Construction of Commercial Agreements in an IT Context

April 30, 2005

The recently reported decision of the Court of Appeal in Vogon International Limited v Serious Fraud Office [2004] EWCA Civ 104 concerned a dispute which arose because one word in a quotation for forensic data recovery work – namely the word ‘database’ – had been interpreted by the respective parties in two entirely different ways leading to a considerable monetary difference between the two competing constructions.



In this article we examine some of the lessons to be learnt from the Vogon case about the interpretation of commercial agreements in an IT context. The theoretical approach of the courts to the interpretation of commercial agreements has been well rehearsed in a number of cases over recent years, most notably in the speech of Lord Hoffman in Investors Compensation Scheme v West Bromwich Building Society [1998] 1 WLR 896 (see in particular p 912). The theoretical principles, however, seldom shed light on the approach adopted by the courts in practice. This article is not intended to be an exhaustive treatment of the rules of interpretation and how they are applied, but is instead intended to give guidance as to the courts’ approach to construing commercial IT agreements in practice.



The Vogon Case



Vogon International offers services including forensic data recovery. In the course of a criminal investigation, the SFO engaged Vogon to recover e-mail data from 33 tapes. Each of the tapes contained backups of data originally held in Microsoft Exchange databases on two e-mail servers. The process involved Vogon assessing the number of databases on the tapes and then configuring a Microsoft Exchange database, populating it with the data from the tapes and recovering individual Personal Store Files (PSTs) from the bulk data files. The 33 tapes contained a total of 49 backups, of which Vogon was ultimately able to process 39.



Prior to the formation of the contract Vogon had provided the SFO with an estimate of what the cost of the data recovery work would be. This estimate referred to an initial set-up charge of approximately £2,000 and a further charge of around £500 for each backup.



Vogon subsequently provided the SFO with a quotation accompanied by a covering letter. The covering letter included the words “Note that at present there only appears to be backups of one database however we cannot guarantee this without carrying out the work.”.



The material part of the quotation itself provided:



“.Set -up and population MS Exchange Database


£1,500 per database.



To process the data from each database to produce one PST file per mailbox


£1,250.00 per database.



Vogon contended that it was entitled to charge for processing 39 backups, on the basis that the words “backup” and “database” were synonymous. The SFO, on the other hand, argued that the word “database” in the quotation referred to the Microsoft Exchange databases on the e-mail servers, of which there were only two. In financial terms the determination of the meaning of the word “database” in the quotation meant the difference between SFO having to pay Vogon £22,500 plus VAT (on the SFO’s case) or £314,375 plus VAT (the sum for which Vogon invoiced).



At first instance HHJ Seymour QC preferred the SFO’s construction. He stressed that the Court’s function was not to determine what the term “database” was capable of meaning but what the parties to this particular contract had intended it to mean. In doing so he held that the phrase “per database” in the quotation had to be read back to refer to the database in respect of which the activity specified was to be carried out, namely “MS Exchange database”.



Judge Seymour also considered the words in the covering letter to the quotation to be significant and to support his conclusion that the parties were referring to MS Exchange databases. His decision placed considerable reliance on the contents of this letter. In particular, the reference to there being no guarantee that the backups were of only one database supported the SFO’s construction. The sentence would have been irrelevant if the applicable fee was to be determined by reference to the number of backups as opposed to the number of MS Exchange databases.



Judge Seymour also had regard to the fact that the work had taken Vogon a total of 20 man days to complete and stated that bearing in mind the approximate cost set out in Vogon’s intial estimate (which was more consistent with SFO’s construction than that of Vogon) the amount Vogon claimed was excessive for ten days’ work carried out by two individuals.



In addition the judge stated that the time estimate provided by Vogon in its covering letter to the quotation supported his construction. This estimate envisaged five working days to set up the server and the database and an additional two to three working days to process each database. These time estimates only made sense if the parties intended “database” to mean each of the two MS Exchange databases rather than the numerous backups.



