If you hear or read the words 'according to a poll' in a broadcast or newspaper story, it's likely that the story is a result of a news release. All of us working in PR have resorted to research stories to get coverage for our organisations. I'm certainly no exception. When judgement day comes, I'll confess to the Valentine's day story I did for Eagle Star life assurance in 1997, claiming that women would love a pension plan for Valentine's day. The shame.

These non stories are a conspiracy between PR people and the media. They rarely have any substance to them, even though they abide by the rules governing market research. But I sense we are all getting bored by the tactic. My 1997 story was picked up by a stack of radio stations and the national press. Were I to repeat it now, I doubt any national would touch it. In the last week, journalists on Metro and The Observer I've spoken to have agreed with my view, even saying they delete emails from a certain insurance company because of the number of silly surveys.

The tactic can result in ridicule. Two weeks ago, the insurer LV issued a story claiming that inexperienced cyclists taking to the roads had resulted in a 29% increase in the number of road accidents involving cyclists. The Independent's report on the story carried the dramatic headline Credit crunch cyclists causing chaos on roads. The CTC, Britain's national cyclists' organisation, quickly condemned LV's story as Mickey Mouse research. Roger Geffen, CTC's respected campaigns manager, said LV's alarming figures appeared to no more than the difference between casualties in winter, when fewer people cycle, and the summer, when more of us take to our bikes. CTC's rebuttal was widely reported in the cycling press. In the words of Roger Geffen:

"Manipulating statistics for a PR stunt wastes the time of the people who took part in the survey. By demonising cyclists and scaring people into staying in their cars, it also undermines the efforts of charities like CTC to encourage more cycling and improve road safety for all”.

Things got worse for LV's PR team. In last Saturday's Guardian, Ben Goldacre devoted his Bad Science column to LV's story, under the headline Perils on the road to PR-reviewed data. Goldacre began:  "I've always fantasised that the insurance industry must possess a vast repository of useful data: the experience of centuries, carefully tabulated by actuaries into secret commercial databases containing a truth about human behaviour and risk that most epidemiologists and social scientists would kill for." LV's story destroyed that illusion, leading him to add, "I shall not be buying shares in this insurance company". (He couldn't anyway, as LV is actually a mutually owned friendly society, but he got his point across.)

LV achieved some media coverage. But it has alienated a sizeable group of people, and damaged its efforts to establish itself as expert on road safety. Not very smart. The release has now disappeared from LV's media centre but the damage has been done.