Integration: Where Deal Value Is Won or Lost, and How the First Hundred Days Are Sequenced

Integration is the work of deciding what two businesses now do together and then actually making it happen: which systems, teams, sites, customers and processes combine, which are left apart, and in what order. Integration is the stage where a promised saving either becomes profit or stays a slide. Integration is also the only part of a purchase that carries on after everybody involved has been paid.

What is integration, and when does it actually start?

Think about two shops on the same street being brought under one owner. The purchase is signed on a Friday. On Monday morning somebody has to answer questions that were never questions before. Do both shops still order stock separately, or does one person now order for both? Do the two tills still cash up into two bank accounts? If a customer walks into the first shop holding a receipt from the second, does anybody accept it? None of those are grand strategic questions. Every one of them has to be answered by a specific person, and until it is answered, somebody is standing at a counter guessing.

The questions on that Monday morning are integration, and the shops version contains the whole of it. Scale it up and the questions get bigger, more numerous and more expensive, but they do not change shape. Integration is not a phase that follows a purchase; it is the set of decisions that a purchase creates, and those decisions exist from the moment the two businesses stop being separate.

Now the part almost everybody misjudges about timing. Completion day is when the money moves and the shares change hands. Integration looks as though it starts there, and it does not. The planning has to run before completion. The decisions that need access to the other business have to be queued up while that access is still being negotiated. A stock ordering process nobody has been shown cannot be mapped. Nobody can decide which of two finance systems survives until somebody has said what is in both. And the baseline that every later saving will be measured against cannot be fixed once the two sets of books have already been mixed together.

Late planning follows a pattern worth naming. While it is under way the pattern looks like ordinary busyness rather than like a failure. The first weeks after a purchase are the only weeks in which everybody expects change, and a business that starts planning at completion spends those weeks deciding what the change will be rather than making it. By the time the plan exists, the acquired business has settled back into its old rhythm, the goodwill of the first fortnight has been used up on meetings, and every subsequent change now has to overcome the answer that things have been fine as they are. The window did not close because anybody opposed the plan. The window closed because nothing arrived while it was open.

Harivansh Packaging Limited has bought 100 per cent of Sundarban Polymers Private Limited, and the transaction is done: an enterprise value of Rs 1,320 crore, less the Rs 180 crore of net debtWhat a business owes on borrowings once the cash it is holding has been set against them. A single figure that nets the two sides. that Sundarban Polymers brought across with it, leaving an equity value paid to the sellers of Rs 1,140 crore. The bridge from enterprise value to equity value is settled elsewhere. Integration starts from the morning after. The money has gone. The sellers have been paid. The chief financial officer, Devyani Kulkarni, now has a business she did not have a fortnight ago, and Ashwin Rege, who led the transaction team, is being asked what happens next.

What exactly does this integration have to deliver?

Before deciding how much of two businesses to combine, the size of the answer has to be in view. Sizing the requirement first is the single most useful move in an integration, and it is almost never made in practice.

Two sources paid for it: Rs 140 crore of cash the buyer was already holding, and Rs 1,000 crore raised as fresh borrowing on a contracted 9.0 per cent. Interest on that borrowing runs at Rs 67.5 crore a year once tax relief at 25.0 per cent is allowed for. The interest sits against what was acquired. Harivansh Packaging Limited earned Rs 225 crore of profit after tax. Sundarban Polymers adds Rs 61 crore, rounded from the Rs 61.35 crore that the exact chain gives. Rs 225 crore plus Rs 61 crore less Rs 67.5 crore leaves Rs 218.5 crore. On an unchanged 18.00 crore shares, earnings per shareThe profit left for shareholders divided by the number of shares in issue. Earnings per share moves when either the profit or the share count moves, and it can fall while total profit rises. moves from Rs 12.50/- down to Rs 12.14/-. The fall is a dilution of 2.9 per cent. Worked on the unrounded Rs 61.35 crore instead, the fall comes out a shade smaller and the direction is unchanged, so the dilution survives either treatment of the rounding.

Read the last two rows slowly. The gross-up is where people stop. The shortfall is Rs 6.50 crore, but that is measured after tax. An operating improvement arrives before tax and is then taxed like everything else, so to be left with Rs 6.50 crore the business has to produce Rs 8.67 crore of extra earnings before interest, tax, depreciation and amortisation, or EBITDAThe letters stand for earnings before interest, tax, depreciation and amortisation. In plain words: operating profit measured before borrowing costs, the tax bill and asset write-downs come out of it.. And that Rs 8.67 crore is net: it is what has to be left standing after whatever the improvement cost to obtain has been paid for.

