: those medians come from large deals integrated deeply, and this example buys a deliberately shallow integration.
Push the same business to a full systems merge and the shape changes completely. The internal time roughly doubles, the one-time work multiplies as you migrate operational data rather than just financial data, and you add ERP or CRM licensing you did not previously carry. A CRM migration after an acquisition is the line most people forget to price at this stage.
So the budget is downstream of a decision you control. Make that decision explicitly, and in writing, before anyone starts costing it, because the depth is much harder to argue down once a number is attached to it.
How do post-merger integration consultants charge? Post-merger integration consultants charge four ways. Day rate is the default, but you will also be offered fixed-scope project fees, monthly retainers, and, at the top end, a percentage of deal value or of synergies. Which one you are offered tells you a lot about what you are buying.
Day rate. The default for strategy firms and independent interims. Consultancy.uk's published UK consulting fee rates run from around £50 per hour for an interim consultant working at an operational level up to £300 or more per hour for a consultant from a leading strategy consulting firm, a spread of roughly six times for work that can look similar on a slide.
Fixed-scope project fee. A defined piece of work for a defined price: migrate this system, build this reporting pack, run this first 90 days. Better for the buyer, because the estimating risk sits with the supplier, but it only works once somebody has scoped the thing.
Monthly retainer. Ongoing capacity, usually with a minimum commitment. Sensible for serial acquirers who will have another deal in six months and want the same team to keep the playbook warm.
Percentage of deal value or of synergies. Common at the top end. Be careful with synergy-linked fees on a small deal, because you will spend more time arguing about attribution than the fee is worth.
Translate the day rate into an engagement before you compare quotes. Two days a week across a three-month integration is about 26 days, or roughly 195 hours at a 7.5 hour day. At the £50 per hour operational end that is just under £10,000; at the £300 per hour strategy end the same calendar is £58,500.
The tier matters as much as the model. A large firm brings brand assurance, a bench, and multi-geography reach, and charges for all three. An independent or boutique brings a named person who does the work, which in our experience is usually the better trade on smaller mid-market deals, provided you check they will still be there in month four.
Whichever model you pick, ask who does the work. If the person in the room is not the person who will run the migration, you are paying for supervision.
This piece prices the work; whether you should buy it at all is a separate question, and we answer it in when to hire a PMI consultant . If the tier question is what you are stuck on, start with who actually does integration work .
What does it cost to not integrate? Not integrating usually costs more than integrating does, eventually, and it arrives as lost margin rather than as an invoice, which is why it rarely gets counted.
The costs of leaving acquisitions unintegrated are real but diffuse: duplicate software and vendor contracts , finance teams rebuilding the same consolidated view by hand every month, decisions delayed because nobody trusts the group number, and legacy systems left running because decommissioning them was never scoped. We have written up where the recoverable money sits in our guide to cost savings after an acquisition .
BTD Consulting's worked example is a useful way to frame the ask. On their numbers, measured against enterprise value rather than the target-revenue base used earlier in this piece, a relatively high integration spend of 5%, aimed at a 10% increase in acquired EBITDA from around 12 months post-close, returns a simple ROI of 27%.
You do not have to accept those specific inputs to use the logic. Put the integration spend next to the value it is protecting, not next to last year's overheads budget, and the conversation changes.
The version that costs the most is the middle: paying for a partial integration that never finishes, so you carry both the project cost and the duplicate running costs. That happens more often than a clean decision to leave a business standalone.
How do you budget for post-merger integration and defend the number? Build an integration budget you can defend by deciding the integration depth first, then costing the four lines (internal time, external help, systems and licensing, one-time work) against it, then adding a contingency band that reflects how much you know about the acquired company's data.
The sequence matters because depth is the variable that moves everything else. Deciding to fully merge two ERPs and then discovering the budget is a different exercise from deciding what you need first. We set out the three levels in choosing your integration level .
In practice it looks like this.
Set the depth deal by deal. A light-touch acquisition where you connect reporting and leave operations alone is a legitimate answer, not a failure of ambition. Most acquisitive groups run a mix.
Inventory before you estimate. Systems, entities, headcount, data quality, contract renewal dates. An integration audit checklist gets most of it, and a week spent here removes most of the estimating risk.
Cost all four lines, including internal time. If the internal number embarrasses you, that is information. It usually means the work is going to land on people who cannot absorb it.
Let data quality set the contingency band. A clean, single-entity target with one accounting system justifies a tight band. Four entities on three systems with no documentation does not, and data migration after an acquisition is where that uncertainty converts into cost.
Budget the run rate as well as the project. Consolidated reporting, licences and support continue after the project ends. A build number with no monthly number attached is an incomplete budget.
For groups doing this repeatedly, the third or fourth deal should be materially cheaper than the first, because the mapping decisions and the migration path are already made. That compounding is the argument for treating integration as a capability rather than a series of one-off projects, and it is the theme of our serial acquisition integration playbook .
