Output-based IT Outsourcing Projects

October 18, 2009

The recent decision of his Honour Justice Byrne in Ipex ITG Pty Ltd (in liq) v Melbourne Water Corporation (No 5) [2009] VSC 383 (11 September 2009), Supreme Court of Victoria involved IT service provider Ipex suing its customer in respect of misrepresentations that were allegedly contained in Melbourne Water’s pre-contractual tender documentation.  The case was brought under ss 51A and 52 of the Trade Practices Act 1974 with Ipex alleging that the misrepresentations constituted misleading and deceptive conduct.

Summary of allegations

The allegations concerned certain information provided by Melbourne Water as to the number of calls for assistance which its IT help desk had received over the six-month period prior to it publishing its invitation to tender (ITT) and request for proposal (RFP).  The information was, according to Ipex, inaccurate and incomplete and misled Ipex about the number of staff that it would have to deploy in order to meet the demand placed on the help desk (the help desk being within the scope of the services outsourced to Ipex).  Ipex said that, as a result of being misled, it incurred costs and expenses which were not taken into account when it calculated its bid price and which Melbourne Water subsequently refused to pay.

Summary of decision

The traditional analysis of a decision of this nature would be to summarise and comment on: (a) each of the allegations, (b) the judge’s findings of fact, and (c) the application of the law to those facts.  However, the bottom line is that Ipex failed on all counts. The judge held that either (a) the alleged representation in question had not in fact been made, or (b) if it had been made, it was not proven by Ipex to be false, and/or (c) in any event, it did not mislead or deceive Ipex because it didn’t rely on it.

Focus on features of output-based outsourcing projects

Accordingly, although I will summarise the alleged misrepresentations, the true importance of this decision for the IT community is to identify why Ipex lost the case so comprehensively.  In my opinion, the fundamental reason is that this case relates to an output-based outsourcing project, not a traditional input-based project, and Ipex simply did not appreciate the important features of output-based IT projects which distinguish them from input-based projects.  The case therefore highlights the critical need for IT service providers to appreciate and fully understand the nature and legal implications of these types of projects. The key attributes of output-based projects (if properly structured and if the contract is properly drafted) are that:

• The customer expresses its requirements in terms of its business requirements and in terms of outputs (or outcomes) – the customer is the master of its business, which it knows better than anyone, and it must articulate these requirements fully and unambiguously to the service provider – the customer therefore mandates what it wants.

• Solution risk is transferred to the service provider – the service provider is the IT expert and must take responsibility for designing and delivering its IT solution, which should not be the customer’s risk – the service provider therefore decides how it is going to meet the customer’s requirements – importantly, the customer should (ideally) not stray into prescribing any aspect of the solution.

• The charging mechanism is based on the service provider meeting the customer’s business requirements, so the customer only pays if its requirements have been met (this is the best way of demonstrating and receiving value for money).

• The service provider is involved directly in managing (and transforming, where necessary) the customer’s business processes, through its control of the IT systems, in order better to achieve the customer’s requirements.

• The service provider typically participates in the success of the customer’s business within an agreed risk-reward incentivisation framework.

Features of the Melbourne Water project

Justice Byrne noted that ‘a feature’ of Melbourne Water’s RFP was that it was expressed in terms that the tenderers were required to provide ‘outcomes’ (or ‘outputs’).  How the outcomes were to be achieved was, to a very large extent, a matter for each tenderer.  This is entirely consistent with the statement above that ‘solution-risk’ is transferred to the service provider under an output/outcome-based outsourcing contract.
Another ‘important feature’ identified by Justice Byrne was that the tenderers were required to submit a lump sum price for providing the outcomes. While there was provision for an adjustment of the price in the event that the scope of work was varied, the wording of the specification was such that a change of this nature would be unlikely.
Ipex was therefore required to assume most, if not all, of the risks involved in providing IT services for Melbourne Water over the three-year term of the project and, logically, Ipex was therefore required to include in its tender price an allowance for these risks (or take the risk of not doing so).  This is a key feature of an output-based project.
I want to highlight how the key features of an output-based outsourcing project influenced the decision that the information that Melbourne Water had provided was neither misleading nor deceptive and that Ipex had no cause for complaint or compensation.

