Belgian contract law, like that of its neighbour countries, makes a basic distinction between obligations to produce strict results and obligations to use reasonable efforts.
In Book 5 of the New Civil Code ("NCC"), adopted on 21 April 2022, these obligations are defined in line with the view that has long been adopted in case law. The NCC has therefore not changed the applicable principles. The distinction exists in the common law and mixed legal systems as well, albeit with specific characteristics. With this caveat, we will refer in this article to “strict obligations” and “best-efforts obligations”.
In addition to these categories, an obligation can also be qualified as a guarantee commitment. This category is not included in the NCC as such, but is mentioned in the bill, where it is called a variant of the strict obligation. (Explanatory Memorandum, Parl.St. Chamber, 2020-2021, DOC No. 55, 1806/001, p. 83, under Art. 5.72.)
A guarantee is a strengthened form of the strict obligation. The debtor of a guarantee warrants the fulfilment of the commitment under all circumstances, including situations of force majeure. The debtor hereby assumes a conscious risk, which is not always within his control.
In the remainder of this text, we will start from the assumption that an IT service provider has taken certain commitments that can be regarded as strict or best-effort commitments, and that the customer can hold him liable as a result of (alleged) improper or late compliance (referred to as "default" in the NCC). After all, the IT service provider is obliged to perform its tasks properly and on time.
It should be noted that all obligations resting on the parties must be assessed individually, in isolation, as a strict obligation to achieve a result or an obligation to use one's best efforts. It is not a question of one overall commitment being assessed in this way . This does not prevent contracts or general terms and conditions from often stipulating that "all obligations" of a party are to be considered as obligations to use their best efforts or to achieve a certain result, possibly with the proviso "unless expressly agreed otherwise for specific obligations".
As previously stated, the distinction between the two categories of obligations determines when there is a contractual liability and also determines the burden of proof of a contractual liability, for example, where the customer claims that an obligation has not been properly or timely performed.
Where there is a strict obligation, the service provider is obliged to achieve a specific result. This is an appropriate way of contracting when there is a sufficiently clear assignment, which can be assessed in detail and is not too complex or experimental. After all, the service provider can then properly assess what the result should be and has the fulfilment of the task mainly in its own hands. On the other hand, a strict obligation to produce a result is not suitable when its correct execution is strongly dependent on external factors, such as the customer’s own cooperation, or when the requested result is still too abstract and can only be made sufficiently concrete during the course of a project.
When he is bound by a strict obligation, the service provider is in principle ipsofacto in default if he cannot achieve the agreed result of his commitment, or cannot achieve it within the agreed timeframe (milestone), unless he can prove force majeure. The question of whether the service provider has tried with sufficient care to achieve that result or deadline is irrelevant here. The customer only has to prove that the promised result was not achieved or not achieved on time. A presumption of default then applies, which can only be rebutted by proof of a force majeure situation.
Force majeure means that the obligation was not performed correctly or on time due to an external cause not attributable to the service provider. These can be the classical situations of force majeure such as natural disasters, war, boycott, epidemics and the like, but also failures of the customer himself or of third parties for which the service provider is not liable (with the exception of subcontractors). When the service provider invokes force majeure, it should be examined whether he did not have the duty to anticipate and prevent certain force majeure situations. For example, certain IT contracts require the service provider to build in disaster protection, and/or supervise involved third parties, co-ordinate the actions of the customer, etc.
If the service provider does not strictly promise the result to be achieved, there is a best-efforts obligation. The service provider is then obliged to make all "normally required" efforts to achieve a certain result. The touchstone is the behaviour of a normally careful service provider in similar circumstances. Thus, failure to achieve the intended result does not mean that the service provider is ipsofacto liable for a contractual default. The customer must prove that the service provider did not act as a normally prudent service provider.