Now put that figure next to the size of the thing it sits inside. The proportion is what should shape the programme. Combined revenue is Rs 4,060 crore and combined EBITDA is Rs 609 crore. Rs 8.67 crore is 1.42 per cent of combined EBITDA. The same Rs 8.67 crore is 6.6 per cent measured against the Rs 132 crore that Sundarban Polymers earns on its own. And it is 0.21 per cent of combined revenue. Rs 8.67 crore is the whole requirement. Not a transformation. Not a doubling. One point four two per cent of the profit the two businesses already make between them.

Both proportions are true and they answer different questions. The 1.42 per cent gives how much the combined result has to move. The 6.6 per cent gives how hard the acquired business itself has to work, and it is more than four times the first figure. A board that hears only the smaller one may conclude that almost nothing needs to happen. A team inside Sundarban Polymers that hears only the larger one may conclude it is being asked for a step change. Neither reading is wrong, and the honest way to state a target is to quote it in rupees first and then say which denominator any percentage was struck on.

How much of two businesses should actually be combined?

Here is where most integrations are decided by default rather than by decision. Nobody sits down and asks how much of the acquired business should be absorbed. Instead a series of individually sensible requests arrive, each one from a person doing their job properly, and the sum of them turns out to be total absorption. The person responsible for finance systems wants one ledger. The person responsible for procurement wants one supplier list. The person responsible for the brand wants one set of packaging. Each request is reasonable. Nobody ever added them up.

How much to combine is a decision with named options, and it should be taken once, in writing, before anybody starts asking for anything. There are four positions worth naming, and real programmes usually sit at one of them rather than between them.

Moving right along that range changes what is at stake. At position one nothing is spent and nothing is saved, and the acquired business carries on exactly as it did. At position two the savings are real but modest and, crucially, they are invisible to customers: nobody outside the two businesses can tell that two payroll runs became one. At position three the operating machinery is shared while the thing customers actually interact with is left alone. At position four everything is touched, including the sales relationships. Sales relationships are the part of a packaging business that took years to build and can be lost in a quarter.

Here is the discipline that turns this from a matter of opinion into a matter of arithmetic. Put the requirement beside the range and ask which positions can plausibly reach Rs 8.67 crore net of what it costs to get there. If a modest position reaches it, then choosing a more disruptive one is spending relationship risk to buy a saving nobody needed. If no position reaches it, that is worth knowing before the programme starts rather than eighteen months into it. Either way, the question has an answer that can be written down, and writing it down is what stops the four separately reasonable requests from adding up to position four by accident.

Move the control to the far right and watch what happens to the third bar. Past a point the cost of getting there rises faster than the saving does, and the net figure falls back below the line even though the gross saving is at its largest. The gross saving and the net saving stop moving in the same direction well before the range runs out. More combination is therefore not automatically better. At the middle of the range the drawn shapes give an illustrative saving of Rs 16.00 crore, an illustrative cost of Rs 6.53 crore and an illustrative net of Rs 9.48 crore, which clears the Rs 8.67 crore line. The three drawn figures are illustrative shapes chosen to show a direction, not measurements of anything.

What do the figures say about whose way of working is better?

Whose way of working is better is the question the range decision usually turns on in practice, and on this transaction the answer is unusually clean. Harivansh Packaging Limited earns Rs 477 crore of EBITDA on Rs 3,180 crore of revenue, a margin of 15.0 per cent. Sundarban Polymers Private Limited earns Rs 132 crore on Rs 880 crore, also a margin of 15.0 per cent. A weighted average of two identical figures is that figure, so combined, Rs 609 crore on Rs 4,060 crore is again 15.0 per cent.

The two margins are identical, so adopting the buyer's processes on the ground that they are the buyer's is an assertion rather than a finding. That sentence is worth reading twice, because it inverts the usual assumption. The default in almost every purchase is that the acquired business moves onto the acquirer's systems, terms and ways of working. Sometimes that default is right, and there are good reasons for it that have nothing to do with quality: one ledger is cheaper to run than two, one supplier list gets a better price, and one set of quality procedures is easier to audit. Cheapness, price and ease of audit are arguments about the cost of running two of something, and they stand up perfectly well.