Specific allegations and misrepresentations

The misleading and deceptive conduct in question was based on six alleged misrepresentations concerning or arising from the number of calls to Melbourne Water’s IT help desk during the six-month period prior to it issuing the RFP. This information was set out in the Help Desk Calls Table (HDC Table) which was included in the RFP.  The six misrepresentations can be further broken down:
1. two were said to be made expressly made in the HDC Table;
2. two were alleged to be implied (ie from the other express representations in the HDC Table); and
3. two were alleged to result from Melbourne Water’s silence (ie withholding or not including relevant information in the HDC Table).

Express misrepresentations

The two express alleged misrepresentations were, in essence:
1. that the average number of monthly calls between October 1999 and March 2000 was 433; and
2. that the additional information as to the categories, numbers and type of calls was accurate.

Implied misrepresentations

The two implied misrepresentations were, in essence:
1. that the average number of help desk calls would be and would continue to be approximately 433 per month during the period of the three-year contract; and
2. that the information was a representative sample sufficient to allow Ipex to cost its proposals on a proper basis.
With respect to the first of these implied representations (which was of a future fact), Ipex claimed that Melbourne Water did not have reasonable grounds for making the representation and, therefore, the representation should be taken to be misleading (relying on s 51A of the Trade Practices Act). Furthermore, Ipex positively alleged that Melbourne Water knew, or ought to have known, that the average number of help desk calls during the term of the project would in fact be significantly greater than 433 per month.

Misrepresentation by silence

Ipex also alleged that Melbourne Water made two further misrepresentations by its silence:
1. withholding relevant information and consciously not disclosing it; and
2. not responding to Ipex’s express requests for further specific information at a tender meeting.

Consideration of each alleged misrepresentation

Express representations: average number of calls per month to the help desk and categories, numbers and types of calls

His Honour dealt with both express misrepresentations together.
The figures for total calls in the HDC Table varied from month to month from a low of 364 in March 2000 to a high of 527 in November 1999, but the average was 433 per month. The evidence was that, from July 1997 to September 1999, the number of calls varied from a high of 980 in August 1997 to a low of 475 in December 1997. Although the general trend of the total number of calls per month was downward, the average over this period of 28 months was 723 per month. During the project, Ipex in fact experienced help desk calls at a greater monthly rate, reaching a maximum of 1,448 calls in March 2002 and averaging 675 per month over the three years of the project.
On all of this evidence one might expect that Ipex had a solid argument that the HDC Table was misleading.  Ipex complained about the disparity and sought payment for the extra costs involved in servicing this number of help desk calls as a variation in scope pursuant to clause 15 of the contract. Ipex’s claim was rejected by Melbourne Water on various grounds but principally because the contract was a lump sum contract and Ipex bore the risk that its input costs may escalate (except if Melbourne Water’s requirements changed – which they hadn’t as they still wanted calls to their help desk to be answered).
His Honour noted that the HDC Table made it clear that the figures related only to the specific six-month period from October 1999 to March 2000.  The fact that the average calls per month both before and after that period was higher did not establish that the information in respect of that six- month period was wrong.  Furthermore, although he noted that the figures for October 1999, January and March 2000 might have been ‘understated,’ they were still ‘within the bounds of tolerance’ which a tenderer would be comfortable to act upon.
Ipex simply did not have a ‘smoking gun’ to prove that the figures for the stated period were wrong.  Another key aspect of an output-based project is that the service provider, knowing and appreciating that its customer is transferring solution-risk to it, must undertake due diligence (or take the risk of not doing so), in particular in relation to price-sensitive matters.
His Honour therefore held that Ipex had failed (the onus being on it) to prove that these alleged misrepresentations were in fact false and, even if they were not 100% accurate in all respects, they were certainly not misleading.