The customer will therefore have to prove a professional failure, which may be a very difficult task depending on the circumstances, as the customer often lacks the specialised knowledge and experience to assess the facts of the case. In practice, this proof usually boils down to an amicable or judicial expert examination in which a specialised third-party expert is entrusted with an assessment of the facts. If, due to the nature of the facts, the customer is unable to provide proof with certainty, it may suffice to provide proof of the probability of a fact (art. 8.6 NCC). Exceptionally, the court may also reverse the aforementioned burden of proof (art. 8.4 NCC). In any case, it is clear that the law of evidence is more favourable to the service provider when an alleged default is assessed within the framework of a best-efforts obligation than within the framework of a strict obligation.
In the context of disputes relating to IT contracts, the classification of obligations as strict or best-efforts obligations is regularly discussed. Indeed, this classification can have a decisive impact on a dispute. This mainly concerns the question of whether certain tasks must be carried out strictly within a set period of time (milestone), or whether certain service levels must be strictly adhered to, with only force majeure being a legal excuse.
When a judge is confronted with a dispute in which the nature of an obligation has to be determined, he will have to ascertain the parties' common intention by interpreting the contract and/or the facts. Several hypotheses are conceivable.
When the parties lay down the qualification contractually, it remains curious that it is an explicit legal qualification that the parties have given to the commitments, and it is possible that they have misjudged that legal qualification. This can lead to problems when an obligation that is very abstract or uncertain, or very dependent on third parties, is nevertheless explicitly defined as a strict obligation. In principle, it is said that the explicit expression of the intention of the parties is decisive for the qualification.
It is sometimes thought that the fact that services are performed at a fixed price or on a “time and materials” basis would have an impact on the qualification, but that is incorrect. However, the vagueness or abstract nature of the specifications may have an impact on the assessment. When certain requirements are still vague and will be specified during the course of the project, this will rather point to a best-efforts obligation. In the absence of sufficiently precise requirements, no specific result can be promised. Conversely, when the requirements are fixed in a strict and detailed way, this can be an indication of a strict obligation, but other factors, such as a strong dependence on third parties, should be taken into account.
The sanctions that the parties have provided for in case of non-fulfilment of certain commitments can also give an indication of the qualification of these commitments. If the parties have included penalty clauses for exceeding the milestonesof a project, this may indicate a strict obligation. In the same sense, the application of service credits for failure to meet the required service levelsof an SLA may indicate this.
An important element in assessing the nature of the commitment, especially where the parties have given no other indications of interpretation, is the alea. Where the result sought to be achieved depends on a (material) aleatory element, a (material) element of uncertainty, there istraditionally a best-efforts obligation, because it is then presumed that a service provider would not seek to commit himself to a specific result .
For example, the execution of a commitment may depend on the actions of certain third parties who have to cooperate in due time, or on third-party products with which the result of the commitment has to be integrated. If these third parties are not subcontractors of the service provider, the service provider is not responsible for them. Moreover, the service provider's commitment may strongly depend on the cooperation of the customer itself, who must cooperate in a timely manner in the progress of a project, including providing timely information, taking decisions, performing tests, etc.
If, on the other hand, the court finds that the service provider had a result (largely) in its own hands, without external risks, the court may consider it to be a strict obligation . This may be the case for simple obligations that involve little or no risk, both in terms of the state of the art and in terms of a possible impact of external factors. The question is then towhat extent theabsence of external factors should be considered. Commitments will always contain an element of uncertainty and depend somehow on the attitude of the customer. This is difficult to delineate with a general rule. Absolute certainty about the chances of success of the promised result is not required to qualify a strict obligation. The more uncertain the performance of the agreed commitment, the more likely it is to be assessed as a best-efforts obligation.
The assessment of an alleged contractual default can become very difficult if an obligation is indeed qualified in a contract or in general terms and conditions as a strict obligation, but in reality the achievement of the result of the obligation is very uncertain, for example if it is heavily dependent on the cooperation of the customer itself or of third parties. The qualification given in the contract is a legal qualification, not a factual one. In that case, the parties have used an ill-considered formulation. In that case, it is still considered that the expressed intention of the parties remains the very first criterion for the assessment .
