The Law relating to Service Credits

April 7, 2016

Well defined service levels are an important part of any outsourcing or technology services contract as they provide a contractual description of the standard of service the supplier will provide to its customer. It is, however, equally important to clearly define the remedies available to the customer in the event the supplier’s performance of services falls below the requisite standard. In most agreements these remedies will include the right for the customer to receive service credits from the supplier, a pre-agreed fixed amount paid by the supplier to the customer.  

However, there is no standard way of drafting for service credits so that they often receive different treatment in agreements. Consider the following: 

‘Service Credits are intended to operate as a price adjustment to reflect the reduced value of the Services as a result of the Supplier’s failure to deliver the Services in a proper, timely and consistent manner. Service Credits are not intended to recompense Customer in respect of Losses suffered as a result of Service Failures or the Supplier’s breach of this Agreement. Accordingly, any payment of Service Credits is without prejudice to Customer’s right to claim damages in respect of the relevant Service Failure, subject always to the restrictions contained in Clause [ ] below (Liability and Indemnities).’

It is clear from the above that service credits are intended to constitute a price adjustment and a non-exclusive remedy.

By contrast, a different approach is taken by this clause:

 ‘The Parties agree that the payment of Service Credits is not onerous or a penalty and constitutes a genuine pre-estimate of loss likely to be suffered by the Customer in respect of a failure of the Supplier to comply with any Service Level’.

Here service credits are expressed as a form of liquidated damages. There is silence as to whether these are the exclusive or non-exclusive remedy available to the customer. 

Alternatively, some clauses just stay silent: 

‘Service credits shall be payable to [the Customer] in accordance with the KPI table set out in Schedule [ ]’. 

In this example, it is not clear what the legal status of service credits is. It is also not clear on the face of the agreement whether additional damages can be recovered for breach of the relevant KPI.  

Which is the best approach – should service credits be described as liquidated damages, a payment adjustment or something else, and what are the legal and commercial consequences of each? Further, to what extent is it possible to claim the service credit and preserve the right to claim additional remedies? 

Liquidated Damages 

Liquidated damages: penalties 

Most IT lawyers would assume service credits are a form of liquidated damages. Historically, the main concern lawyers have had with regard to expressing service credits as liquidated damages is fear of being caught by the rule against penalties. This has led many to take considerable pains to ensure that the service credits they draft reflect a genuine pre-estimate of the loss likely to be suffered by the customer. Such an approach appeared to be supported by the analysis of Arden LJ (admittedly in the minority) in Murray v Leisureplay [2005] EWCA Civ 963, who seemed to suggest that the court should carefully analyse the reasons for the liquidated damages clause and actively intervene in the case of inappropriate clauses. 

While such caution is not wholly misplaced, recent case law and particularly the Supreme Court’s complex decision[1] in the joined cases of Cavendish Square Holding BV v Makdessi and Parking Eye Limited v Beavis [2015] UKSC 79 (‘Cavendish Square/Beavis’) suggests courts are unlikely to intervene in all but the most extreme cases. 

The Supreme Court carried out a thorough review of the law relating to penalties which Lords Neuberger and Sumption, who together gave the leading judgement, noted: ‘… is an ancient, haphazardly constructed edifice which has not weathered well.’  

In considering the application of the rule, they noted firstly that it applies only in circumstances of breach; by contrast, it does not affect the nature of any primary obligations in the contract which do not arise as a result of breach: 

 ‘The penalty rule regulates only the remedies available for breach of a party’s primary obligations, not the primary obligations themselves’ (at [13]). 

