What switching cost is

Every supply relationship that lasts long enough produces, as a side effect, a barrier to exit. The data are in their format. The integrations point at their endpoints. People have learned their interface. Processes have moulded themselves around their limits.

Switching cost is the sum of everything you would have to spend, once, to leave: the migration project, extracting and converting the data, re-integrating the systems, retraining, and the period in which two systems run in parallel while neither works well.

It is a number nearly every company has estimated at least once, usually to conclude that it is high and therefore one does not move. And then the spreadsheet gets filed, as though the story ended there.

The cost you will never pay is the price you will pay forever

It does not end there. Because you are not the only one who knows that number.

A rational vendor who knows what leaving costs you also knows exactly how much it can raise the price while remaining your best option. It can charge more than the alternative — considerably more — and you will stay, because the exit bill is higher than the premium you will pay. Every year. For as long as the relationship lasts.

Hence the inversion that gives this page its title. Your exit cost is not an expense you will bear: it is the price you will pay for not bearing it. It appears on no invoice under the heading "lock-in". It appears, spread thin and invisible, inside the fee you renew.

The conversion is exact, not rhetorical. If your exit cost is S, the annual cost of the best alternative is A, and you reason over a horizon of Y years at a cost of capital r, then the sustainable lock-in premium is the percentage π at which the present value of the premium paid for Y years exactly equals S. Above that threshold migrating pays; below it, it does not. A vendor who knows it stops one cent below.

The scale of the exit cost

What is striking is how little it takes to build serious lock-in. The table starts from the leanest exit cost imaginable and adds, one at a time, the items nearly every estimate forgets. The alternative's annual cost stays at 100,000, the horizon four years, the cost of capital 8%.

Every row adds an item no contract names, and every item turns into pricing power for the vendor. An exit cost of two years of fee — which in enterprise software is ordinary, not extreme — lets the incumbent charge 60% more than the alternative while remaining the economically correct choice.

Note that none of these items is a price. They are consequences of how the system was adopted: how many integrations, how much system-specific training, how much dependent operation. You do not buy lock-in: you build it, one project at a time.

How this calculator works

Five values are needed, and none depends on currency: enter the amounts in yours and read the results in the same one.

The annual cost of the alternative is the fee you would pay the best available competitor for the same function. It is not a list quotation: it is what you would really pay, after the negotiation you would run if you seriously intended to move.

The migration project cost gathers the one-offs: extracting and converting data, rewriting integrations, parallel licences, consultancy, training. The disruption cost is what you lose during the transition: dual running, errors, the productivity that does not return for some months.

The decision horizon is the number of years over which you compare the two roads. It is not the software's useful life: it is how far the decision-maker looks. The discount rate is your cost of capital, and it matters because the exit cost is paid today while the savings arrive later.

How to read the results

Total exit cost is the sum of the two one-off items. It is the number usually computed, and the one people usually stop at.

The maximum price the incumbent can charge is the same information turned outward: it is the fee beyond which migrating genuinely pays, and therefore the natural ceiling of any renewal demand. The annual premium is what you hand over every year if the vendor has got there.

The sustainable lock-in premium is the verdict metric, and it is dimensionless: a percentage, identical in any currency and at any scale. Below 15% lock-in is weak and you negotiate the renewal. Between 15% and 35% it is moderate. Between 35% and 65% it is strong: the incumbent can apply that premium and you will stay anyway. Above 65% you are not negotiating a renewal, you are receiving a price announcement.

Why a short horizon makes you more captive

Here lies the most useful consequence, and the least intuitive.

The sustainable premium is the exit cost divided by the present value of one year of saving, multiplied by the number of years over which that saving accrues. Lengthen the horizon and the denominator grows: the same barrier to exit spreads over more years of benefit, and the premium the vendor can sustain falls.

At the reference values, looking one year ahead grants the incumbent a premium of 162%: practically any price. Looking four years ahead cuts it to 45%. Looking eight brings it below 27%.

It follows that decision myopia is not neutral with respect to bargaining power: it is the vendor's principal ally. An organisation appraising renewals against the current financial year has already surrendered, before sitting down, most of its leverage. And the same organisation, were it to accept a multi-year horizon, would find that the migration which looked unaffordable becomes the obvious choice.

Why expensive capital deepens lock-in

There is a second, quieter effect, and it strikes those who can least afford it.

The exit cost is paid now. The savings arrive in later years. Discounted at a high rate they are worth less, so a larger premium is needed to justify migrating. The premium the vendor can sustain rises with your cost of capital.

At the reference values, a firm discounting at 8% grants the vendor 45%; the same firm, identical in every other respect, discounting at 20% because credit is dear, grants it nearly 58%. The vendor has done nothing different: it is the customer's balance sheet that deepens the lock-in.

It is an unpleasant regularity and worth naming: financially fragile companies pay more for the same software than solid ones do. Not because they receive a different price list, but because their ability to leave is lower, and the market prices it.