First implied misrepresentation: Future numbers of help desk calls

The first of the implied misrepresentations was that Melbourne Water, by providing the historic figures of help desk numbers, made an implied representation that the average number of help desk calls would be, and would continue in the future to be, approximately 433 per calendar month during the period of the contract. This allegation was based on what Ipex described as ‘a notorious industry practice’ that the number of help desk calls over a period of time remained constant providing there was no change to the IT environment.  The fact that Ipex acknowledged this proviso is crucial.

Relying on the number of calls in the 28-month period from July 1997 to September 1999 (723 per month), Ipex argued that Melbourne Water had no basis for making this implied representation.
The relevance of the proviso to this alleged misrepresentation is that output-based projects are not about ‘minding the keys’ or keeping the IT infrastructure ‘ticking over.’  The customer wants its service provider to be a pro-active manager of its IT systems in order to improve its business processes and its competitiveness and to deliver cost benefits. If Ipex seriously thought that the business and/or the IT environment would not change in the future then it had misunderstood the nature of the project and its role in the project from the outset.  His Honour found that there was nothing in the RFP to lead Ipex to believe that the Melbourne Water IT environment would remain constant over the three-year term of the contract and, to the contrary, the RFP itself warned the tenderers that they must accommodate future changes in a ‘dynamic environment’. Accordingly, his Honour concluded that this misrepresentation wasn’t in fact made and could not be implied.
Furthermore, even if one could extrapolate a representation to the effect as alleged, his Honour considered that Melbourne Water would have had reasonable grounds for making it.  The expectation of Melbourne Water was consistent with and reflected the expectation of all customers that award a tender under an output-based structure, ie they expect the new IT service provider to examine critically the current IT environment and recommend and make improvements.  This was indeed one of Ipex’s functions under the RFP. The customer’s not unreasonable expectation is also that the new provider will perform better than the incumbent legacy provider (where there is one).  In these circumstances, it was not unreasonable for Melbourne Water to expect that the number of calls would reduce down to a number no greater than that shown in the HDC Table and, hopefully, less.  His Honour noted that Ipex indeed provided encouragement for this expectation in the material which it provided in the course of the tender process.
The judge went on to consider the question whether, if in fact the implied misrepresentation had been made, Ipex relied on it.  It is well settled that a misrepresentation that is not in fact relied on cannot, logically, mislead or deceive.  In answer to this question, his Honour was satisfied that the raw information about the help desk calls was used by Ipex in formulating its bid price but not on the basis of the alleged ‘notorious industry practice’ that the numbers do not change.
To summarise, his Honour held that the alleged misrepresentation wasn’t made; but if it was, there were reasonable grounds for making it; but if there weren’t, Ipex didn’t rely on it in any event!

Second implied misrepresentation: the HDC Table contained a representative sample

Ipex argued that, because Melbourne Water expected that the tenderers would calculate their tender price from the information contained in the HDC Table, it impliedly represented that the information was proper and sufficient for that purpose. Relying again on the number of calls from July 1997 to September 1999 (723 per month), Ipex submitted that the monthly figures in the HDC Table for the six months from October 1999 to March 2000 were obviously not a representative sample, they were not ‘proper and sufficient’ and they misled Ipex when it calculated its costs.
His Honour was satisfied that the information in the HDC Table was intended for the assistance of tenderers in their task of preparing their tenders. He also noted that a substantial scope of the services involved the handling of help desk calls during the period of the contract.  However, a feature of an output-based tender process is that the customer, Melbourne Water in this case, allows (indeed should require) tenderers to approach the achievement of the stipulated objectives in their own way and to introduce innovation and change where they consider it to be cost effective. Changes of this kind are clearly likely to affect the number and type of help desk calls both in the short term and in the long term. In the short term the numbers might increase; in the long term, innovative changes should produce greater efficiencies and reduce the number of calls.
Accordingly, his Honour held that the number of calls in the HDC Table did not convey an implied representation that these were representative numbers that were ‘proper and sufficient’ for the purposes of calculating a tender price. 
The fundamental flaw in this case was that Ipex’s costing methodology used a traditional input cost basis and it was therefore inappropriate for an output-based project.  The evidence was that Ipex used a costing methodology that it had developed for an earlier project (the G8 tender).  However, there was an important difference between the basis for the G8 tender and the Melbourne Water tender. The G8 tenderers were provided with the general number of help desk calls upon which they should base their cost and extra calls would warrant a price adjustment, ie the customer in the G8 project did not transfer solution-risk in excess of the stated number of calls, quite the opposite, the risk reverted to the customer.  As already noted, the Melbourne Water tenderers were required to build the risk of extra calls into their price. This meant that they were required to make their own assessment of the likely demands upon the help desk over the three-year contract period. His Honour emphasised that the philosophy underlying the Melbourne Water tender was that the successful tenderer was to achieve an output (or outcome) in terms of the extensive performance requirements set out in the RFP. Assuming that the numbers of calls shown in the HDC Table was to be taken as a representative sample of the experience of Melbourne Water upon which the tenderers were to rely, it could (and should) therefore only be relied upon as indicative of the existing position. How this might be used for the purpose of predicting the future was a matter for the tenderers. In this sense, it was therefore sufficient for them to perform this task.
As to the question whether, if the implied misrepresentation had been made, Ipex relied on it? The evidence did not support such a finding and therefore this allegation also failed for want of reliance.