The real issue was how you determine if a clause is penal or not. Fundamentally, they concluded that the test here is whether the clause is ‘unconscionable and extravagant and one which no court ought to allow to be enforced’.[2]  

For Lords Neuberger and Sumption too much attention had been given to Lord Dunedin’s test in Dunlop Pneumatic Tyre Co Ltd v New Garage and Motor Co Limited [1915] AC 79, which has ‘achieved the [misplaced] status of a quasi-statutory code in the subsequent case law’; they also noted that the other judges in the House of Lords in Dunlop did not expressly approve Lord Dunedin’s formulation and the relevant clause in that case failed all but the last of Lord Dunedin’s four tests, ‘which may prove helpful or even conclusive’ (at p 87) to determine if the clause was unconscionable or extravagant:

  • If the sum is extravagant and unconscionable in comparison with the greatest loss that could conceivably be proved to have followed from the breach;
  • The breach consists of non-payment of money, the clause will be a penalty if it requires on breach payment of a sum greater than ought to have been paid;
  • Presumption of a penalty where a single lump sum is made payable by way of compensation on the occurrence of one or more or all of several events, some of which would cause serious damage and others but trifling damage;
  • On the other hand, it is no obstacle to a stipulated sum being a genuine pre-estimate of damage, that the consequences of the breach are such as to make pre-estimation an impossibility. 

Lords Sumption and Neuberger considered these to be a ‘useful tool for….simple damages clauses in standard contracts. But they are not easily applied to more complex cases’ (at [22]). To deal with these, the leading judges in the Supreme Court attempted to determine the rationale for deciding whether a clause is ‘unconscionable’. That rationale they found in the speech of Lord Atkinson in Dunlop. Lord Atkinson emphasised the wider interests of Dunlop in enforcing the damages clause which went beyond recovery of damages.

The Court of Appeal had previously[3] approved the so called ‘new approach’ based on a broader test of commercial justification found in a line of earlier cases: 

  • Lordsvale Finance v Zambia [1996] QB 752, where Coleman J had to consider a provision entitling a lender to interest of 1.5% and a further unexplained 1% in the event of breach. Rejecting the suggestion that the extra 1% was a penalty, Coleman J noted that, although it did not represent a pre-estimate of loss, there was a good commercial reason for the extra interest charge, given that money is more expensive for a less good credit risk. The extra 1% was therefore justifiable. 
  • Murray v Leisureplay [2005] IRLR 946, where Mr Murray’s contract of employment provided he was entitled to one year’s salary and continued pension benefits on termination. At first instance the judge held this amounted to a penalty. The Court of Appeal rejected this approach. The amount was not a genuine pre-estimate but was commercially justifiable taking into account the other provisions of the agreement and taking into account that the company valued his services highly and to that extent included generous reassurances regarding dismissal. That such assurances exceeded the amount he could recover ordinarily for breach of contract did not render the clause penal unless it did so extravagantly. 

The Supreme Court endorsed this ‘commercial justification test’ with some reservations. In particular they had concerns as to the ‘artificial categorisation’ that had occurred between a clause that was ‘in terrorem’, and the emphasis on finding a ‘genuine pre-estimate of loss’. In the view of the Supreme Court these were irrelevant to the real question, which is whether the clause is penal (ie whether it is unconscionable or extravagant by imposing some detriment for breach out of all proportion to any legitimate interest). 

‘A damages clause may properly be justified by some other consideration than the desire to recover compensation for a breach. This must depend on whether the innocent party has a legitimate interest in performance extending beyond the prospect of pecuniary compensation flowing directly from the breach in question’ (at [28]). 


‘The true test is whether the impugned provision is a secondary obligation which imposes a detriment on the contract breaker out of all proportion to any legitimate interest of the innocent party in the enforcement of the primary obligation. The innocent party can have no proper interest in simply punishing the defaulter. His interest is in performance or in some appropriate alternative to performance. In the case of a straightforward damages clause, that interest will rarely extend beyond compensation for breach, and we therefore expect Lord Dunedin’s four tests would usually be perfectly adequate to determine its validity. But compensation is not necessarily the only legitimate interest that the innocent party may have in the performance of the defaulter’s primary obligations’ (at [32]).[4] 

By way of example, in Beavis they found this legitimate interest in: (a) the way the parking scheme was intended to ensure a reasonable turnover of customers at the shopping outlets served by the parking and to prevent misuse of that parking facility by commuters who would leave their cars longer than two hours and had no interest in shopping; and (b) the need for the scheme to be self financing or else to make a profit. They also noted that, since the penalty rule is an interference with freedom of contract, the courts should normally uphold what the parties have agreed where the parties are of comparable bargaining power and have been properly advised. As Lord Mance neatly stated: 

