The Alignment Illusion: Why the Business–IT Gap Never Closes — and What Closes It Instead
A problem that stays at the top of the list for a decade, despite concerted effort, is not unfinished; it is misdiagnosed.
Executive Summary
For most of the past decade, “alignment” has sat at or near the top of every list of the concerns that occupy information-systems leadership. The surveys are remarkably consistent: whatever else changes from year to year, the distance between what the business wants and what its systems deliver returns as the unsolved problem. And each year the field answers with more machinery — steering committees, IT strategies cascaded from the corporate plan, chargeback and service-level agreements, maturity models that promise to lift a firm from one rung of alignment to the next.
The persistence is the point. A problem that stays at the top of the list for a decade, despite concerted effort, is not unfinished; it is misdiagnosed. This paper makes the case that alignment, as the field commonly frames it, is a structural illusion — the image of a solution generated by the very arrangement that produces the problem.
The argument in brief. To speak of aligning “the business” and “IT” is to assume two separate entities that a well-engineered linkage can bring into correspondence. That separation is not a natural starting condition to be managed better; it is a choice of organisation. Once made, it turns alignment into a permanent, asymptotic pursuit — forever approached, never reached — because each mechanism built to close the gap also hardens the two sides it spans. The more seriously a firm takes alignment, the more it institutionalises the division that makes alignment necessary.
The recommendation is not to abolish technical professionalism or shared infrastructure; that objection is serious and is answered directly below. It is to stop applying one model to two different problems. Where technology is a utility, alignment is the wrong word for what is really a question of efficient supply. Where technology differentiates, the separation is fatal, and the answer is not a better bridge but the removal of the seam: organising around business capability with technology embedded in it, funding outcomes rather than a function, and making the accountable business leader the owner of the system rather than its customer.
The Metaphor Is the Mistake
Consider the word itself. To align is to bring things into a straight line — axles, gears, the sights of an instrument. The metaphor is mechanical, and it smuggles in an assumption: that there are two distinct components, each with its own separate existence, which the right linkage can bring into correspondence. Adopt the word and you have already accepted the premise. The task then becomes building the linkage — the committee, the cascade, the shared scorecard — while the two components remain exactly where they were, now merely joined by machinery.
This would be harmless if the premise were sound. But in most enterprises there is no coherent object called “the business” standing apart from an object called “IT,” waiting to be aligned. A bank is its systems: its products are records in a database, its service is the behaviour of an application, its controls are lines of code. When the systems stop, the business stops — not metaphorically, but at the counter and on the telephone. The recent, expensive experience of preparing for the century date change made this plain to every board that lived through it. For the better part of two years, firms traced the tendrils of their old systems into every corner of operations and found there was almost nothing the technology did not touch. The lesson available in that moment was profound: there is no business that is separable from its systems. And yet, the date safely passed, the two-camps model reasserted itself within the year, as though the discovery had never been made.
“Alignment is the name we give to the friction between two organisations we chose to create, and then mistook for a law of nature.”
The separation, in other words, is manufactured and then naturalised. The moment a firm stands up a distinct IT function — its own budget line, its own director, its own planning cycle, its own culture and career ladder — it creates two constituencies with different incentives and different languages, and alignment becomes the permanent problem of reconciling them. It is worth being precise about the mechanism, because the orthodox response mistakes it entirely.
A Problem That Refuses to Leave the Top of the List
The productivity paradox has hung over this field for years: through a decade of extraordinary investment in information technology, the returns to that investment have been strangely hard to find in the aggregate figures. Many explanations have been offered — measurement lag, the mismanagement of implementation, the time it takes for organisations to learn how to use what they have bought. But set the paradox beside the alignment literature and a more uncomfortable reading appears. If the missing ingredient were alignment — if the returns were being lost in the gap between business intent and technical delivery — then a decade of concerted alignment effort should have begun to close it. Instead the gap and the paradox coexist, each stubbornly persistent, which suggests that the machinery aimed at the first is not touching the cause of the second.
There is a tell in the survey data that the field has been slow to read. When the same concern tops the list of executive priorities year after year — through recession and recovery, through the mainframe’s long decline and the arrival of the web — the constancy is itself diagnostic. Genuine problems that receive serious, sustained investment tend to move: they get better, or they are at least displaced by newer problems. A concern that does neither, that sits immovably at the summit while fortunes are spent addressing it, is almost certainly being addressed at the wrong level. The alignment problem behaves exactly like a symptom being treated as though it were the disease.
Anatomy of an Alignment Programme
The pattern is easier to see in the particular than in the abstract. Consider a composite drawn from the kind of programme that has become almost standard in large financial-services firms — no single institution, but a picture true to many.