The NCC states that the court can only re-qualify a qualification given by the parties to their contract if the qualification is incompatible with the terms of the contract or with mandatory rules or rules of public order (art. 5.68 NCC). This rule is more concerned with the qualification of a contract as a whole, which may be contrary to the operational reality as reflected in certain contract clauses, which often occurs in order to circumvent certain legal rules. Nevertheless, it seems to us that this provision also confirms the limited authority of the court to re-qualify obligations. If the judge does not proceed to such re-qualification, the only way out for the service provider is to invoke force majeure (cfr. supra).
A Service Level Agreement or "SLA" is the strict obligation par excellence in the context of IT contracts. An SLA includes very precise and quantified commitments (service levels), which are constantly measured against indicators. Performance is then measured (usually on a monthly basis) against detailed levels. Although this simplifies matters somewhat, the credo of the SLA is: "If you can't measure it, you can't sanction it". If a quantified commitment cannot be achieved, the service provider commits a breach of contract. He can then only invoke force majeure.
Aleatory elements of the service are often already taken into account when defining the indicators of the service levels. For example, in many SLAs a distinction is made between the promised responsetime and the restore time. When an incident occurs in the IT service, the service provider promises to provide feedback to the customer and to start the restore process within a set period of time. The maximum timeframe is determined according to the severity and impact of the incident in question. This response time is entirely in the hands of the service provider's service desk, and is almost always regarded as a strict commitment. For the restore time, the time required to actually resolve the incident and restore the service, this may be more difficult, depending on the complexity of the incident and the environment in which it occurs.
When it comes to complicated processes, where software and/or systems can be highly dependent on other software or systems, and/or where different environments and different players are involved, the diagnosis may take more time and the recovery time may be difficult to promise in general figures. Therefore, the recovery time is often explicitly qualified as a best-effort obligation, which implies that the service provider must act carefully and professionally with regard to the recovery, and must pursue the requested levels in a professional manner, and this in function of the urgency and the impact of the incident. In addition, assumptions are often included in SLAs, which may limit the aleatory element of the service.
Finally, it should be noted that some SLAs deviate from the basic philosophy mentioned above. Thus, it may happen that all commitments are qualified as best-efforts obligations.
More and more, IT projects are carried out in an "agile" manner. This means that at the start of the project, the requirements are formulated in a rather vague and abstract way, and they are gradually refined and filled in by the service provider and the customer. This is done on the assumption that there will be intensive cooperation between the service provider and the customer. The end result cannot be specified from the start and is therefore uncertain. Apart from the vagueness and the dynamic, evolving and flexible (possibly even experimental) nature of a project, external factors (not least the customer’s own attitude) may influence the outcome. Such projects usually imply that the service provider is bound by best-efforts commitments. He must strive for the best possible outcome in the given circumstances, like a good professional.
In the context of agile projects, it is strongly recommended to deal (albeit partially) with the uncertainty of the end result by means of an appropriate arrangement for the acceptance of what is being completed. In such projects, it is therefore best to model the contractual framework as much as possible with well-reflected contract terms so that the operational reality is correctly taken into account. When such a project fails and the parties accuse each other of negligence, the court must take into account the interaction that took place between the parties (or that did not take place where it was required). Certain external factors that prevented the commitments from being properly performed may then be categorised as "force majeure".
When parties draw up IT contracts, they should always take sufficient account of the operational reality and not limit themselves to common style formulas. The roles of the parties, the degree of competence of the service provider and of the customer, the purpose and clarity of a project, the experimental nature of the activity, the dependence on external factors, the flexibility required or not required in the development of possible solutions, are all factors that must be taken into account in order to draw up a realistic contract. They also strongly influence the required intensity of the parties' commitments in the performance of the contract.