Don't start from a percentage. The commonest shortcut in a rushed deal process is to guess a synergy number as "2% of combined revenue" or "10% of the target's cost base" and then work backward to a story. That runs the logic in reverse. Build the ledger bottom-up, function by function, the way the integration-planning office would, since it'll later be held accountable to the number.
One saving, one line, one owner. That's the rule, and every entry in the ledger names three things. (a) The specific action that produces the saving or the revenue: close this facility, renegotiate this vendor contract, cross-sell this product through that channel. (b) A single named owner accountable for delivering it. (c) The specific cost-center, headcount pool, or revenue line it draws from, and that pool must appear on no other line in the ledger.
This one habit is what makes double-counting detectable at all. A ledger built loosely, with overlapping descriptions and no named pool per line, can't be audited even in principle.
Advertisement
How do you size cost and revenue synergies?
Cost synergies, the credible family. You size these from the two entities' actual cost bases, never from a benchmark percentage. Procurement means combined purchasing volume against shared vendors, sized from actual spend data in both companies' AP ledgers. Overhead and duplicate public-company costs (redundant board, audit, listed-entity compliance, executive layers) are sized from the org chart and actual headcount. Don't estimate them. Footprint covers overlapping facilities, warehouses, or offices, sized from a facility-by-facility overlap map. Logistics covers shared routes and consolidated carrier contracts. Systems means retiring a duplicate ERP, CRM, or cloud footprint once migration completes. These lines are largely within management's control, and decisions deliver them. So diligence should demand the specific decision behind every line, along with the dollar figure.
Revenue synergies, the incredible family. Cross-sell (the combined entity's customers buy each other's products), bundling/upsell, and expanded bid capacity (a combined balance sheet or footprint wins tenders neither could win alone). Every one of these depends on customers, competitors, and sales cycles behaving as hoped. They arrive latest of all the synergy types. They're also the line item most often inflated to justify a premium a bidder has already decided to pay. So size every revenue-synergy line at a deliberately conservative fraction of its "headline" potential, and expect it to take the hardest questioning in the stress tests that follow.
Sizing method, stated on every line. A defensible ledger states how each line's number was derived. That might be a percentage of a named, verifiable cost base (never of combined revenue), or a headcount count cross-checked against an actual redundant-role census instead of an estimate. It might be a facility-by-facility overlap map, or a customer-account-level overlap analysis for cross-sell. A line with no stated sizing method is still a placeholder. Label it as one until it earns a real number.
Stress test one: double-counting
Double-counting is the most common and least discussed defect in a synergy ledger. From inside any single workstream it's invisible. It only shows up when someone reads the whole ledger across, line by line, against a single master list of named cost pools. It comes in two distinct shapes.
Shape 1, the same saving claimed on two lines. Two workstream leads, working in parallel with no shared master list, each claim credit for eliminating the same cost. You'll see it in a few classic patterns. An IT workstream and an overhead workstream both claim the salary of the same shared-services team. A facilities workstream and a logistics workstream both claim credit for the same warehouse closure. The third is harder to spot, because it involves the target's own "base-case" standalone forecast, which already assumed some scale-driven cost improvement before the deal was even announced. A procurement workstream then claims the same margin uplift, so the synergy case doubles an improvement that was already baked into the diligence team's base case.
The fix is mechanical. Reconcile every line against a single master cost-center and headcount register, and reject any pair of lines that reference the same pool.
Shape 2: interacting synergies that partially cancel. This one is subtler, and possibly more dangerous, because it produces two individually defensible-looking lines that are jointly wrong. Realizing one synergy reduces another. Take a facility-consolidation synergy that closes a nearby plant, next to a logistics synergy that assumed shorter average shipping routes. They're in direct tension. Closing the plant lengthens the average route, which erodes some or all of the logistics saving. A ledger that books both lines at their stand-alone, un-netted values overstates the truth. Don't strike either line. Model the combined effect of both decisions together, then re-derive the logistics line in the world where the facility decision has actually happened.
