By Bryan J. Kaus
Plans are worthless, but planning is everything.
— Dwight D. Eisenhower, 1957
On September 21, 2023, the unitholders of Magellan Midstream Partners voted on whether to sell their company to ONEOK for $18.8 billion.
Fifty-five percent of the outstanding units voted yes. For a merger the board had recommended, that is a thin margin. Energy Income Partners, one of Magellan’s largest holders, had spent the summer campaigning against the deal on the grounds that it handed ONEOK roughly $3 billion of tax shield that belonged to Magellan’s owners, and that the synergies offered in return were not worth it.
The synergies were specific. ONEOK’s chief executive, Pierce Norton, had put at least $200 million a year into the base case and said the number could reach $400 million or more within two to four years. About $100 million of the base was overhead that would disappear when two corporate functions became one. The rest was commercial: bundling natural gas liquids services with crude and refined products, using Magellan’s export expertise, and pulling demand through ONEOK’s butane blending business.
Nobody doubted the overhead. The commercial number was the argument. The cost savings were visible on an org chart; the rest was a story about how two pipeline networks would behave once they belonged to the same company.
The deal closed four days after the vote. The arguing stopped, and the proving began.
What a Synergy Actually Requires
Two years later, in its December 2025 investor presentation, ONEOK reported it had realized more than $350 million in annual adjusted EBITDA synergies from the Magellan assets, against the $200 million base and ahead of the $400 million ceiling it had set for a window with two years still to run. It identified up to $350 million more in near-term opportunity and put the longer-term potential at $700 million or better.
The categories matter more than the total, because they describe work rather than arithmetic. Corporate cost savings. Operating efficiencies. Enhanced blending. Logistics savings. Marketing optimization. And in the next tranche, more logistics, connectivity to refined products markets ONEOK could not reach before, and contract bundling.
These are management’s numbers, in a company-defined adjusted EBITDA for which ONEOK itself says it cannot provide a GAAP reconciliation. They are not free cash flow after integration costs and the capital spent to connect the systems, and they are not the return on the deal. But as a description of where value came from, they are useful precisely because none of it arrived with the closing documents.
A pipeline does not move more product because its owner changed. A terminal does not blend more butane into gasoline because of a name on a deed. Butane is a natural gas liquid, and ONEOK had a great deal of it. Gasoline tolerates a certain amount of butane in winter, and a refined products system with the right storage and the right customers can absorb it profitably. Magellan had the gasoline system. Before the deal those two facts belonged to different companies with different commercial teams, different scheduling, and different contracts.
Making the blending synergy real meant somebody understanding the vapor-pressure limits at each terminal, which customers’ contracts allowed what, where the butane could physically get to, and what it was worth there against Mont Belvieu. In June 2024 ONEOK bought roughly 450 miles of Gulf Coast NGL pipeline from Easton Energy for one stated reason: to accelerate the Magellan synergies by connecting its Mont Belvieu assets to its Houston-area ones. The pipe was tied in. Then the commercial work could start.
A synergy begins as a hypothesis on a slide. It becomes money when someone changes how the physical system runs and the change shows up in the results.
The Same Problem at Three Hundred Employees
The scale of the Magellan deal can make the lesson feel like a large-company lesson. It is not.
In December 2010, Sentinel Capital Partners bought Chase Industries, a Cincinnati manufacturer of made-to-order specialty doors: the double-swing impact doors in a restaurant kitchen, the corrosion-resistant doors in a chemical plant, the cold-storage doors on a walk-in freezer. Chase had been in business since 1932 and had perhaps 300 employees. Sentinel’s thesis was to use it as a platform to consolidate a fragmented industry, and Chase’s managers invested alongside the fund.
In August 2011 Chase bought Chem-Pruf, a fiberglass door maker in Brownsville, Texas, that owned the top of the corrosion-resistant market where Chase’s own brands competed at the bottom. Rather than fold Chem-Pruf into Cincinnati, Chase moved its own smaller corrosion-resistant lines to Chem-Pruf’s Texas plant. The acquirer consolidated into the acquired.