How this differs from the cost of a migration

If you have used the software migration cost calculator, you have already met the left-hand side of this equation: what moving costs, and how long the benefit takes to repay the project.

That calculator answers "does the migration pay for itself?". This page answers a different, complementary question: "what is it worth to my current vendor that I will not migrate?". The first number decides a project; the second decides a price, every year, including the years in which the project is not undertaken.

The distinction matters because the two quantities move in opposite directions. Reducing migration cost — through portable data, decoupled integrations, non-specific skills — does not merely make a move possible: it lowers, above all, the price you pay if you stay. It is an investment that yields even when it is not exercised, like an option. To look at a system's whole spend over time, the total cost of ownership calculator lines up the recurring items.

Limits of the estimate

This tool is a simplified model for educational purposes, not contractual advice. It assumes the two solutions are functionally equivalent and that price is the only difference, which is almost never true: if the incumbent offers something the alternative lacks, that difference must be added to the sustainable premium before crying scandal.

It assumes the vendor is rational, informed and unconstrained: in reality many incumbents leave part of the premium on the table, out of reputation, framework agreements, or simply because they have not computed your exit cost. The number you read is the maximum sustainable, not a forecast.

It ignores future exit costs from the new vendor, which are rarely nil, and it neglects the execution risk of the migration, which is the most honest reason not to run it. Use it to know your number before entering a renewal negotiation, and to see where investing to lower it pays; for the contract, get advice.

The proof: the premium is your exit cost

When people say a one-off expense "turns into" an annual rent, it is usually a metaphor. Here it is not, and it can be checked line by line.

Take the reference case: exit cost 150,000, alternative at 100,000 a year, four-year horizon, capital at 8%. The calculator says the incumbent can apply a 45.29% premium, that is 45,288.12 a year. Discount those four payments and add them up.

The final cell is not an approximation. It is exactly the exit cost we started from, to the cent. By charging you 45.29% more for four years, the vendor is extracting, in present value, precisely the sum you would have spent to be free of it.

That is the whole page in one table. You do not have two options, an expensive one and a cheap one: you have two options that cost the same, and in one of them the money stays with you as future freedom, while in the other it already belongs to the vendor. Whoever "decides not to migrate in order not to spend" has already spent.

What lengthening the horizon is worth

The exit cost is given. The cost of capital, nearly. But the horizon over which you compare the two roads is yours to choose, and nobody imposes it.

The table holds everything fixed — exit cost 150,000, alternative at 100,000, capital at 8% — and changes only how many years you look ahead.

The first step is worth more than all the others combined. Moving from a one-year to a two-year horizon halves the vendor's pricing power, and it requires no investment: it requires only that the procurement committee stop measuring renewals against the current financial year.

The reading for whoever negotiates is direct. Turn up at the renewal having authorised a one-year appraisal and your counterpart can ask almost any figure and be right. Turn up with a multi-year mandate — and a credible quotation for the alternative in hand — and the ceiling drops eighty points before you open your mouth.

How an exit cost is built, and how it is dismantled

Exit cost is not a property of the vendor: it is a property of how you adopted it. The same two companies, running exactly the same software, can have exit barriers differing by a factor of five.

It is built every time the short road is chosen. The integration written against the proprietary API instead of against an abstraction. The custom field that exists only there. The report nobody would know how to rebuild elsewhere. The certification your engineers earned on that product rather than on the craft. Every shortcut is a brick in the wall, and no single shortcut looks important while you lay it.

It is dismantled with four moves that do not require changing vendor. Exportability: demand contractually a complete extraction in an open, documented format, and test it for real once a year, not on paper. Decoupling: a thin abstraction layer between your systems and the vendor, which moves the rewrite from a hundred integrations to one. Transferable skills: train people on the domain, not on the interface. Exit assistance: negotiate it on the way in, when you have power, not on the way out, when you have none.

None of these moves presupposes that you want to leave. That is exactly the point: they lower the price you pay by staying. They yield every year, including the ones in which you never exercise them.

The lock-in that appears in no contract

Search the contract for the word and you will not find it. Term clauses, guaranteed minimums, early termination penalties are explicit lock-in: visible, negotiable, and the small part of the problem.

Serious lock-in is tacit and accumulates in three layers. The first is the data: not so much their ownership, which modern contracts grant you, as their format, their history, the internal identifiers every other system points at. An export that hands you rows without semantics is, in practice, a non-export.

The second is the integrations. Every connection between the vendor and the rest of your application landscape is a piece of work to redo. This is where exit costs grow quadratically, because you do not migrate a system: you migrate a system and everything that talks to it.

The third, and the most underrated, is the processes. Through years of working around a product's limits, the organisation has built procedures that make sense only inside those limits. Changing vendor then means redesigning the processes, and the cost of the redesign is charged to the migration even when those processes were already wrong.

It is worth separating the three items when you fill in the calculator. The first is fought with formats, the second with architecture, the third — the dearest — only by admitting that part of the exit cost is not a cost of changing: it is organisational debt that the change finally forces you to settle.