The unstated version of the argument is that the larger business must be the better one, and that version does not stand up. Size is not evidence of quality. Harivansh Packaging Limited is roughly three and a half times the revenue of Sundarban Polymers Private Limited and converts revenue into operating profit at precisely the same rate. If anything, a smaller business achieving the same margin is doing so without the buying power that comes with Rs 3,180 crore of revenue. The absence of that buying power is a point in its favour rather than against it, and the record does not carry enough to press the point.

Where two businesses are equally profitable, the burden of proof sits on whoever wants to change something. That is a practical rule and it is easy to apply. Somebody proposes moving the acquired business onto the buyer's planning process. Fine: what does the change deliver, what does it cost, and what evidence is there that the current process is worse? If the answer to the third question is that the buyer is bigger, the proposal has not yet been made. If the answer is that running two planning processes costs a named amount, the proposal has been made and can be weighed.

What does integrating cost, and where does that cost land?

The household version comes first. A household decides to combine two kitchens into one because running two is wasteful. Before anything is saved, there is a carpenter to pay, a bigger fridge to buy, duplicate equipment paid for once already to throw away, and three weeks of eating out because the kitchen is a building site. Every one of those costs lands before a single rupee of the saving arrives, and a household that reports saving money on groceries without mentioning the carpenter has described the year inaccurately.

Combining costs money before it saves money, the spending is real, and it lands in the year it happens rather than being spread across the years the saving runs for. Redundancy costs, system migration, dual running while both systems are live, project management, travel, consultancy support, relabelling, requalifying a plant with a customer: all of them are cash out, and most of them appear as operating cost in the profit and loss account for the year they occur. Some may be capitalised, and whether a particular cost can be is an accounting judgement that belongs to the accounting layer rather than to integration planning. The direction and the timing are what matter to an integration plan.

A saving reported without its cost of achievement has been reported at the wrong number. This happens constantly and it is rarely dishonest. The programme reports what it delivered, and what it delivered is the gross saving. The cost of running the programme sits in a central overhead line that somebody else is responsible for. Both figures are correct in their own report and neither report contains the subtraction. The board sees a delivered saving, the accounts show profit that has not moved, and nobody can explain why.

The instruction that fixes it is unglamorous and completely effective. The cost of integrating belongs in the same document, in the same place, as the saving it produces. Not in the same pack. On the same line, with the net figure computed and shown. If a Rs 4 crore annual saving cost Rs 3 crore to obtain in the first year, that fact belongs beside the Rs 4 crore rather than four slides later. Somebody who sees the two figures separately will not do the subtraction, and somebody who sees them together cannot avoid it. Putting the cost beside the saving also protects the programme: a saving that genuinely cost very little to obtain looks far better when the comparison sits alongside.

There is one more constraint on what integration can spend, and it comes from the balance sheet rather than from the programme. Before the purchase, Harivansh Packaging Limited carried net debt of Rs 600 crore against Rs 477 crore of EBITDA, or 1.26 times. The purchase absorbed the whole Rs 140 crore of cash and layered Rs 1,000 crore of fresh borrowing on top, taking its own net debt to Rs 1,740 crore. Buying 100 per cent also brings the acquired business's Rs 180 crore of net debt with it, so the consolidated figure lands at Rs 1,920 crore.

An integration that needs heavy up-front spending is being planned by a business that has just used all of its cash and taken on Rs 1,000 crore of new borrowing. On a consolidated basis leverage has moved from 1.26 times to 3.15 times, and there are exactly two honest pairings here rather than one. Consolidated, Rs 1,920 crore sits above the combined Rs 609 crore, for 3.15 times. Standalone, the buyer's own Rs 1,740 crore sits above its own Rs 477 crore, for 3.65 times. Which of the two is meant has always to be stated. On either basis the practical consequence is the same: an integration cost is not free money, it competes with debt service, and a programme designed as though the buyer still had Rs 140 crore of spare cash is being designed against a balance sheet that no longer exists.

How to plan Acquisition Integration, in six steps

The procedure is short. A planning method that cannot be held in the head is not a planning method, it is a document. Each step produces one written artefact that somebody can be shown.