 ‘What is necessary in each case is to consider, first, whether any … legitimate business interest is served and protected by the clause and second, … assuming such an interest to exist, the provision made for the interest is nevertheless in the circumstances extravagant, exorbitant or unconscionable. In considering what is extravagant, exorbitant or unconscionable, I consider that the extent to which the parties were negotiating at arm’s length on the basis of legal advice and had every opportunity to appreciate what they were agreeing must at least be a relevant factor’ (at [152]). 

Lord Toulson used similar language to describe the test. Lord Hodge was of the same mind, but used slightly different (though broadly equivalent) terminology.[5] 

Application to service credits 

A failure to meet service levels will normally constitute a breach of a primary obligation. The secondary obligation is to pay service credits. 

If those service credits are designed to deter breach (were ‘in terrorem’) or were not a genuine pre-estimate of loss, there was previously scope to argue that these amounted to a penalty and should not be capable of being enforced. Cavendish Square/Beavis and the line of wider ‘commercial justification’ cases has done away with that distinction. Merely demonstrating that a service credit regime is intended to deter breach or the amount of service credits have been calculated without much thought given to any likely loss resulting from the service level failure will not render those service credits unenforceable as a penalty unless wider consideration is given to whether the service credits are imposed in order to protect some legitimate interest and they are unconscionable and extravagant.  

By way of example, prior to the Cavendish Square/Beavis line of cases, it was always conceptually difficult to accept the idea of ramping service credits. For example: 

Breach of service level x will result in £20 service credit on day 1

A further breach of service level x on day 2 will result in a service credit of £30

An additional breach of service level x on day 3 will result in a further service credit of £40. 

In this example the amount of the service credit could never have been a genuine pre-estimate because the same breach of service level results in a different service credit. Following the Cavendish Square/Beavis line of cases, it seems much more likely that such a service level regime will be upheld: 

1.     the parties have agreed it and, absent any inequality of bargaining position and abuse of it, the courts will give effect to the parties’ agreement; and 

2.     it is not unconscionable or extravagant.  

The customer may not have suffered any actual loss equal to £40, but the customer has a legitimate interest in ensuring that its IT systems function efficiently and in avoiding repeat failures for the broader benefit of its business. 

Likewise, many service level regimes look to deal with different types of service level failure. Some of these envisage accrual of different service credit points depending on the importance of the service level that has been breached, for example, availability of a contracted service (cf. the phone being answered on three rings). Others fail to distinguish between these service level failures, punishing all of them in the same manner. 

In the last case, under the old law prior to the Cavendish Square/Beavis line of cases, unless it could be demonstrated that these were failures of the same kind[6] or that the consequences of each failure are such as to make pre-estimation an impossibility, there was a risk of falling within the presumption set out in the third of Lord Dunedin’s tests for a penalty. 

Following the Cavendish Square/Beavis line of cases, it would appear more likely that such a regime will be upheld, particularly if the draftsman of the service credit regime includes a rationale for the differing treatment and includes an acknowledgement from the supplier that they consider such service credits are not unconscionable or extravagant and reflect the needs of the customer for a service free of accumulated minor defects. 

That said, circumstances can be envisaged where a claim for service credits could be unconscionable and extravagant eg the levying of service failure points in Mid Essex Hospital Services NHS Trust v Compass Group [2013] EWCA Civ 200. The relevant service credit regime in that case was based on price adjustment rather than liquidated damages and the application of the service credit regime was successfully challenged on the basis that the Trust had not observed the mechanism for calculation set out in the agreement. That said, some of the minor breaches were punished by the Trust with service deductions worth in excess of £584,000. These included breaches for failures such as having chocolate mousse out of date by one day, out of date bagels and fluff on the inside of the RCN notice board (which the supplier did not even have access to). These sound unconscionable and extravagant by any measure.  