In November 2012 Chase bought Eliason, a Kalamazoo company that had been family-owned since 1952 and dominated restaurant traffic doors, Chase’s primary category. Sentinel had identified Eliason during its original diligence on Chase, two years earlier, and waited until the platform could carry it. The integration ran in both directions: Chase’s China sourcing went to Eliason, and Eliason’s lean manufacturing discipline went into Chase’s other plants.
Then came the decision that defines the case. Sentinel and Chase’s board made Eliason’s chief executive, Jeffrey Stark, the CEO of the combined company, and moved several other Eliason executives into leadership roles. The acquirer handed the keys to the people it had just bought, because they were the people who knew how to run the business it now wanted to be.
When Sentinel sold Chase to Audax Group in September 2014, after almost four years, it reported that revenue had doubled and EBITDA had tripled, with both acquisitions integrated.
Those are a private equity firm’s own case-study figures, for the whole enterprise, in a period when U.S. construction and food service were recovering. They are not a measure of what integration alone contributed. What they do show is the sequence of operating decisions behind a number that reads, from the outside, like financial engineering: where to make what, which disciplines to move in which direction, and who should be in charge when the dust settled.
A $18.8 billion pipeline merger and a $300-employee door maker have the same problem. Someone has to decide, and then someone has to do it.
When Commercial Opportunity Outruns Operating Capacity
Now consider a case in which the deciding and the doing fall apart.
A large equipment and infrastructure company buys a specialized industrial services firm. The target has a strong name in its niche, technicians who took years to train, specialized equipment, and customers it has served for two decades. Its commercial reach is small. The buyer has a sales force covering hundreds of accounts the target has never called on.
The thesis writes itself. Introduce the service to the larger customer base, raise revenue, fill the equipment schedule, expand margin.
Now look at what the target was built to do. Its estimating depends on two people who have been pricing this kind of work since the founder started the company. Its pipeline lives in a spreadsheet and in their heads. Its technicians cannot be hired off the street; a qualified one needs certifications and field time measured in years. Its equipment fleet caps the number of jobs it can run in a month. Its contracts, pricing, and revenue recognition look nothing like the parent’s.
The buyer’s salespeople, meanwhile, know how to sell what they have always sold. They do not know what makes this service a fit for a customer, what questions to ask, or when to walk away.
None of this says the deal was wrong. Some of it is exactly why the buyer saw an opportunity. It says the opportunity requires capabilities the transaction did not deliver.
So what happens if the parent turns the sales force loose on day one?
Inquiries surge. Most are unqualified. The two estimators drown. Proposals slip from days to weeks. Customers who were curious lose interest. Worse, the business starts accepting work it has not properly priced and cannot properly staff, and the first few jobs for the parent’s marquee accounts go badly. The pipeline report looks better every month while the ability to convert it gets worse.
A revenue synergy becomes an operating liability in about ninety days.
The alternative is slower at the start and faster everywhere after. Pick the three or four customer segments where the service is most valuable. Write qualification criteria so a salesperson knows which opportunities deserve a technical conversation. Build an estimating process that does not depend on two people, with pricing discipline and a clear owner for each account. Tie it to equipment availability, technician capacity, training, and scheduling. Some of that is investment ahead of revenue: another estimator, a scheduling system, a training program, possibly a second rig. Pilot it in one region. Fix what breaks. Then expand.
The goal is not a busier sales organization. It is a repeatable system for winning and delivering profitable work. The revenue is downstream of the system.
Someone Has to Connect the Decisions
Most acquisitions have an integration plan. There are finance, commercial, HR, technology, legal, and operations workstreams, each with owners, deadlines, and milestones. That structure is necessary. It is not sufficient.
Finance wants overhead out. Commercial wants growth now. Operations wants capacity and service protected. HR wants one pay structure and one org chart. IT wants one system by year-end. Each initiative is defensible inside its own discipline, and each one changes the conditions under which the others will succeed or fail.
Who decides whether the administrative headcount comes out before or after the new commercial systems are live?
Who notices that the sales push cannot proceed until the estimating capacity exists?
Who rules that the procurement savings are not worth switching away from the one supplier whose reliability holds production together?
Who answers for an initiative that improves one department’s numbers and quietly raises costs in another?
Those questions fall into the spaces between organizational responsibilities. The integration can be on schedule, every box green, while the economics of the combined business move the wrong way.
