Marks and Spencer v Interflora: Court of Appeal throws out hand-picked witness evidence
Marks and SpencervInterflora
- Court
- Court of Appeal
- Judgment
- 20 November 2012
- Citation
- [2012] EWCA Civ 1501
Courts will reject witness evidence gathered by picking only the answers that suit your case.
Interflora wanted to call witnesses it had handpicked from a survey, keeping only the answers that helped its case. The Court of Appeal said no, and M&S won this part of the fight.
What happened
Interflora sued M&S for trademark infringement after M&S bought "Interflora" as a Google AdWords keyword, so that M&S flower delivery adverts appeared alongside Interflora's own search results. To support its case, Interflora ran survey exercises and then wanted to call as live witnesses only the people whose answers backed its argument, without using the surveys themselves as reliable statistical evidence. The High Court had allowed most of this evidence. M&S appealed against that decision.
What did the court decide in Marks and Spencer v Interflora?
The Court of Appeal allowed M&S's appeal. It held that evidence from witnesses selected this way was not admissible unless shown to have real value, where the likely benefit justifies the cost of using it, a tougher test than the previous approach of letting such evidence in unless it was shown to be worthless. The court also found the survey questions used to find these witnesses were leading, which tainted the answers they produced and, on its own, justified excluding the evidence.
What we learned from this
- What I'd take from this is that if you're relying on witnesses to prove customers were confused, you can't just keep the ones who agree with you and throw away the rest.
- If you're gathering consumer evidence for a trademark dispute, it's worth checking the questions you ask aren't leading and that the process could survive scrutiny in court.
- Evidence that costs a lot to produce needs to show real value before a court will let it in, not just the absence of obvious flaws.
- This ruling was about what evidence could be used, not about whether M&S's advertising actually infringed Interflora's trademark, that question was sent back for a later decision.
Things to be aware of
- This is a ruling about what evidence could be used in court, not a decision on whether M&S's AdWords use actually infringed Interflora's trademark.
- The test for trademark infringement confusion is different from the 'substantial number confused' test used in passing off claims, and evidence built for one doesn't automatically work for the other.
- The underlying infringement question was sent back to be decided separately, without the excluded witness evidence.
The businesses involved
Marks and Spencer PLC
UK retailer offering its own flower delivery service. Bought 'Interflora' as a Google AdWords keyword, which triggered the trademark dispute.
Interflora Inc
Operates a flower delivery network of independent florists and brought the trademark infringement claim against M&S over its AdWords use.
- Note
- Registered outside the UK
What happened next
- The underlying dispute began with Interflora suing M&S at the High Court over its sponsorship of the 'Interflora' keyword.Source: Pinsent Masons, March 2009
- Following further proceedings, Interflora later won its trademark case against M&S over the Google AdWords advertising.Source: Theguardian, May 2013
- The dispute concerned M&S's own flower delivery service competing with Interflora's independent florist network.Source: Rpclegal, May 2013
How we can help
If something in this case sounds familiar, talk it through with us. The first call is free for new clients.
This page is for information purposes only and does not constitute trademark advice or guidance. Summary drafted with AI assistance from the published judgment and checked by The Trademark Helpline (how we write these). Judgment: [2012] EWCA Civ 1501 on Find Case Law, published under the Open Justice Licence. Company details from Companies House; trademark counts from the UK trademark register. Last reviewed 27 October 2026. Report an error.