A mid-sized insurer, weary of the perennial complaint that IT and the business were pulling in different directions, launches a two-year alignment initiative. It does everything the literature recommends. A business-IT steering committee is established, meeting monthly, chaired at executive level. The IT strategy is formally cascaded from the corporate strategy, each technology objective traced back to a business goal in a handsomely bound document. Chargeback is introduced so that business units see the cost of what they consume; service-level agreements are written so that expectations are explicit. An assessment is commissioned, and the firm’s alignment maturity is judged to have risen over the two years from the second of five levels to the third — real, measurable progress against the very framework the discipline prescribes.
And underneath, the opposite is happening. Because the governed, central portfolio can only move at the speed of its committee, the business units do what business units under pressure always do: they route around it. A product manager who cannot get a change prioritised builds the calculation in a spreadsheet that quietly becomes business-critical. A department stands up its own database on a server under a desk. Reporting the central function cannot deliver is assembled by hand each month by an analyst whose formal job title bears no relation to the work. By the end of the programme, of the twenty-two initiatives the steering committee had formally ranked as priorities, nine have gone live. Meanwhile the shadow estate — the departmental systems that never appear in the IT budget — is estimated, when someone finally troubles to look, at somewhere between a third and a half of official IT expenditure.
Read the two results together. The measured relationship improved; the actual gap widened. This is not a story of a programme executed badly. It is a story of a programme executed well, doing precisely what its design causes it to do.
Why the Machinery Cannot Win
The reason the machinery cannot close the gap is that it governs the visible relationship while real demand escapes into the invisible one. Every alignment mechanism is, at bottom, an interface between two parties — and an interface presupposes, formalises, and then perpetuates the very boundary it sits on.
- A steering committee is a negotiation between two sides. It makes the border more orderly; it does not remove the border. Indeed it gives the border a room, a calendar, and a set of officials whose standing depends on its continued existence.
- A chargeback regime turns a colleague into a supplier and a supplier into a customer. It sharpens cost visibility and, in the same motion, converts a shared enterprise problem into an arm’s-length transaction in which each party optimises its own position rather than the whole.
- A cascaded strategy assumes the business strategy is settled first and the technology strategy follows from it. But where technology is genuinely a source of advantage, the causation runs both ways: what is newly possible reshapes what is worth wanting. A one-way cascade structurally forecloses the very contribution that differentiating technology is supposed to make.
Each of these is a reasonable act of management. Together they produce a paradox anyone who has run such a programme will recognise: the harder a firm works at alignment, the more elaborate its apparatus of separation becomes. Ceremony accumulates. The committee spawns sub-committees; the service-level agreement spawns a disputes process; the maturity model spawns an annual assessment. The organisation grows steadily better at managing the relationship between its two halves, and no closer to not having two halves. This is the asymptote: motion without arrival, effort converted into the upkeep of the very division it was meant to end.
The harder an organisation works at alignment, the more sophisticated its machinery of separation becomes — and the more permanent the division it was built to overcome.
The Objection That Must Be Answered
The strongest case against this argument is not sentimental attachment to the IT department. It is that specialisation exists for good reasons, and they are hard ones. Technology is a real discipline. Systems must be secure, resilient, and maintainable long after the enthusiasm that launched them has faded. Infrastructure carries economies of scale that only a shared, professionally run function can capture; common standards are what stop the estate fragmenting into a thousand incompatible islands. Anyone who has watched the spread of end-user computing — the proliferation of departmental databases and business-critical spreadsheets that no one can audit, reconcile, or recover once their author has moved on — knows precisely what happens when delivery is handed wholesale to the business and the specialists are pushed aside. Dissolve the function, the objection runs, and you do not get fusion; you get anarchy dressed as empowerment.
The objection is correct. It is also fatal only to a position this paper does not hold. Two observations turn it from a refutation into a refinement.
The first is that the anarchy the objection fears is not a hypothetical consequence of dissolving the IT function; it is the current, observed product of the alignment model. The spreadsheet sprawl, the servers under desks, the unauditable departmental databases — these are not what happens when the specialists are removed. They are what happens when a governed central function, held at arm’s length by the very machinery meant to align it, cannot serve the demand in front of it, and that demand escapes into the shadows. The end-user computing mess is not an argument against reorganisation; it is the alignment model’s own handiwork, and the most vivid evidence that the model is failing.
The second is that embed the technology and abandon technical stewardship are not the same proposition, and the distinction is the whole of the recommendation. There is a layer of technology for which shared, disciplined, professional stewardship is exactly right — and a layer for which it is exactly wrong. The error of the alignment discourse is to apply a single organising model to both.