Liquidated damages; exclusive remedy 

The second issue relating to service credits that exercises lawyers is whether the service credits should be the exclusive remedy. 

In the context of liquidated damages clauses for delay in construction and other contracts, it has been held that the nature and effect of the relevant clause will depend on its construction, but will normally operate as a limitation on the supplier’s liability[7]. Two judgments of Ramsay J (also in the context of building contracts) where liquidated damages have been linked to late delivery have supported this interpretation. In Chattan Developments v Reigill Civil Engineering Contractors [2007] EWHC 305 (TCC), Ramsay J noted: 

‘ … when there is a valid and enforceable liquidated and ascertained damages clause within an agreement, those damages are the sole and exclusive remedy for the particular breach to which they relate. ….Unliquidated damages are not recoverable because the parties’ agreement of liquidated damages replaces the remedy which would otherwise be available for breach…. The question of whether unliquidated damages could be recovered was a matter for the interpretation of the agreement from which it was possible to find a clear intention to exclude that remedy’. 

Similarly in Biffa Waste Services v Maschinenfabrik Ernst Hese Gmbh [2008] EWHC 6 (TCC), Ramsay J noted, ‘… the general position is that a liquidated damages clause in a contract covers ‘all the damages for non-completion’ or constitutes an exhaustive agreement as to the damages which are or are not to be payable by the contractor in the event of a failure to complete the works on time’. 

It is submitted that a similar approach should be taken with the interpretation of service credit clauses where service credits are expressed (or should logically be interpreted) as a form of liquidated damages. Normally, as a matter of construction and in the absence of any wording to the contrary, the service credits will be the exclusive remedy for the failure to meet the relevant service levels. But what if the clause expressly provides that other remedies are available? By way of example: 

 ‘In addition to the payment of service credits [where the service credits are expressed as liquidated damages] pursuant to clause [z], the Customer shall also be entitled to recover any costs associated with remediating the Supplier’s failure, including the [X] and [Y] costs and the right to terminate this agreement for cause ….’. 

In this situation the drafting seeks to make clear that the parties intend the liquidated damages are not the exclusive remedy. Two alternative interpretations of this clause are possible: 

  • If the liquidated damages already reflect a genuine pre-estimate of loss, there will be no additional damages to claim, so the additional right to recover costs will never kick in. Suppliers can take comfort from this, but would (understandably) rather not have this provision included. At the very least they should ensure that any liquidated damages paid are taken into account in respect of any general damages claim. 
  • The service credits are likely to be construed as a payment which occurs in addition to the actual losses arising from the breach of service levels. These will be at risk of challenge only if it can be demonstrated that there is no legitimate interest in the customer receiving these service credits and the amounts are not extravagant and unconscionable. 

Price Adjustment Mechanism 

Historically, some lawyers have had reservations with expressing service credits as liquidated damages. In particular they have been aware that: 

  • parties frequently do not have the rule against penalties in contemplation when they calculate the amount of service credits – customers do not usually investigate or take any steps to quantify the loss they could suffer and the worry has been that any figures inserted into the agreement are fairly arbitrary; and  
  • in many cases, failure to deliver an aspect of a service would not result in any definable loss or damage.  

A common way to mitigate these risks has been to express service credits as a price adjustment mechanism. The idea is that the price payable by the customer should flex depending on the supplier’s performance against service levels. The service credits reflect the different value for the services provided.  

Not surprisingly, customers’ lawyers have used this approach because of several perceived benefits: 

(a)     it was believed this would avoid the penalty rule altogether since the service credit is not rendered in circumstances of a breach of contract; and 

(b)   it allows for potentially larger service credits to be claimed which bear no relationship to any loss suffered as a result of failure to comply with a service level and instead are referable to differing levels of service. 

The Supreme Court’s decision in Cavendish Square/Beavis has meant that in most cases the draftsman will have a reduced risk of a service credit provision expressed as liquidated damages constituting a penalty, even where there is no associated loss. That said, use of the price adjustment mechanism may be appropriate where the amount of the service credits is intended to be more than merely compensatory, ie potentially larger service credits could still be phrased as a price adjustment and be enforceable. 

