EIS Gmbh v LELO Oceania Pty Ltd (Costs)

This was a costs judgment delivered after the real patent fight had already been decided. In the earlier liability trial, EIS failed on infringement, none of the LELO products were found to infringe the relevant claims, and the patent was held invalid and should be revoked because the claims did not comply with ss 40(2)(a), 40(2)(b) and 40(3) of the Patents Act. The respondents therefore won the case overall.

But that did not end the matter. Costs became a separate dispute because the respondents had not succeeded on every argument they ran. Their revocation case succeeded only on limited grounds. Other invalidity grounds failed. Claims they made about statements by EIS failed. Their attacks on some experimental evidence failed. The Court also criticised the involvement of one of their experts.

So the practical question became: should a party that won overall still recover all of its costs when parts of its case added substantial expense but did not succeed?

That is the commercial value of this judgment. It is not mainly about patent doctrine. It is about litigation discipline, case management and the financial consequences of running too many points in a technical dispute.

Downes J dealt with three main questions.

First, should the Court make separate costs orders for EIS's infringement claims and for the respondents' revocation cross-claims, or should it make one global costs order? EIS wanted separate orders. The respondents wanted one overall order.

Second, if the respondents were the successful parties overall, should their costs still be discounted because they had pursued a number of unsuccessful grounds and evidentiary attacks that occupied a large part of the trial?

Third, should the respondents receive indemnity costs after 19 September 2024 because they had made a settlement offer that was better than the result EIS ultimately achieved?

Those questions were not answered by simply identifying the winner. The Court approached them through practical case-management principles, the overlap between the claim and cross-claim, and the reasonableness of the parties' conduct at the time decisions were made.

The Court rejected EIS's request for separate costs orders. It held that a global costs order was appropriate because the respondents' revocation cross-claim was brought as a defence to the infringement proceedings and the defence and successful invalidity claims had a common substratum of fact. The Court identified the proper construction of the patent claims as that common foundation.

That point matters in practice. In many patent cases, infringement and invalidity are not cleanly separable. The same claim construction issues, technical evidence and expert analysis often sit underneath both sides' positions. Where that is true, a court may see little value in trying to split the legal work into neat compartments after the event.

The Court also took a practical view of what would happen next if it accepted EIS's approach. Trying to disentangle the costs of EIS's case from the costs of the respondents' case would likely increase the expense of any taxation process. In other words, the exercise of separating costs would itself create more argument and more cost. The Court said that approach had no real utility.

For businesses, this is a useful reminder that costs orders are often shaped by practicality. If your claim and the other side's defence or cross-claim are built on the same technical and factual platform, the Court may prefer one broad order rather than a highly segmented one.

The respondents argued that there should be no discount because they could not have known in advance which points would succeed and because parties should not be discouraged from vigorously defending infringement proceedings, including by cross-claiming for invalidity. The Court did not accept that submission in full.

Downes J identified several substantial parts of the respondents' case that failed. Their cross-claim did not succeed on best method, utility, novelty or inventive step. The Court said those topics occupied a large proportion of the evidence and the time spent at trial. The respondents also failed on claims concerning statements made by EIS. Their attack on the Outerspace experiments and the Cobalt experiments was rejected as well.

The Court found that these were not minor side points. They had caused EIS to incur significant additional costs, including extra trial time. The judgment also focused on expert evidence. The respondents had called two experts instead of one, and the Court said Mr Duff should not have given evidence for reasons explained in the liability judgment.

His involvement, including his own experiments and criticisms of others' experiments, expanded the factual and legal issues EIS had to address, generated evidentiary objections, and prolonged the concurrent expert evidence session.

The Court referred to modern case-management principles and said apportionment of costs where a party succeeds on some claims is consistent with those principles. It then made an important broader statement for patent litigation: parties should be encouraged to pursue their best points only, not all arguable points. The possibility of a discounted costs order may create that incentive.

The Court also rejected the suggestion that the claims about statements made by EIS were too minor to matter. It noted that those claims had been the subject of two interlocutory applications before trial and had been run in closing submissions in a way that expanded beyond the pleaded case.

Finally, although the respondents had successfully defended infringement, the Court noted that one of their primary pre-trial and trial arguments was not accepted in the way they had advanced it. They had argued that the devices needed to be measured when used on the clitoris in order to establish infringement. While the Court accepted their construction argument, it did not accept that proposition as necessary to establish infringement.