How does PMI Stack price integration work? PMI Stack prices integration work in three stages. A fixed-price paid diagnostic comes first, then a fixed-price execution phase quoted against the roadmap that diagnostic produces, then included hypercare with an optional monthly retainer.
PMI Stack is an integration practice that works with acquisitive groups on the systems, data and consolidated reporting side of a deal. The staging exists so that nobody quotes a build before someone has opened the systems.
The diagnostic is a scoped audit of the systems, entities, data and reporting across the group, ending in an integration level recommendation and a roadmap. It is fixed price, sized from the intake scope. For a mid-market group we price that diagnostic in the low four figures, and it is credited against the build setup if you proceed within 60 days.
The execution phase is then quoted as a fixed price against that roadmap, with the estimating risk on us. We would rather absorb an estimating mistake than hand a client an open-ended hourly bill.
After go-live there is a short hypercare period included, meaning a week or two of hands-on support while the new setup beds in, plus an optional monthly retainer for groups that want capacity kept warm for the next deal. If consolidated reporting is the pressing problem rather than a full systems merge, that is a narrower and cheaper starting point, described in consolidated financial reporting .
Two honest limits. We do not do pure strategy consulting, and if what you need is a pair of hands, someone else will be cheaper.
Common questions about post-merger integration cost Is integration cost a percentage of deal value or of revenue? Both get used, but target revenue is the more reliable base for smaller deals. Deal-value percentages get quoted because deal value is the number everyone already has in front of them. It is still the wrong base: a bargain purchase does not arrive with a smaller chart of accounts.
Why do small acquisitions cost proportionally more to integrate? Small acquisitions cost proportionally more because the floor of integration work does not shrink with the target, so the same tasks land on a much smaller revenue base. The practical consequence for acquisitive groups is that buying four small businesses costs more to integrate than buying one business of their combined size.
Who pays post-merger integration costs, the buyer or the seller? The buyer pays post-merger integration costs in almost all cases, because integration runs after close and it is the acquirer's programme. Transition services the seller provides are the usual exception, and those get priced in a transition services agreement negotiated alongside the sale rather than in your integration budget, so keep the two numbers apart.
How do you reduce post-merger integration costs? You reduce post-merger integration costs by reducing the depth, not the day rate. Deciding up front that financial reporting is the only thing you will merge removes whole workstreams, which is worth far more than shaving a supplier's quote. What does not work is squeezing the day rate on a scope nobody has defined yet.
Are integration costs capitalised or expensed? Integration costs are mostly expensed rather than capitalised, but check each line rather than the project as a whole. Deal costs are the clear-cut part: apart from the costs of issuing debt or equity, which follow their own rules, IFRS 3 requires an entity to account for acquisition-related costs “as expenses in the periods in which the costs are incurred and the services are received”, rather than adding them to the cost of the business combination (IFRS 3 agenda decision, July 2009 ).
Post-close integration spend sits outside that rule and gets treated line by line. Most of it lands in the profit and loss account: internal time, data cleanup, severance, decommissioning. Some system build work can be capitalised as software instead.
Confirm the split with your auditor before you present the budget, because it changes how the deal year reads to a lender.
How long do post-merger integration costs last? In the integrations we run, most of the one-time cost lands in the first six to twelve months, with a tail for decommissioning and any system consolidation you deferred. The run-rate costs, licences, reporting and support, continue indefinitely. We cover project timing in how long post-merger integration takes .
Should the integration budget come out of the deal model? Yes, and it should be in the model before you sign. Integration cost that appears post-close reads as an overrun even when the number was always reasonable.
Can you run a post-merger integration without external help? You can run an integration without external help, but only if you have genuine internal capacity and have done it before. The failure mode is rarely incompetence. It is that nobody is given the time: the work goes to whoever is closest to it, alongside a full workload, and then stretches across quarters instead of weeks. The first 100 days after an acquisition is where that usually becomes visible.
What is the biggest driver of integration cost? People are the single biggest integration cost driver, by a distance, as the EY-Parthenon severance and retention finding earlier in this piece shows. So the honest first question in any budget conversation is whether the acquired management team is staying, and whether anyone is being paid to leave. That answer moves the number more than any system decision you make afterwards, and it is usually settled before anyone opens a spreadsheet.
So what number should you put in front of the board? The number you built from the work, not one lifted from a benchmark. Size the four cost lines against the depth you have actually chosen, then sanity-check the total against the published sector medians and be ready to explain any large gap in either direction.
Then run one check before you present it. If the internal time line is smaller than the external quote, the budget is probably wrong, because the work still has to be done by somebody and that somebody is on your payroll. It is the line I see underestimated most often.
There is no honest headline post-merger integration cost, and anyone who offers you one has quietly picked your integration depth for you. Build it yourself and it stops being a guess. Boards rarely reject a defended integration budget; they reject the one that arrives in pieces, three months after the deal closes.
If you are pricing a live deal and want a second opinion on the scope before you commit to a figure, our system unification page sets out what the diagnostic covers.
.png)
