Tuesday, June 25, 2013

Giant sucking sound: vacuum claims false

Dyson, Inc. v. Bissell Homecare, Inc., 2013 WL 2936453 (N.D. Ill.)

Dyson sued Bissell for false advertising; here the court grants Dyson’s partial motion for summary judgment and denies Bissell’s motion to exclude/disqualify certain experts.

Bissell advertises that consumers can improve their respiratory health by using Bissell vacuum cleaners, touting the specific air filtration performance of eleven different models on its website, packaging, and products.  Open-system (OS) vacuum cleaners are not sealed and leak some air from gaps and seams; most of the claims at issue related to OS models, though Dyson also challenged claims regarding Bissell’s Healthy Home vacuum, a closed-system model in which the unit is sealed.

In 2010, Bissell advertised that its OS vacuums had a “HEPA Media Filter,” and that the filters or vacuum cleaners “capture[d] over 99.9%” of certain allergens.  Dyson argued that HEPA is the term used in the US to indicate that a filter or filtration system traps 99.97% of dust and other particles that are 0.3 microns in size. But, Dyson continued, the statements were false because the OS vacuums didn’t meet HEPA standards or capture over 99.9% of allergens. And Bissell allegedly falsely claimed that the Healthy Home vacuum was “airtight” and that it captured 100% of certain allergens.

Dyson commissioned independent testing of the Bissell vacuums using allegedly industry standard tests, which showed that the filtration performance of the OS vacuums didn’t come close to HEPA level.  The testing also supposedly showed that the filters used in the vacuums didn’t even themselves meet the HEPA standards (or capture 99.9% of allergens, of course).  Further, the testing allegedly showed that the Healthy Home vacuum wasn’t airtight and didn’t perform at HEPA levels.

After Dyson sued, Bissell added new disclaimers indicating that only the filter, and not the vacuum cleaner as a whole, met HEPA standards, and revised the disclaimers in 2012.

Dyson also alleged that its unique designs and superior filters made its vacuums far superior to those of Bissell in terms of allergens.  Bissell, knowing the importance consumers place on healthy products, allegedly deceived consumers, who can’t independently verify its claims, thus harming Dyson, Bissell’s direct competitor.  Dyson sued under the Lanham Act as well as under Illinois’s
Uniform Deceptive Trade Practices Act, its Consumer Fraud and Deceptive Trade Practices Act, and common law unfair competition.

The court first evaluated Bissell’s challenge to an expert report on Dyson’s survey.  While Dyson bore the burden of establishing admissibility, and while a consumer survey must comply with principles of professional survey research, consumer surveys are rarely so flawed as to be completely unhelpful to the trier of fact and therefore inadmissible; shortcomings generally go to weight instead.

Dyson’s expert, Kivetz, examined certain advertising claims that the court agreed were representative of the claims at issue.  On four panels of the the package for one of its vacuums, Bissell stated: “HEPA Media Filter*—The filter captures 99.9% of pollens and ragweed from the air passing through it.” The packaging also included Bissell’s initial disclaimer on the lower left corner of the back panel, which states: “ *The filter media, not the vacuum as a whole, complies with the High Efficiency Particulate Air (HEPA) filter specification.”

The test cell saw that packaging, while the control group viewed modified packaging with a disclaimer directly below the claims stating  “*The vacuum during operation does not meet the HEPA filter specifications.”  (Wow, that’s a really defendant-favorable control ad.)  Respondents were asked what, if anything, was being communicated to them about the HEPA Media Filter and about the vacuum cleaner as a whole, with more specific followups.  Responses were coded by an independent coder.  Kivetz concluded that the challenged claims caused net 27% deception, and that Bissell’s initial disclaimer didn’t work.

Bissell argued that the survey questions didn’t address the relevant issue and were based on undefined terminology.  Respondents were first asked “what message or messages, if any, does this package mainly communicate to you about this product?” and “Anything else?” Then the filter question: “Does or doesn't this package communicate to you anything about a HEPA Media Filter?” A no response ended the survey.  A yes response led to: “What does this package communicate to you about a HEPA Media Filter?” and then “What else, if anything, does this package communicate to you about a HEPA Media Filter?” Next: “Do you think that:

This package does communicate to you whether the vacuum provides HEPA level performance during operation

This package does not communicate to you whether the vacuum provides HEPA level performance during operation, or

Do you not know?”

If the respondent chose the first option (not clear if it was rotated in practice), there was a follow-up:

 “What does this package communicate to you about whether the vacuum provides HEPA level performance during operation?” and “What else, if anything, does this package communicate to you about whether the vacuum provides HEPA level performance during operation?”