That is why integration leadership is a different job from running the steering committee. It means understanding how the acquisition was supposed to make money, seeing the dependencies among the initiatives, and having the authority to decide or escalate when they conflict.
It is not a job for one person acting alone; a serious integration needs capable leaders in every function and inside the acquired business. Nor does every acquisition need an outside executive. Plenty of companies have deep internal benches and a history of doing this well. The role can sit with a senior executive from the acquirer, with an integration leader hired for the transaction, or in some situations with an experienced outside operator brought in to build the operating model and hand it to a permanent team.
The arrangement matters less than the mandate. A person cannot be accountable for an economic outcome they have no authority to influence, and an integration is not resourced simply because every workstream has a name beside it.
Experience Matters, Especially When the Plan Changes
There is now evidence on who should hold that mandate.
In an April 2026 NBER working paper, Sinan Gokkaya, Xi Liu, and René Stulz examined the business-unit leaders, the executives actually responsible for integrating and operating acquired companies, across a large sample of U.S. transactions. Leaders with prior experience in the target’s industry were associated with more successful integrations and with better realization of the deal’s cost and revenue synergies. Other attributes of those executives, including general deal experience, were unrelated to the outcome.
The finding that matters most for this argument is where the effect showed up. Industry experience in the integration leader improved integration planning and execution. It had no measurable effect on the earlier stages of the deal.
That is working-paper research, and industry experience by itself does not make anyone a good leader. But the shape of the result is the shape of this essay. The capabilities that make a transaction succeed and the capabilities that make an integration succeed are different, and companies routinely staff the second with whoever was good at the first.
They are different because the plan will not survive intact. Some synergies will not show up. Some initiatives will cost more than modeled. Some practices should be left alone longer than planned and others changed sooner. The original model is the starting point, not an instruction manual reality is obliged to follow.
This is what Eisenhower meant. The line gets quoted as a dismissal of planning; he meant the reverse. The value of planning is in the people it produces, people who understand the objective well enough to abandon the plan when conditions change and still get where they were going. That is what a leader who knows the acquired industry brings. When the estimating backlog hits, or the key dispatcher quits, or the blending economics turn out to depend on a terminal nobody put on the slide, they can recognize the problem for what it is and act before a small deviation becomes a material loss.
Sometimes the right call is to accelerate. Sometimes it is to delay a change that would put operations at risk. Sometimes it is to challenge the original assumptions, commit capital to a capability the model underestimated, or admit that a projected synergy is not coming. None of those is a sign of failed integration. They are what competent leadership looks like once conditions have become clearer than they were at closing.
The danger is confusing adherence to the plan with progress toward the objective.
A buyer who has done the work knows what the deal was for. A buyer who bought a slide deck finds out when the numbers miss.
Measure the Value, Not the Activity
The second place integrations go wrong is in what gets reported.
Organizations are good at tracking whether things got done. The org chart is approved. The ERP is migrated. Procurement contracts are consolidated. Headcount reductions are complete. Training is at 100 percent. Each may be necessary. None is evidence that the acquisition is creating value.
Suppose the model assumed $5 million a year in procurement savings. The procurement team renegotiates the contracts and reports the target met. But the new supplier runs longer lead times, so the plants carry more inventory. Deliveries slip, expedited freight shows up on the P&L, and production stalls twice in a quarter. The initiative hit its number. The business pocketed a fraction of it.
The same thing happens on the revenue side. An acquired business can show a sharp increase in sales that, once the thinner margins, the extra working capital, and the strain on capacity are counted, does almost nothing for free cash flow. Headcount cuts that drive overtime and contractor spend overstate the net the same way.
So management should be able to say, for every initiative in the model, which actions produce the benefit, how it is measured, when it lands, what it cost to get, and where the offset is likely to appear.
The same discipline has to be applied to the base business. A company can hit every integration target and still miss the thesis because the acquired operation deteriorated underneath it. A strong market can hide a weak integration for years.
That is why ONEOK’s list matters more than its total. A board told “350millionofsynergies”haslearnedonething.Aboardtold”350 million, of which this much is overhead, this much is blending volume at these terminals, this much is logistics on these routes” has learned something it can check.