How do you check a synergy ledger for double-counting?
The reconciliation checklist. Run four checks before any ledger leaves the diligence room. (1) Every line traces to a single named cost-center, GL line, or headcount pool that appears on no other line. (2) The sum of every line's claimed headcount reduction is cross-checked against an independent, actual redundant-role census. If the ledger claims more heads than the census finds redundant, it's either double-counting or wishful thinking. (3) Every revenue-synergy line is checked against every cost-synergy line for a shared resource claimed by both. Watch for the sales-force trap here. You can't both keep a target's sales team intact to drive cross-sell and eliminate it as a cost synergy. (4) One-time cash/working-capital effects, such as a one-off inventory drawdown, are kept in a separate ledger from recurring P&L run-rate effects like the ongoing carrying-cost saving on lower average inventory. Fold both into one undifferentiated number and then annualize it, and you've double-counted the one-time item forever.
Stress test two: dis-synergies
A synergy ledger with no negative lines hasn't been stress-tested. It was drafted by people paid to want the deal to close. Five dis-synergy patterns recur across almost every deal.
Customer attrition / channel conflict. Some customers dual-source on purpose, to avoid depending on a single supplier. Once two suppliers merge into one, they're structurally likely to shift spend away. Distributors cut loose by a footprint-consolidation plan stop buying entirely. Size this from the diligence team's actual customer-overlap analysis: which accounts are common to both entities, and which of those have a stated dual-sourcing policy. Don't guess at it.
Key-talent flight. Specialist staff are flight risks: the people who hold the vendor relationships, the technical know-how, or the customer trust the synergy case depends on. They're especially likely to go once the retention or earn-out period that kept them in place lapses. In acquisitions of small, founder-led targets, departures cluster right at the vesting cliff. Where a specific revenue or capability line depends on named individuals staying, risk-weight that line's realization down explicitly. Don't take it at face value.
Stranded costs. Sometimes a cost the ledger assumed would disappear is contractually locked in after all. It might be a shared-services agreement with a remaining term, or a lease that can't be broken without a penalty. It might be a long-term supply contract with a minimum-purchase commitment written against the very facility slated for closure. These costs don't vanish. They turn a "permanent saving" line into a "saving, delayed until the contract lapses" line. That line belongs in the timing section below, so don't just delete it from the ledger.
Regulatory-mandated remedies. A competition authority's approval can come with a condition: divest an overlapping plant, a customer relationship, or a brand, to preserve competition. Every dollar of synergy riding on the divested asset leaves the ledger entirely. A rationale that ignored this risk back at the "why not" stage now pays for it twice. It pays once in the deal-approval delay, and again in the shrunken synergy pool.
Integration distraction. Bandwidth management spends running the integration is bandwidth it doesn't spend running the base business. That shows up as a real, if temporary, drag on the combined entity's underlying performance during the integration window. It's typically modeled as a modest, explicitly one-time percentage haircut to combined EBITDA in the first year, kept separate from any of the named synergy or dis-synergy lines above.
Stress test three: timing and phasing
Merger modeling already teaches you to phase a synergy stream for discounting: the percentage of run-rate realized in year 1 versus later years, so a present value can be computed. Here you learn the layer underneath that phasing. It's the operational reason different synergy types ramp at structurally different speeds, along with the sequencing dependencies that make some delays inevitable.
Why synergy types ramp at different speeds. Corporate-overhead and duplicate-listed-entity-cost synergies can land almost immediately. A board seat, an audit engagement and a listing can all be eliminated at or near Day 1, because nobody else's cooperation is required. Procurement synergies ramp over two to four quarters, gated by vendor contracts coming up for renewal. You can't break a fixed-term supply agreement early without cost, so the saving arrives on the contract's own clock. Footprint and headcount synergies take the longest of the cost-side lines, twelve to twenty-four months. Statutory notice periods, lease-exit timing, and the physical logistics of consolidating a facility all gate them.