Bissell argued that the later questions didn’t address the relevant question of whether participants believed the claims related to the vacuum as a whole or merely to the filter.  Instead, the survey induced respondents to erroneously draw a distinction between the vacuum during operation and not in operation.  Bissell identified a “handful” of verbatim responses indicating that those respondents believed that HEPA level performance could only be achieved if the vacuum was running.  But this handful indicating “possible confusion” on the part of a few respondents didn’t sufficiently show that the survey didn’t address the relevant issue.  “ Confusion on the part of a few Participants is the unavoidable, albeit undesirable, by-product of asking Participants open-ended, non-leading questions during a Survey.”

Bissell argued that the questions didn’t raise the possibility that filtration performance claims might be based on filter-only testing, but that was “an objection to Kivetz's failure to use leading questions to suggest the response desired by Bissell.”  Plus, the progression of questions began broadly, then focused on the filter alone, then asked whether the claims communicated anything about the filtration benefits of the vacuum as a whole—which was the point of interest.

Bissell next objected to the use of “HEPA level performance” in the questions as an undefined term.  It argued that HEPA standards differ depending on what is being tested or described.  But the survey wasn’t designed to test consumers’ understanding of the phrase “HEPA level performance” or to determine whether the vacuum or filter actually achieved HEPA level performance.  Instead, the survey tested whether consumers believed that Bissell was making claims about the vacuum as a whole.  The precise definition of “HEPA level performance” was therefore not relevant to this survey, and it was close enough in meaning to Bissell’s claims (including the 99.9% claims) to serve as an appropriate proxy for them.  Mead Johnson & Co. v. Abbott Laboratories, 201 F.3d 883 (7th Cir. 2000), didn’t apply because the survey wasn’t being used to define the meaning of that phrase.

Bissell also challenged the coding of the survey.  It contended that many responses indicated neutrality or uncertainty about whether respondents believed the vacuum provided HEPA level performance during operation.  The coding was done independently by a coder who didn’t know the purpose of the survey.  This masked procedure made Bissell’s claims of bias unconvincing.  “Moreover, Bissell has not presented any legal authority supporting the proposition that survey responses must be coded in a completely objective manner, and certain case law in fact suggests otherwise.”  It also didn’t matter that the coder wasn’t given “relevant information about filtration standards, HEPA, and HEPA protocols,” since the coder wasn’t evaluating how consumers defined HEPA level performance.  Likewise, it didn’t matter that Kivetz didn’t know that the HEPA standard referenced in Bissell’s first disclaimer was a test protocol for filter media alone; survey respondents would likely also lack such knowledge.

Bissell then argued that presenting only the packaging to respondents didn’t properly replicate market conditions, which should have included multiple products and point of sale materials.  But this wasn’t a trademark case, and there was no precedent that comparative products and literature are required to replicate market conditions in a false advertising case.  And “case law suggests that ‘standards for replicating market conditions for a consumer confusion survey in a trademark infringement case cannot be applied wholesale’ to cases alleging false advertising.”  A mall intercept survey using an actual package and a price card similar to the marketplace reality, along with instructions to examine the package as if the respondent was at the store thinking of buying it, were fine.  (Though I don’t think there was any problem with this survey, I don’t understand the rationale for distinguishing false advertising and trademark here.  There could be one, I suppose, having to do with what questions consumers are asking themselves when they make decisions—but then we’d be forced to confront the fact that trademark doesn’t require materiality, though it should.)

Bissell then pointed out that Kivetz criticized a survey in another case as unreliable for allowing participants to continue to view the packaging while the answered the survey questions, then did the same thing here.  But Kivetz identified material distinctions in the respective surveys, specifically the category of products at issue.  Here, the products were vacuums, “the purchase of which would likely involve deliberate and thoughtful decision-making by a consumer before its purchase because of its significant cost,” whereas the previous survey involved low-cost beverages.  In the context of a big-ticket purchase, market conditions are better approximated by allowing respondents to view the stimulus throughout the survey.

None of the criticisms warranted exclusion of the survey.

Likewise, Bissell failed to disqualify Dyson’s technical expert on filtration, Susan Goldsmith, even though she’d done work for Bissell as well, once more than ten years before the case began and once during the case, before Dyson disclosed her as an expert.  Goldsmith’s relationship with Bissell was far from exclusive. Instead, Goldsmith performs services for the entire vacuum industry, including for direct competitors.  She’s one of the world’s leading experts on vacuum cleaner testing, and few experts are capable of performing her services.  Her company was possibly the only independently accredited lab in the world with both the expertise and the expensive, specialized equipment required to run the tests at issue in the case.  The fact of her previous work for Bissell, including her exposure to confidential information, wasn’t sufficient to disqualify her. 

The court also noted that Dyson first hired her for work relating to this case.  Only after Goldsmith got results adverse to Bissell’s position did Bissell hire her as a consultant.  And by the time Bissell hired her, it should have been well aware of her work for Dyson related to the litigation based on Dyson’s initial disclosures, not to mention Bissell’s awareness that she worked for competitors.  It would be inequitable to disqualify her under these circumstances.