1. Decide the degree of combination and write it downPick one of the four positions and record it in a sentence anybody can quote back. Not a principle, a position. Everything that follows either implements that position or is a request to change it, and a request to change it goes back to the person who took the decision rather than being absorbed quietly into the plan.
2. List the things that must be true on the first dayShort, specific, testable. Each item is something a person can be sent to check on the morning itself. If an item cannot be checked by asking one question of one person, it is not a day one item and it belongs on the ordinary plan.
3. Identify what depends on whatThis is the step that produces a sequence rather than a wish list. The acquired business cannot move onto one ledger before the chart of accounts has been mapped. Buying cannot be consolidated before both contract books have been read. The dependencies decide the order, which is why the order is discovered rather than chosen, and why an integration plan built by listing priorities usually falls apart on contact with the first dependency nobody had noticed.
4. Put a named person against each part of the planA name, not a department and not a committee. The test is whether somebody could walk into a room, ask that individual how their part is going, and get an answer rather than a referral.
5. Fix the baseline every saving will be measured against, before completionFreight cost per tonne, headcount by function, spend by supplier category, whatever the promised savings sit on. Once the two sets of books are combined, the separate starting point stops existing and no amount of later reconstruction brings it back honestly. This step has a deadline that the others do not, and the deadline is completion day.
6. Say what will deliberately not be touched in the first yearWrite the list. Publish it. This is the step almost nobody records, and it is the one that stops the plan swelling into everything. Without it, every improvement anybody in either business has ever wanted eventually finds its way into the programme, because the programme is the only vehicle currently moving.

The sixth step deserves its bold sentence because of how it fails when it is missing. Nobody adds items to an integration plan maliciously. People add them because the plan is the one place where change is currently possible, and a good idea that has been waiting three years for a slot finally has one. Each addition is defensible on its own. The aggregate is a programme that has stopped being about the Rs 8.67 crore and has become a general improvement effort with no end and no test.

What genuinely has to work on the first morning?

A request for a day one list produces forty items. A request for the four that would cause visible harm before lunch if they failed produces a list that is much shorter and much more useful. Four things genuinely have to work on the first morning, and almost nothing else belongs on that list.

Work through why each of the four is on the list. A person who does not know who they report to cannot escalate a problem, cannot get a decision, and will spend the day asking colleagues who also do not know. A customer who does not know who to call will call somebody, and if that person cannot help, the call becomes an impression about the new arrangement that lasts a very long time. A supplier who does not get paid stops supplying, and in packaging that means a line stops. And a person who cannot get an answer about their own pay will assume the worst available answer. The assumption is both reasonable and corrosive.

All four items are about certainty rather than about improvement. None of them makes anything better. Each of them stops something breaking. A list of forty things has no order and a list of four has nothing else on it, so treating everything as urgent is precisely how the genuinely urgent items get missed.

Notice too what is not on the day one list. The absences are instructive. One email system is not there. A shared email system is a real irritation, it will be raised constantly, and it can wait three months without a single customer noticing. One brand is not there either, and on this transaction it may never be there at all depending on which position on the range was chosen. The strongest test to apply is simple: if it failed on the first morning, would a customer, a supplier or an employee suffer something concrete before the end of the day? If not, it is not a day one item.

Who is actually accountable for integration?

Here is the failure that shows up more reliably than any other, and it does not look like a failure while it is happening. Integration that belongs to everybody belongs to nobody in particular, and that is the most common reason integrations fail. There is a steering group. There is a weekly meeting. There are workstream leads. A decision that affects two businesses cannot be taken by a person who represents only one of them, so everybody is engaged, the meetings are well attended, and nothing gets decided.

The fix is a single named individual with the authority to decide between the two businesses rather than merely to escalate between them. The difference between deciding and escalating is the whole of it, and it is not about seniority. A very senior coordinator with no decision rights produces exactly the same queue as a junior one. The test is whether, when the acquired business says its planning process should survive and the buyer says its own should, this person can say which one survives and have that stand.

A lack of authority produces not disagreement but a queue: decisions waiting on two chief executives who are both busy running businesses. Each individual decision is small. The chief executives are perfectly willing to take them. But they have between them a few hours a month for this, and the queue arrives at a few decisions a week, so the queue grows. Then the workstreams start working around the queue by making local decisions that are individually sensible and collectively inconsistent, and six months later somebody discovers that two teams built two different answers to the same question.