Two Kinds of Technology, Two Different Models
The estate any large firm runs is not one thing. Part of it is utility: the payroll system, the general ledger, electronic mail, the network, the commodity processing of standard transactions. No customer will ever choose the firm because its payroll runs well, and no competitor is meaningfully disadvantaged by running the same package. For this layer the right questions are the supply questions — is it reliable, is it secure, are we buying it efficiently — and here the professional, centralised, standards-driven function is not merely defensible but ideal. Tellingly, alignment is not even the right frame for it: one does not align with the electricity supply, one procures it well.
The other part is differentiating: the systems that constitute the firm’s actual advantage, that a competitor could not simply buy and switch on. Here the utility model is fatal, because advantage of this kind cannot be specified in advance by one side and delivered by the other. It has to be invented jointly, in the same room, by people who understand the market and people who understand what is technically possible, each changing the other’s sense of what is worth doing. This is the layer where separation destroys value — and where, precisely because there are not two things but one, alignment names an impossibility rather than a goal.
| Dimension | Utility technology | Differentiating technology |
|---|---|---|
| What it is | Payroll, ledger, mail, network, commodity transaction processing | The systems that are the firm’s advantage |
| The right question | Are we supplying this reliably and efficiently? | Are business and technology inventing this together? |
| Governing model | Supply — benchmarks, total cost of ownership, service levels, selective outsourcing | Fusion — shared accountability, joint teams, funding by outcome |
| Where “alignment” fits | Unnecessary: it is a purchasing relationship | Impossible: there are not two things to align |
Seen this way, the outsourcing debate that has divided the field resolves cleanly. Outsourcing is a sound answer to the utility question and a ruinous answer to the differentiating one. Firms that outsource the utility layer sharpen their focus; firms that outsource the differentiating layer sign away the joint capability on which their future advantage depends, and convert an internal problem they might have solved into a contractual one they cannot.
What to Do Instead
The prescription follows from the diagnosis. It is not a new alignment mechanism — the field has enough of those. It is a change in the unit around which the enterprise is organised, funded, and held to account.
- Sort the estate before you reorganise it. Divide the technology honestly into the utility layer and the differentiating layer. Most firms will find the split uncomfortable: much of what the IT function treats as strategic is in fact utility, and some of what the business dismisses as trivial is quietly differentiating. The sorting is itself clarifying.
- For the utility layer, stop aligning and start supplying. Manage it as a service: benchmark it, cost it honestly, hold it to explicit levels, and outsource the parts where an external provider genuinely does it better and cheaper. Here the professional central function belongs, and here the language of supply — not alignment — is the correct one.
- For the differentiating layer, remove the seam. Do not build a better bridge between business and technology; put them in the same team, accountable to the same outcome, funded as one. The measure of success is not how well the two sides collaborate but that there are no longer two sides to speak of.
- Fund outcomes, not the function. The annual IT budget, negotiated as a block between two constituencies, is one of the great engines of separation. Fund the capability or the value stream instead, with the technology cost carried inside it, so that no one can any longer ask whether “IT spend” is aligned to “business priorities” — the question dissolves for want of two separate things to compare.
- Put ownership where the accountability already sits. The differentiating system should be owned by the business leader accountable for the outcome it serves, not consumed by that leader as a customer of a separate provider. Ownership means the authority to shape it and the responsibility when it fails — the opposite of the arm’s-length service relationship the service-level agreement institutionalises.
- Make technical fluency a general-management competence. For as long as understanding technology is treated as a specialist enclave, the two cultures will regenerate no matter how the boxes are drawn. The leaders of the differentiating business must be fluent enough in what is technically possible to help invent it — not to write the code, but to think natively about technology as a raw material of strategy.
The goal is not a business better aligned with its technology, but an enterprise in which the distinction has stopped being the organising fact of its life.
None of this is free, and it is worth being candid about the cost. Sorting the estate provokes turf disputes; funding by outcome disturbs a budgeting process that many careers are built around; embedding delivery strands the central function with the least glamorous half of its old remit. These are real prices. But they are one-time costs of reorganisation, paid to escape a recurring cost that never ends — the permanent overhead of running, staffing, and forever re-tuning the machinery of a separation the firm chose and then forgot it had chosen.
What Alignment Was Trying to Tell Us
The alignment discourse was never foolish. It was a true observation aimed at the wrong cause. The friction it named is real: in most large organisations, business intent and technical delivery genuinely do pull against each other, expensively and chronically. The error was to read that friction as a gap between two things that needed a better bridge, rather than as a seam within one thing that needed to be removed.
A decade of committees, cascades, chargeback, and maturity models has taught us — at considerable expense — the one thing the machinery could never admit about itself: you cannot align your way out of a structure. You can only reorganise your way out of it. The firms that will pull ahead in the years to come are not those with the most mature alignment programmes. They are the ones that stop trying to align two halves of an enterprise, and quietly arrange never to have had two halves at all.