Leadership has to separate what the acquired business would have done anyway, what the integration changed, and what the market did to both. Financial results alone will not do it. Operating indicators will: customer retention, on-time delivery, utilization, safety, contract margin, qualified pipeline conversion, working capital, and whether the people the business depends on are still there. Those move before the P&L does.
A completed project and a realized benefit are not the same thing, and an integration report that cannot tell them apart is a status update, not a measurement.
Build a Business That No Longer Needs the Integration Team
The last test comes after the visible work is done.
An acquisition may need extraordinary attention during its transition. Processes have to be built, leadership aligned, constraints cleared. But that intensity should not become permanent. A successful integration leaves behind a business that can operate and improve on its own: clear accountability, capable managers, reliable systems, and the financial discipline to evaluate its own performance.
For a founder-led business, that usually means developing the next layer of leaders and building processes that no longer route through the founder’s desk every morning. For a specialist firm inside a larger parent, it means the commercial, technical, and operating capabilities to serve a customer base several times the size of the one it had. Chase Doors came out of Sentinel’s ownership with the acquired company’s CEO running it and a manufacturing discipline the platform had not possessed going in. That is what leaving a business stronger looks like.
For a serial acquirer there is a further economy. A company that rebuilds its integration approach from scratch every time pays for the same lessons repeatedly and strains the same managers each round. One that develops a repeatable discipline gets better at identifying, buying, and improving businesses over time, which is the whole thesis of a platform strategy. The trap on the other side is a rigid methodology that substitutes for understanding the particular company being bought. The process should become more repeatable. The judgment should stay responsive to the business in front of it.
An integration leader, internal or external, should be judged on that as much as on the first year’s numbers. The most durable contribution is an organization that keeps creating value after the integration effort has ended, and does not need the person who led it to stay.
Where This Would Be Wrong
The case for integration leadership rests on being able to credit the integration, and not the market, with the result. For ONEOK that is a fair challenge. The two years after the Magellan close were good years for Permian volumes and NGL throughput, and ONEOK’s reported earnings now blend Magellan with EnLink and Medallion, both bought in the interim. Chase Doors grew through a construction recovery. Nobody outside either company can cleanly separate the integration from the environment.
So the honest position is that both results should be treated as management’s results.
What can be said is narrower and more useful. Overhead savings do not depend on commodity prices. Blending margins depend on physical connectivity that did not exist until the Easton pipeline was tied in. Logistics savings come from rerouting product across a combined system, which is a scheduling decision, not a market one. Consolidating a product line into the acquired company’s plant, or putting the acquired company’s CEO in charge, are choices somebody made and could have made differently. The categories these companies report are, by their nature, things people did rather than things that happened to them. If the results were mostly market, the categories would read differently.
That is the test for any acquirer’s synergy claim. Not the total, but whether the components describe decisions.
The Point Taken
The value of an acquisition is not realized at closing, or when the organizations are combined, or when the last milestone is reported complete. It is realized when the business produces the economic improvement that justified the price, and when that improvement holds.
Getting there takes financial discipline, commercial judgment, operating knowledge, and an honest understanding of the people and capabilities the business runs on. It takes someone whose job is to connect those things rather than let them develop in parallel and collide.
That person is a capital allocation decision.
A company will commit billions to an acquisition, negotiate the price to the decimal, and structure the financing with care. The same seriousness is owed to the questions that decide whether the business delivers. Who will lead the integration? Do they understand the business? Do they have the authority and the resources to resolve competing priorities? Is the company prepared to invest in the capabilities that are missing? Can it tell real economic improvement from integration activity? And will the changes leave behind a stronger enterprise than the one it bought?
Most acquirers spend months on the price and an afternoon on those.
Magellan’s unitholders were asked to trust that the commercial synergies were real. Fifty-five percent of them did. Two years later the number came in at $350 million and rising, and the only reason it did is that people inside the combined company did the unglamorous work of making two systems into one. Sentinel’s investors were asked to trust that a door company could become a platform. It did, under the leadership of the man whose company it bought.
A compelling thesis is a good beginning.
The measure of the deal is not what management believed the business could become. It is what the organization built.
© 2026 23.5 Strategies; The Point Taken™