The decision and reasoning of Judge Seymour on the issue of the construction of the contract was unanimously upheld by the Court of Appeal. Although the case did not involve either court in a lengthy analysis as to the general approach that ought to be adopted in construing commercial IT contracts, the case is an apt reminder of some of the applicable principles and, in particular, of the imperative of reaching a result that reflects commercial reality.



The Use of Expert Evidence in Contractual Interpretation


IT contracts, perhaps more than any other type of contract, often contain references to specialist and technical terms. It appears to be established that in cases where there is a dispute between the parties as to the meaning of a technical term, not only is the court permitted to use expert evidence for assistance, but it ought to do so. The case of Baldwin & Francis Ltd v Patents Appeal Tribunal [1959] A.C. 663 concerned the construction of a technical specification (see also Cooper (Max) and Sons Pty Ltd v Sydney City Council (1980) 54 ALJ 234 at 239, PC, per Lord Diplock). In Baldwin Lord Reid said:



Technicalities crop up in a great variety of cases and, when the parties are not in dispute about them they are often explained without evidence.If the court finds that there is a genuine dispute and real doubt about any technical matter, then I do not think that it can proceed farther without evidence”.



In Vogon both parties had called expert evidence to shed some light on how the term database is understood in the IT world. Interestingly, neither court relied on the evidence concerning the way the term is used in the industry. The expert evidence had indicated that, although “database” has no single and invariable meaning, it is generally recognised as being a structured or organised store of data, held within a computer system in such a way that the data can be accessed or manipulated in the form in which it is stored. In other words there is no need for the data to be restored in order for it to be accessed and manipulated. However a backup, it was agreed, is merely a snapshot copy of the data held on the database at the time the snapshot is taken. Unlike the data on the database itself, the backup data is not accessible, and cannot be manipulated, unless the backup is restored.



The courts could legitimately have paid more attention to the expert evidence in Vogon. Although the term “database” is often used loosely and can mean different things in different contexts, there are a number of uncontroversial criteria that must be satisfied for a collection of information to be described as a “database”. Judge Seymour was correct in saying that the question that had to be decided was not what the word was capable of meaning but what it meant in the context of this particular contract. Nevertheless, Vogon was a specialist IT contractor and the SFO had its own IT experts involved in the investigation. As such, the contract was made between IT-literate parties and it would have been open to the courts to have had more regard to the expert evidence in supplying reasons for its decision.



The Subsequent Conduct of the Parties


There can be little doubt that it is generally not permissible for a court to look at the subsequent conduct of the parties when it is interpreting a written agreement, unless that subsequent conduct amounts to a variation or gives rise to an estoppel (Wickman Tools v Schuler A.G. [1974] AC 235; HL; see also Lewison, The Interpretation of Contracts, 3rd edn, § 3.12).



In Vogon, as we have mentioned, Judge Seymour had regard to the fact that the work done by Vogon had only taken two people ten days to carry out. He used this as part of his reasoning in finding that the SFO’s construction was to be preferred. Vogon’s claim for £314,375 plus VAT was regarded as being excessive for 20 man days of work. Lord Justice May in the Court of Appeal agreed with this reasoning.



Strictly speaking, a stringent adherence to the rules of construction would have precluded the courts from looking at the time it actually took Vogon to carry out the work post-contract. As Lord Parmoor said in Union Insurance Society of Canton Ltd v George Wills & Co [1916] AC 281, “It is immaterial to the construction of the contract to consider subsequent events. The intention of the parties must be gathered from the language of the contract, the subject-matter, and the circumstances in existence at the time it was made.” (emphasis added). However, by looking at the time Vogon had subsequently taken to carry out the work, the courts were able to reach a decision that made commercial sense and that prevented the SFO from having to pay excessive rates for the work done. In assessing the plausibility of the competing interpretations, it would have been unrealistic to divorce the actual time taken from the assessment of the price, as it was only by reference to the time and manpower needed to get the work done that the courts could come to a view on which price was the more realistic.