The court then turned to the parties’ cross motions for partial summary judgment on the falsity of the original statements about Bissell’s OS vacuums; the Lanham Act and state law claims were identical, except that for an Illinois CFDTA claim a plaintiff must also prove that the defendant intended consumers to rely on the deception.  Where an ad explicitly or implicitly represents that tests prove a claim, a plaintiff can show falsity by showing that the tests don’t prove the proposition for which they’re cited.  Bissell’s use of the term HEPA and reference to a specific numerical percentage implicitly (NB: and necessarily) indicates that Bissell was relying on tests to support its claim.

The parties disagreed over whether there was one definition for “HEPA.”  But Bissell failed to provide sufficient evidence to create a material dispute of fact: in the US, HEPA means that a filter or filtration system has a minimum efficiency of 99.97% for 0.3 micron particles when tested at the application's rated air flow.  Plus, even if the court applied Bissell’s suggested alternative standards, Bissell didn’t show that the ads would’ve been accurate under them anyway.

Bissell admitted that it didn’t base its statements on any tests performed on the OS vacuums themselves.  Instead, it based its statements on tests conducted on the filters alone.  However, Bissell requested that these standalone tests be done at air flow rates less than the actual rates created by the OS vacuums.  Even Bissell’s own expert thought this wasn’t ideal.  The court found it undisputed that, even if the claims were about the filters alone, Bissell still didn’t have adequate testing to support its claims.  Dyson’s own testing showed that when the filters were tested separately, they didn’t meet HEPA standards.  Bissell didn’t refute these results and didn’t create a genuine issue of fact about them.

Bissell argued that it used the term “Media” in its initial claims, which is understood in the industry to refer to the paper filter and thus notified consumers that its claims related only to the filters. But the court considered the statements in context: “The undisputed facts in this case and the evidence presented by Dyson clearly show that Bissell used the 2010 Statements to advertise and sell vacuum cleaners, not filters.”  Thus, Bissell’s claims were false and unambiguous representations that the vacuums operated at a HEPA filtration level and captured 99.9% of allergens.  Even if “Media” had been a clear reference to the paper in the filters, Dyson still clearly established literal falsity.

The court also found that the statements were material.  Bissell’s consumer research on the importance of respiratory health to improve its marketing showed materiality; Bissell’s studies showed that filtration efficiency was important to consumers.  Plus, the statements appeared on the front packaging of some vacuums, which was where Bissell highlighted prominent selling features.  An employee testified at his deposition that the filtration efficiency was one of the “key features that a customer [is] interested in....”  Plus, the statements related to health; particles less than 2.5 microns in size can result in adverse health effects in humans if inhaled.  No reasonable trier of fact could avoid finding materiality.

In addition, there was no disputed issue on probable injury to Dyson. Dyson wasn’t seeking its own lost profits; it need only show probable injury to prevail.  Bissell was Dyson’s number one competitor in the industry, both on volume and dollar market share terms.  Dyson and Bissell vacuums are sold in many of the same places and often appear side-by-side.  The false statements here would’ve made cheaper Bissell vacuums more appealing than more expensive Dyson vacuums advertising the same high filtration efficiency rates.  Thus, Dyson suffered a probable and discernible competitive injury as a result of the false statements, as a matter of law.

Dyson didn’t move for summary judgment on the later ads with disclaimers, though Bissell did; the court denied Bissell’s motion for failure to explain why it was entitled to summary judgment.  Dyson had shown adequate evidence that the later ads were false to go to a trier of fact. Likewise, there was enough evidence that the closed-system Healthy Home vacuum didn’t capture 100% of allergens to go forward.

Finally, the court rejected Bissell’s affirmative defenses, including release—that was based on a 2006 release ending earlier litigation.  The release couldn’t have contemplated later-made ad claims for new products like these, and in any event Illinois doesn’t enforce contracts exculpating persons from the consequences of their wilful and wanton acts, and with respect to the initial statements, the facts showed that Bissell’s false advertising was wanton and willful.

Acquiescence, estoppel, laches, and a statute of limitations argument also failed for essentially the same reasons: Dyson never condoned Bissell’s conduct and took action promptly.  For laches, Bissell argued that it was prejudiced by incurring marketing expenses, but didn’t show that any specific expenses were connected to any Dyson delay.  “[U]nder Bissell's own version of [the] facts, Bissell was able to sell OS Vacuum Cleaners for years before Dyson challenged the current models, which indicates, if anything, that Bissell may have profited from any delay on the part of Dyson.”  And the court borrowed Illinois’s three-year statute of limitations under the UDTPA/CFDTPA and found that Dyson acted promptly when it knew or should have known of its claims (including by engaging in testing to disprove the filtration claims).


  1. Any follow up on this case? I understand there has been a decision, but I can't find it.

  2. Do you have an update of this case? I heard that there has been a decision, but I can't find it anywhere.

  3. The opinion appears to have been vacated because of the parties' settlement. The materials should be up on the RECAP website.

  4. Thanks. Do you have a link to the site?

  5. You can search for the case on http://archive.recapthelaw.org/