Revenue/cross-sell synergies take longest of all, eighteen to thirty-six-plus months. They're gated by sales-cycle length and by the time a retrained sales force needs before it can sell an unfamiliar product credibly.
How should you phase synergies over time?
Sequencing dependencies. Some delays come from physics, and no amount of optimism moves them. You can't close a redundant plant until the receiving facility has validated it can absorb the volume. You can't cut a shared-services headcount until the new combined system it depends on has actually gone live. So a synergy that "looks late" on a slide is a red flag only if there's no structural dependency behind the delay. Interrogate the one that looks late for no stated reason. It usually means no one is actually driving it.
The Day-1/100-day/Year-1/Year-2+ frame. A usable phasing ladder has four rungs. Day 1 covers the actions that need no cooperation and no notice period. Think corporate-cost eliminations, plus contract terminations with no penalty. The first 100 days bring quick wins, renegotiated vendor terms coming due, and redundant senior-executive compensation. Year 1 holds the actions gated by notice periods and system-migration completion. Year 2 and beyond is full run-rate: footprint fully rationalized, revenue synergies beginning to show, systems fully migrated. Build every synergy line against this ladder explicitly. Don't assume one blended "50% year one, 100% thereafter" ramp for every line regardless of type, because that blended shortcut is what hides the sequencing risk above.
Stress test four: cost-to-achieve
Synergies aren't free. Every one of them costs money to collect, and that cost (cost-to-achieve, or CTA) deserves its own ledger. Phase it across the same years as the synergies it enables, instead of folding it into a single lump-sum guess.
The CTA categories. There are four. Severance and retention covers redundant-role payouts, plus retention bonuses for the flight risks you flagged in the dis-synergy work. Systems and data migration means consolidating ERPs, CRMs, or cloud platforms. That cost typically precedes the IT-synergy line it enables, so it shows up as a negative before the saving shows up as a positive. Real-estate exit covers lease-break penalties and decommissioning costs for closed facilities, timed to follow the physical closure. Professional/integration-office fees are the legal, consulting, and program-management costs of running the integration itself. Keep them apart from the deal-advisory fees paid to close the transaction. Those are a one-time transaction cost, and they don't belong in synergy capture.
Phase it, don't lump it. CTA is rarely one number in one year. It's a schedule that follows the synergy ledger's own timing. Severance for a Year-2 facility closure gets paid in Year 2, and Year 1 never sees it. In the early years it frequently exceeds the synergies realized so far. For many deals, that makes the near-term net-synergy number negative in Year 1. The headline "run-rate synergy" figure hides that completely. A board should hear it explicitly before it signs off.
The reconciling bridge: gross to net, twice
Everything above exists to produce one artifact: a bridge. It starts at the naive number a workstream lead first pitches and ends at a number a board can actually rely on, with every step shown and every total tying out. Here's the structure, applied the same way whatever the currency:
Naive gross (sum of first-draft lines) → minus double-counting removed → Double-count-adjusted gross (full run-rate) → apply the operational ramp, year by year → Year-by-year realized, gross → minus dis-synergies (recurring and one-time) → Year-by-year net → minus cost-to-achieve, phased → Year-by-year net of CTA.
Two of the worked examples in the full module (Trishul Industrial Supplies in India ₹, Northfield Analytics in US $) build this bridge end to end, in full, for two different deal profiles. Both close on the number a walk-away decision actually turns on, the breakeven synergy-realization percentage: premium paid ÷ capitalized value of the risk-adjusted run-rate synergy.
How do you read the breakeven realisation rate?
Reading it is simple. A number comfortably below 50% signals a deal with real margin of safety. Near or above 90%, the deal needs near-flawless execution to be worth what is being paid for it. And above 100% (WE6 shows this case) the deal destroys value even if every synergy on the ledger arrives as planned, with zero execution slippage. That last case is the cleanest "why not" verdict you can produce.
No amount of optimism about execution can rescue a deal whose fully realized synergy case is smaller than the premium being paid for it.