Giving Meaning to All Parts of the Contract


As a general rule, the parties to a contract are to be taken to have intended that each word they use is there for a reason. There is therefore a presumption “towards treating words as adding something, rather than as mere surplusage” (S.A. Maritime et Commerciale of Geneva v Anglo-Iranian Oil Co. Ltd [1954] 1 WLR 492; per Somervell LJ). Notable situations where this presumption is weaker include charterparties, leases and standard form contracts where the bank of clauses has been built up over a period of time (Lewison, § 7.03). However in a “one-off” contract such as that in Vogon, where the contract documents had been especially prepared for the particular circumstances, it seems that the courts will try to give effect to all parts of the contract.



Vogon’s covering letter to its quotation had referred to the SFO’s tapes containing “backups of one database”. The fact that Vogon had itself drawn this apparent distinction between a backup and a database was significant as it indicated that the two concepts were different. This therefore tended to support the SFO’s construction because Vogon’s interpretation (which effectively treated a backup and a database as being synonymous) would have required part of its own quotation to be ignored entirely.



In his judgment, Lord Justice May also made reference to Vogon’s first estimate of the cost of the work. This gave an estimated cost of the work on a “per backup” basis as opposed to “per database”. However, having regard to the number of backups there were, this would have produced a total price close to that contended for by the SFO. It was therefore far more consistent with the SFO’s construction than with Vogon’s. Finding for Vogon would have involved disregarding the effect of this document.



Need for a Commercially Sensible Result


Perhaps the most important lesson to be gleaned from Vogon is that although the traditional canons of construction are important in interpreting commercial contracts, the courts will be reluctant to uphold a construction that confounds common sense or that leads to a commercially unrealistic result. The principle was encapsulated by Lord Diplock in The Antaios [1985] AC 191 at 201:



“I take this opportunity of re-stating that if a detailed semantic and syntactical analysis of words in a commercial contract is going to lead to a conclusion that flouts business commonsense, it must be made to yield to business commonsense.”



A similar approach was adopted more recently by Lord Steyn in Mannai Investment Co. Ltd. v Eagle Star Life Assurance Co. Ltd [1997] AC 749 at 771.



The contract therefore has to be construed within the framework of its commercial purpose, provided this is the commercial purpose of both parties (Bank of Nova Scotia v Hellenic Mutual War Risks Association [1990] QB 818; CA), and the courts will construe the document on the basis that commercial parties intend a commercially understandable result. Vogon is one of the clearest recent examples of the primacy afforded to this principle by the courts.



The evidence at trial had revealed that the purpose of the contract had been for Vogon to carry out the work within a ten-day deadline. Indeed an uplift in Vogon’s fees had been agreed for the work to be carried out within this deadline. Bearing this fact in mind, the contract only made business sense if “database” meant “MS Exchange database” as opposed to “backup”. This was because the covering letter to Vogon’s quotation had referred to set-up of the initial database taking five days and then each subsequent database taking two to three days each to process. This only fitted in with the envisaged ten-day time scale if there were no more than two databases. Had the parties intended “database” to mean “backup” (of which there were known to be a large number) it would have been impossible to carry out the work within the deadline. As a result the commercial purpose of the agreement would have been frustrated.



This principle can also be seen at work in a bare analysis of the different contract prices that resulted from the parties’ respective constructions. The courts adopted an interpretation that produced a price for the work that was considered to be broadly in line with the market rate as opposed to one that would have created an excessive price, several times greater that that shown by the expert evidence to have been reasonable (May LJ at [22]).



Interaction of Rules



In Vogon one can therefore see in practical terms the interaction of a number of principles governing the interpretation of commercial contracts in an IT context. The decision is indicative of the modern approach adopted by the courts, in which the traditional rules of interpretation are not applied dogmatically but rather are used as guidelines to lead to a commercially realistic outcome.



Duncan McCall and Thomas Crangle are barristers and members of the IT Group at 4 Pump Court, Temple, London, EC4Y 7AN. www.4pumpcourt.com