Brands looking to optimise online advertising spend could find new opportunities by honing in on micro-level purchase data to reach audiences in a wider range of categories.

New data shown exclusively to Marketing Week reveals consumers buying certain products are more likely to buy goods in other seemingly unrelated categories that brands could dismiss as irrelevant.

Data from Skimlinks, which uses machine learning to observe behaviour and create profiles from 1.4 billion users across its publisher network, shows that purchases do not happen in isolation.

For example, consumers who are looking to buy a large appliance for their kitchen, such as a washing machine or dishwasher, are eight times more likely to buy a coffee machine than the average consumer. However, those with no allegiance to a particular brand of appliance have almost no likelihood of buying a Nespresso coffee maker. Conversely, those looking for a Bosch appliance are 57 times more likely to buy a coffee machine and within that category they are eight times more likely to buy a Nespresso-branded product.


The same pattern is seen when looking at the difference between coffee machines that use pods and those that use ground coffee. Bosch appliance shoppers are significantly more likely than the average appliance buyer to purchase a pod-based machine.

The research also shows that people who are in the market for a Miele washing machine are 10 times more likely than average to buy a larger range-style cooker and 30 times more likely to be in the market for a dedicated wine chiller.

“From a customer’s point of view, [the purchase journey] is seamless,” says Edward Thomas, head of audience at Skimlinks. “They are not necessarily thinking about these [purchases] as discreet decisions, so what [the data shows] is how connected things are. By being a shopper for one product category, you are incredibly likely or equally unlikely to be [a shopper] in another.”

There are some clichés and obvious insights that are borne out at a micro level within the data. People in the market for a motorbike and associated accessories are  10 times more likely to be reading content about divorce, while a consumer that has bought maternity clothing in the past three months is 20 times more likely than the general population to buy a car seat.

Within homes, the TV is a central focal point but it is also an important purchase moment that triggers re-evaluation of other products.

Perhaps predictably, people who are in the market for a new TV are 19 times more likely to buy surround sound and DVD or Blu-ray players. With all those devices they are 26 times more likely to need a universal remote. However, less obviously, consumers are also seven times more likely to re-evaluate their broadband and entertainment package at the same time.

Family stereotypes

Regarding purchase behaviour within families, the results again conform to certain stereotypes. Shoes are generally bought by the person who will wear them but female shoppers looking to buy women’s shoes are 4.8 times more likely to buy baby and children’s clothes; male consumers buying men’s shoes are 2.9 times more likely to buy similar items. The same buyers of women’s shoes are also 25% more likely to be buying men’s underwear than their male counterparts.

“Some people believe that [a woman being] the primary shopper in a household is a concept that is slowly declining, but based on the data today this [is still the case],” says Thomas.


Having a child obviously has an immediate impact on parents’ purchase behaviour but it also has a long-term effect on how they buy. Being in the market for maternity clothing, for example, logically predicts that a shopper will purchase nursery furniture within the next three months – they are 18 times more likely to do so than the average shopper.

In addition, consumers in the market for nursery furniture are 15 times more likely to buy a digital camera within the following six weeks.

Marketers, of course, need to plan ahead but understanding the path consumers tend to follow within certain categories, and how these choices may influence their purchase behaviour in other sectors, can help brands better target their message and budget online to avoid wastage.

Enter the ‘insight innovation’ category at the Marketing Week Awards. For more details, click here

  • John Hamilton

    interesting theres no ref to the bathroom ???? 10% of GDP

  • This is a really good article! It’s interesting what connections can be discovered and implemented when you take a look at the stats that reveal an almost “natural” upsell. Similar to the would-you-like-fries-with-that approach, this method could be a greater benefit for everyone involved.