That reduced the force of their submission that their defence had succeeded exactly as run.

Taking all of those matters together, while still recognising that the respondents were the successful parties overall, the Court fixed a 40% discount. EIS therefore had to pay 60% of the respondents' costs.

The respondents also sought indemnity costs after 19 September 2024 based on a settlement offer sent by letter. The Court first noted that the letter was not a formal Notice of offer to compromise under the Federal Court Rules, so the usual rule-based consequence under r 25.14 did not automatically apply.

That did not end the issue, because an informal Calderbank offer can still support indemnity costs. But the Court emphasised that a better-than-trial offer is not enough by itself. The offeror must show both that there was a genuine offer of compromise and that it was unreasonable for the offeree not to accept it.

The Court referred to the usual factors for assessing reasonableness, including the stage of the proceeding, the time given to consider the offer, the extent of compromise, the offeree's prospects of success as at the date of the offer, the clarity of the offer's terms, and whether indemnity costs were foreshadowed.

Some of those factors favoured the respondents. The offer was clearly expressed, it gave EIS two weeks to consider it, it warned that indemnity costs might be sought, and there was no dispute that the offer was better than the result EIS eventually achieved at trial.

Even so, the Court held that EIS had not acted unreasonably in rejecting it. At the time of the offer, much of the evidence supporting the eventual finding of lack of clarity had not yet been adduced. EIS had only just completed its evidence in answer, reply evidence had not yet been filed, and the joint expert conclaves that produced JER 1 and JER 3 had not yet occurred. So EIS did not yet have access to some of the pivotal material later relied on in the liability judgment.

The respondents argued that they had already told EIS long before the offer that EIS could not prove infringement because the claim language required measurement when the device was used on the clitoris. But the Court pointed out that, although it accepted the respondents' construction argument, it did not accept the broader proposition in the way the respondents had advanced it.

The Court also noted that EIS had been grappling with that contention before trial and that the respondents maintained it at trial without success, notwithstanding that their own expert did not agree with it.

In that context, the Court concluded that it was not unreasonable for EIS to refuse the offer. Indemnity costs were therefore refused.

For business owners, the practical lesson is about controlling the shape of the dispute. Patent litigation can become expensive because each additional invalidity ground, expert opinion, experiment, interlocutory application and side allegation creates more documents, more evidence and more hearing time. This judgment shows that the Court may later ask whether those additions were really justified.

If you are enforcing a patent, this case is a reminder that a settlement offer should be assessed against the evidence actually available at the time. A later loss does not automatically mean your refusal was unreasonable. But you should be able to explain, contemporaneously, what evidence was still missing and why the offer was not commercially acceptable then.

If you are defending a patent claim, the judgment is a warning against overloading the case. A revocation cross-claim may be entirely appropriate, but the Court may still reduce your costs if you pursue too many grounds that fail or if your expert strategy expands the case unnecessarily. The Court expressly endorsed the idea that parties should pursue their best points only.

That means businesses should ask their legal team for more than a merits view. They should ask for a live costs-risk map. Which issues are central? Which are backup points? Which expert evidence is essential? Which allegations are likely to increase cost without materially improving the result? Those questions can affect the commercial outcome almost as much as the legal merits.

The costs judgment was delivered on 29 October 2025. It followed the liability judgment published on 18 September 2025. The costs issue was determined on the papers after the Court sought submissions from the parties. The orders required EIS to pay 60% of the respondents' costs of the originating application and the notice of cross-claim in each proceeding, as taxed if not agreed, subject to any costs orders already made in either proceeding.

The judgment also records that the respondents' settlement offer was sent on 19 September 2024 and gave EIS two weeks to consider it. That timing was important because the Court assessed reasonableness by reference to the state of the evidence at that date.

This page is based on the Federal Court's costs reasons in EIS Gmbh v LELO Oceania Pty Ltd (Costs) [2025] FCA 1313. Those reasons repeatedly refer to the earlier liability judgment, EIS Gmbh v LELO Oceania Pty Ltd (Liability Trial) [2025] FCA 1111, for the detailed technical background and fuller discussion of the patent issues.

This page is general information only and not legal advice. Patent disputes and costs outcomes are highly fact-specific, especially where expert evidence, experiments and settlement offers are involved.