Before wowing the books world with The Tipping Point, Blink and Outliers, Malcolm Gladwell came to prominence as a staff writer for the New Yorker, for whom he has worked since 1996. What the Dog Saw is a collection of 19 of his favourite pieces for the magazine, tackling wildly disparate subject matter, from intelligence failures in the build-up to 9/11 to the success of Heinz ketchup.

It's clear from the off that part of Gladwell's success is down to an engaging tone that makes the seemingly mundane utterly absorbing. Witness the opening article, a profile of Ron Popeil, the founder of kitchenware firm Ronco and inventor of the Showtime rotisserie, which has racked up sales worth more than $1bn thanks largely to late-night infomercials showcasing Popeil's remarkable salesmanship skills. It would be easy to sneer at one of QVC's biggest moneyspinners (the network's executives are awestruck as Popeil grosses more than $1m in a single hour-long slot), but Gladwell focuses on Popeil's infectious enthusiasm for creating products that will make cooking easier for people.

Elsewhere, he deftly reveals how the rise of feminism between the 1950s and 1970s was played out in hair dye advertising slogans, from Shirley Polykoff's teasing, enigmatic "Does she or doesn't she?" for Clairol in 1956, to Ilon Specht's altogether more self-assured "Because I'm worth it" for L'Oréal 17 years later. Gladwell asks: "In writing the history of women in the postwar era, did we forget something important? Did we forget hair?"

He really hits his stride, however, when he tackles broader issues and finds connections between seemingly unrelated things. "The Picture Problem", for example, is a wide-ranging essay exploring how improvements in imaging techniques across a range of fields haven't necessarily resulted in more accurate or effective results. During the second Gulf war, much was made of American planes' missile-guidance technology, which enabled them to deliver bombs to within 13 metres of their targets; 50 attempts were made to kill Saddam Hussein and his senior advisers by this method. But without accurate information as to the Baathist leaders' precise whereabouts, not one was successful.

Similarly, more sophisticated mammograms have resulted in soaring detection rates for early-stage tumours in milk ducts, the vast majority of which are promptly removed, in 30% of cases resulting in mastectomy. Yet, counterintuitively, instances of invasive breast cancer have continued to rise. Only a small percentage of these tiny tumours would have gone on to pose a threat if left untreated, but, Gladwell argues, not knowing which ones is resulting in chronic over-treatment. The message is that without supplementary information to back up what we see, clearer pictures sometimes only cloud the issue.