How to negotiate once you know your number

Knowing the sustainable premium changes the negotiation, but not in the naive way. There is no point telling the vendor you know what it can charge: the point is to make its number smaller, before you sit down.

The first move is having a real, priced alternative. Not a name, a quotation. The calculator starts from the alternative's annual cost because without that number the sustainable premium is infinite: with no alternative there is no ceiling, and any demand is acceptable by definition. Half the work of a renewal is done beforehand, with a vendor you will not choose.

The second is bringing the right mandate. A one-year mandate is worth, at the reference values, a 162% premium; a four-year one, 45%. It is the single variable that gives you the most leverage and the only one that depends entirely on an internal decision.

The third is spending part of the premium instead of paying it. If the incumbent is collecting 45,000 a year in premium, investing a fraction of it in exportability and decoupling lowers the exit cost, hence the premium, hence the price — and the return on that investment is measured in points of fee, every year, forever.

The fourth is knowing when the number says stay. A premium below 15% means lock-in is not the reason for your price, and that a migration would be value destruction dressed up as independence. The calculator serves as much to decide to leave as to decide, knowingly, to stay.

Mistakes that hand over pricing power

The first is confusing the fee with the cost. Two offers get compared on list price, and the lower one wins; then it turns out one of them carries twenty proprietary integrations. Future exit cost should be estimated before signing, while it is still a choice, and not afterwards, when it is a fact.

The second is telling the vendor you have no intention of changing. It happens often, out of courtesy or laziness, and closes the negotiation before opening it: an incumbent that knows it has no competition applies the full premium, and it is rational to do so.

The third is treating the exit cost as a fixed number. It is not: it is the result of architectural choices that can be reversed, often at a cost far below the annual premium they are generating. Whoever regards it as immutable has already accepted the rent.

The fourth is measuring the success of a negotiation by the discount obtained against the opening demand. The opening demand is what the vendor calculated it could ask. Getting ten per cent off a fifty per cent premium means having paid forty-five. The correct benchmark is not the demand: it is the price of the alternative.

Frequently asked questions

What is switching cost?

It is the sum of everything you would spend once to move to another vendor: the migration project, extracting and converting the data, re-integrating the systems, training, and the disruption of the transition period. It does not include the new vendor's fee.

How is the sustainable lock-in premium calculated?

Exit cost = project cost + disruption cost. Annuity factor = (1 − (1+r)^−Y) ÷ r. Sustainable premium = exit cost ÷ (annual cost of the alternative × annuity factor). It is the percentage extra the incumbent can charge before migrating becomes worthwhile.

Why does the exit cost become a price?

Because a vendor who knows it knows how far it can raise the fee while remaining your best choice. You will never pay the exit cost: you will pay it as an annual premium, for as long as the relationship lasts. It is the same amount, converted from a one-off into a rent.

What does a 45% premium mean?

That the incumbent can charge 45% more than the alternative's annual cost and you will stay anyway, because the present value of that premium over the chosen horizon exactly equals your exit cost. Above that threshold migrating pays; below it, it does not.

Why does a short horizon raise the premium?

Because the exit cost spreads over fewer years of saving. At the reference values, looking one year ahead grants the vendor a 162% premium; four years cuts it to 45%; eight brings it below 27%. Decision myopia is the vendor's principal ally.

Why does the cost of capital deepen lock-in?

Because the exit cost is paid today and the savings arrive later: discounted at a high rate they are worth less, so a larger premium is needed to justify migrating. Financially fragile companies pay more for the same software, because their ability to leave is lower.

How does this differ from the migration cost calculator?

That one answers "does the migration pay for itself?" and decides a project. This one answers "what is it worth to my vendor that I will not migrate?" and decides a price — every year, including the years in which the project is not undertaken.

How do you reduce the exit cost?

With data exportable in open formats, integrations decoupled behind an abstraction layer, skills that are not vendor-specific, and contracts providing for exit assistance. Every point of reduction lowers the price you pay even if you stay: it is an investment that yields without being exercised.

Is the computed premium what I will actually pay?

No: it is the sustainable maximum, not a forecast. Many vendors leave part of the premium on the table out of reputation, framework agreements, or because they never computed your exit cost. The number tells you where the ceiling is before you sit down to negotiate.

Does the calculator work in any currency?

Yes. Amounts are treated as pure numbers: costs in and results out in the same currency. The lock-in premium is a percentage, hence dimensionless, and it does not change if you multiply every amount by the same constant.

What if the incumbent offers more features than the alternative?

Then part of the premium is paid for those features, not for lock-in. Estimate the annual value of the functional difference and add it to the alternative's cost before comparing: the residual premium is the one attributable to the exit barrier.

Does this tool replace contractual advice?

No. It is a simplified model assuming functional equivalence, a rational vendor, and no future exit cost from the new vendor. Use it to know your number before a renewal and to decide where to invest to lower it; for the contract, get advice.

Updated July 2026