The household version is a house being renovated by two people who are both at work all day. Every choice waits for the evening, the builder loses a day each time, and eventually the builder starts choosing tiles himself. He is not wrong to. Somebody had to decide and nobody was available.

Why does this go wrong on people rather than on systems?

Ask anybody who has run one of these where the difficulty was, and they will not say the software. The answer is surprising. The software is the part with the budget, the plan and the specialist team, and those three are exactly the reason. A system migration has somebody responsible for it, a budget, a specification and a finish date, and every one of those things pulls it toward completion. It may run late and it may cost more than expected, but it is a bounded problem that somebody is being measured on.

The questions that actually stall an integration have none of those properties. Who decides what the acquired plant runs next quarter. Whose planning process survives. Whether the person who ran finance at Sundarban Polymers still runs finance, or reports to somebody who used to be a peer. Whether a manager whose scope just halved has been demoted. The fate of the sales incentive scheme that the acquired sales team has worked to for six years. None of those has a budget line, a specification or a finish date. All of them are being discussed in corridors from the first week.

The striking thing is that these questions are answerable, and the usual failure is that nobody answered them rather than that they were hard. Somebody could say, in the first fortnight, that the acquired finance lead continues to run finance for that business and reports to Devyani Kulkarni, that the incentive scheme runs unchanged to the end of the current year and is reviewed after that, and that no scope changes are planned before April. None of those answers requires new information. The answers require a decision and somebody willing to state it in public.

Something fills the gap when the questions go unanswered, and it is not idleness. People are excellent at inferring answers from evidence, and in the absence of a statement they will use whatever evidence is available: who was invited to which meeting, whose office moved, which system was chosen, whose title appeared first on an organisation chart. Every one of those is read as a signal and most of them mean nothing at all. RetentionKeeping specific people in place after a sale, usually because they hold knowledge or relationships that nobody has written down anywhere. of the people who know the acquired customers is exactly what a buyer needs most and exactly what an unanswered question erodes fastest. The people with the most options leave first.

There is a connection here to what the buyer paid for. Rs 1,140 crore of equity value went to the sellers of Sundarban Polymers Private Limited against a net worth of Rs 320 crore, leaving Rs 820 crore of goodwillThe amount by which a price exceeds the net worth of what was bought. Goodwill sits on the buyer's balance sheet until an allocation exercise splits part of it into identified intangibles. before any allocation to identified intangibles, an accounting exercise set out under purchase price allocation. The Rs 820 crore represents relationships, know-how and position, and every one of those walks about on two legs. The buyer did not purchase a set of machines at a premium. The buyer purchased a business whose value is substantially held by people who can resign.

When is integration finished?

Most programmes, asked when they end, give a date. On that date some things were delivered, some were quietly dropped, and nobody wrote down which was which. The honest answer is that integration is finished when every promised change has been measured against its baseline and either delivered or formally written off in writing.

The written-off half matters as much as the delivered half, and it is the half people resist. Writing off a promised saving feels like admitting failure. Writing one off is not failure: it is the only way of distinguishing between a change that did not happen and a change that is still coming, and that distinction is exactly what a board needs. A line that was promised at Rs 2 crore, attempted, and found to be worth Rs 0.4 crore should be recorded as delivered at Rs 0.4 crore with Rs 1.6 crore written off, and then it stops taking up space. A programme that never writes anything off carries every disappointment forward forever. Old integration reports read like archaeology for exactly that reason.

A programme with no stated end does not end dramatically. The programme does not fail. Attendance at the weekly meeting drops from twelve to eight to four. The pack gets shorter. At some point somebody notices the meeting has not happened for two months and nobody reinstates it. Nothing was ever declared finished and nothing was ever declared abandoned, so the only honest statement anybody can make about the promised Rs 8.67 crore afterwards is that they do not know.

One more thing about the clock, connecting back to what the transaction itself already set running. The deal terms established that an earn-outA slice of the price paid only if the acquired business reaches an agreed figure in a defined period after the sale, so part of what the sellers receive depends on later performance. and a warranty period both reach past completion, so there are obligations under the agreement running alongside the integration. The earn-out and the warranty period have their own dates, set by the document. The integration does not, unless somebody sets one. The asymmetry is worth noticing: the lawyers gave the commercial obligations an end date and nobody gave the operational ones one.