This last point does require some qualification however. In the Supreme Court, Lords Sumption and Neuberger considering criticism that the application of the penalty rule may depend on how the relevant instrument is phrased, noted (at [14-[15]) that it is the substance of the provision not its form or label that is important. In a short passage, they reserve the well-established equitable jurisdiction of the courts to overrule or ignore the apparent purpose or manner of working of a clause. Effectively therefore price adjustment mechanisms are also potentially subject to the rule against penalties.[8] Referring to clause 5.1 of the sale agreement with Cavendish Square, a clause that Lords Neuberger and Sumption categorised as a price adjustment provision, they concluded (at [77]) that: 

‘We do not doubt that price adjustment clauses are open to abuse and if clause 5.1 were a disguised punishment for the Sellers’ breach, it would make no difference that it was expressed as part of the formula for determining the consideration.’[9]  

Price adjustment: exclusive remedies 

Cavendish Square/Beavis looks also to affect the issue of the exclusivity of the remedies available.  

Prior to the Supreme Court’s decision many, in the absence of judicial authority on the point, had considered there were limitations to the use of price adjustment mechanisms. Where the parties have agreed an adjustment to the price based on differing levels of performance of the services, any failure to achieve the service levels will not amount to a breach of contract. The logical consequence is that the only remedy available to the customer is the service credit, not a general damages claim (because there has been no breach). Without specific drafting to cover the point, use of a price adjustment mechanism would risk the customer losing any opportunity to claim damages or to terminate the agreement for breach. Effectively the reduction of the price payable becomes the exclusive remedy available to the customer. I shall call this the ‘Pre-Cavendish Reasoning’. 

In Cavendish Square/Beavis, Lords Sumption and Neuberger offered the following obiter comments. Referring to clause 5.1 of the sale agreement to Cavendish, they stated (at [74]): 

‘It is not a contractual alternative to damages at law. Indeed in principle a claim for common law damages remains open in addition, if any could be proved. The clause is in reality a price adjustment clause’. 

In that case, there were two provisions: one consisted of a set of restrictive covenants and the other (clause 5.1) stated that an element of the consideration due in respect of the sale of the company would cease to be payable if the covenants were breached. Makdessi breached the covenants and the Supreme Court held that it would not oblige payment be made to him because the relevant clause was a price adjustment clause and was not abusive.[10] The logic of Lords Sumption and Neuberger appears to have been that breach of the covenants could have given rise to a loss that was greater than the amount that could have been provided for by the price adjustment. To the extent that that loss has not ‘abated’, to use the language of the court (at [76]), there remained a claim that could be brought, though they helpfully go on to say, ‘the issue does not arise [in this case] and was not argued’. 

It may be possible to distinguish the approach taken by Lords Sumption and Neuberger in Cavendish Square/Beavis. As noted above, in that case there were two separate provisions. Breach of the covenants clause would always give rise to a right to a free-standing right to damages. By contrast, where a service credit is phrased as a payment adjustment, the payment adjustment is usually specifically linked to the service credits due in that clause. On that basis, the Pre-Cavendish Reasoning set out above should still hold true.  

That said, depending on the construction of the relevant service credit provisions, it may still be arguable that a price adjustment mechanism in respect of service credits does not limit the damages or remedies available for breach. Fundamentally, the question of exclusivity of remedy becomes an issue of construction. 

A number of consequences flow from this: 

(a)   It is in both parties’ interests to be clear on the face of the document that any service credits phrased as a payment adjustment are or are not the exclusive remedy available for breach. 

(b)   A customer who is prepared to accept that service credits are the exclusive remedy will naturally wish to ensure that they adequately reflect the diminution in the quality of the services and are not merely trivial. In doing so, it will still need to be mindful of the court’s equitable jurisdiction to strike out clauses which are extravagant or unconscionable. 

(c)   If the service credits are not intended to be the exclusive remedy available, it is in both parties’ interests to be clear where the price adjustment begins and ends (ie it should be clear where poor performance constitutes a breach of contract). From a customer’s perspective, it will be important to be clear that there is a threshold or floor at which point the price adjustment mechanism ceases to apply and additional remedies kick in.  