First misrepresentation by silence – withholding earlier call numbers

Melbourne Water accepted that it did not disclose to (and consciously withheld from) Ipex additional information about help desk calls which it had received from July 1997 in monthly reports provided to it by its legacy IT services provider.  Ipex submitted that the failure to disclose the prior help desk calls history meant that the limited information in the HD Calls Table was misleading, ie it contained only half the story.
On the evidence, the period prior to October 1999 was all but irrelevant because of changes to Melbourne Water’s IT environment (and help desk requirements) that resulted from carrying out another project.  It was therefore not misleading to withhold this earlier (and irrelevant) information.  Indeed, arguably, it would have been misleading to have done so.

Second misrepresentation by silence: Not responding to requests for information

The final allegation was based on the failure of Melbourne Water to provide information as to help desk calls sought by Ipex at a pre-tender meeting. It was accepted by Melbourne Water that certain requests for information were not answered at the meeting and, although they were deferred to another meeting, they were not later responded to.
Insofar as the various requests were relevant to help desk calls (which was not clearly the case), Melbourne Water chose not to provide the requested information because they took the view that it would be unhelpful for the tender process if tenderers were provided with the experience of the legacy provider. This would arguably lead them to follow the ‘old’ approach to providing IT services and the intent of Melbourne Water was to encourage tenderers to propose innovative solutions (a hallmark of output-based projects).
His Honour accepted this explanation as being consistent with the inherent nature of the project and consistent with other aspects of the RFP.  Accordingly, this allegation was also dismissed.


It is clear that Ipex did not fully appreciate the nature of output-based IT projects and how they differ from input-based projects.  His Honour Justice Byrne clearly did so and Melbourne Water’s legal team should be given full credit for preparing and presenting their client’s case in such a way as to bring these features to the fore.

Of course, an output-based procurement approach is not always appropriate for every type of IT outsourcing project.  It is likely to be less appropriate for the procurement of off-the-shelf or commoditised IT systems and services, whereas it is very likely to be a good candidate for longer term and high value projects and/or where the customer wants its service provider to be its IT ‘business partner’ and provide a service that:
• meets its business requirements measured in terms of outputs/outcomes, not inputs;
• offers demonstrable value for money, with payments linked to said outputs/outcomes, not inputs; and
• is managed pro-actively, ie the service provider is both required and incentivised to identify opportunities to:
o increase the customer’s competitiveness;
o improve the customer’s business processes; and
o deliver cost savings.

Some might say that that’s what IT projects should always be about.

Julian Gyngell is a Solicitor & Attorney at Kepdowrie Chambers in Wahroonga New South Wales: julian@gyngellslaw.com and www.gyngellslaw.com

Copyright © October 2009. Julian Gyngell. All rights reserved.