 By way of example, the agreement could contain a trigger (either a single outage or persistent performance at a given level) or a threshold amount of service credits accrued over a defined period at which point the price adjustment mechanism would cease to apply and the customer would reserve the right to terminate and/or claim damages at large. In the absence of such a floor or threshold, there is a risk that the customer will find it impossible to demonstrate that the relevant performance fell outside the agreed price adjustment mechanism and constitutes an actionable breach. 

What if a clause is silent as to whether service credits constitute liquidated damages or a price adjustment? 

It is likely that, in the absence of a clause that spells out the basis on which service credits are claimed, the question will be one of interpretation using the tried and tested methodologies recognised by the courts.[11] This means that: 

  • If the clause is interpreted as providing for service credits on the basis of either liquidated damages or a price adjustment, the exclusivity of the remedy will also be one of construction.
  • In both cases, the clause may be struck down only if the amounts involved are ‘extravagant or unconscionable’. In all likelihood however regardless of the bases on which service credits are phrased, it is unlikely that the courts will wish to intervene, particularly where the parties have been properly advised and one party is not seeking to take advantage of another. 

Justin Harrington is a Partner in the Technology and Outsourcing Department at Blake Morgan LLP:




[1] As noted in this article, there are a number of apparent contradictions within the leading judgment of Lords Sumption and Neuberger with whom Lord Clarke agreed. The other judges gave similar but broadly concurring judgments (though Lord Toulson gave a dissenting judgment in respect of Parking Eye).

[2] Per Lord Halsbury in Clydebank Shipbuilding [1905]

[3] See Court of Appeal judgement of Clarke LJ in Cavendish Square Holdings v Makdessi [2013]EWCA 1539. Note however that in the Court of Appeal their lordships held in Cavendish that the relevant clause was penal, even after applying the commercial justification test.

[4] This section of the judgment gives rise to several questions. What is a ‘straightforward damages clause’. Having stated earlier that Lord Dunedin’s tests were not necessarily approved of by the majority, and have been excessively relied upon, it seems strange to then go on to say that they will in most cases be ‘perfectly adequate’. Does this mean that there are two tests or one, depending on whether the clause is intended to be compensatory or not?

[5] Lord Hodge referred to ‘exorbitance or manifest excess compared to the innocent party’s commercial interests’ (at [248]) or ‘exorbitant or unconscionable’ (at [255]).

[6] See Parker LJ in Dunlop at p 93.

[7] Temloc v Errill Properties 39 BLR, Court of Appeal

[8] In theory this would also be true of service credits phrased as a series of conditions precedent eg a payment of x will be made by the customer if conditions y and z are fulfilled, failure to meet y and z would not amount to breach, but failure to meet a condition precedent for payment. A service credit regime could be structured so that a base level payment will be paid for a standard level of acceptable services and additional increments (traditional ‘service debits’) paid for meeting improved levels of performance.

[9] Lord Mance by contrast (at [181]) sees price adjustment provisions as falling within the penalty rule. Lord Hodge similarly commented (at [227]): ‘There is no reason in principle why a contractual provision which involves forfeiture of sums otherwise due, should not be subjected to the rule against penalties, if the forfeiture is wholly disproportionate either to the loss suffered by the innocent party or to another justifiable commercial interest which that party has sought to protect by that clause’. It should be noted that Lords Hodge and Clarke were more open minded as to whether this clause amounted to a price adjustment, but nevertheless concluded that the rule against penalties applied to it (see [270] and [291] respectively).

[10] See [75]-[77]. The logic here is curious and circular. On the one hand their lordships appear to be suggesting that this clause is not penal because it is a price adjustment. On the other they suggest that even price adjustments are subject to abuse ‘if a disguised punishment of the Sellers’ breach’ ie if a penalty clause.

[11] See Lord Hoffman in Investors Compensation Scheme v West Bromwich Building Society [1997] UKHL 28.