EBay predicts parents’ purchases

Discussion
Nov 17, 2015
Tom Ryan

EBay Advertising U.K. has launched a predictive targeting tool for identifying and segmenting "parent" shoppers. The launch is part of a larger push to use online behavior to better engage shoppers across life stages.

The "New Mums" targeting product is based on "real shopping behavior" and taps insights from eBay’s 18 million monthly unique users. Rather than use a one-size-fits-all approach to the Baby & Toddler category, the product enables brands to target gifters and other shoppers across five distinct shopper groups:

The expectants: Peak purchases during this period include nappies, nursery furniture and buggies.

New arrivals: Dummies (pacifiers) and comforters top shopping lists along with DVDs and other home entertainment for parents "going out for a night on the sofa," according to eBay.

Tiny tots: Stair gates, child locks and new car seats top wish lists for parents managing the adventures of one-to-two year-olds.

eBay UK mums

Source: eBay Internal Data 2015

Toddler time: With many parents of two- and three-year-olds having a second child on the way, a bigger family car and a tablet loaded with kid-friendly distractions become big sellers.

Pre-school mums: With kids turning four and set for school, "big changes" are underway as nurseries are converted into kids rooms and kids start craving scooters, bikes and determining their favorite toys.

"We analyze millions of data points over several years and create correlations between buying and browsing behaviors," said eBay in explaining how the tool works. "The results enable us to pinpoint the specific purchases a consumer is primed for and, in this case, what stage of motherhood they are at."

The New Mums product is the first part of the eBay Advanced Targeting predictive offering. The tool is designed to better target consumers across life-stages, seasonal events and shopper milestones – such as "premium" and "sales" Christmas shoppers; "home-movers"; "holiday makers"; and "new car buyers."

Alessandra Di Lorenzo, commercial director, eBay Advertising UK, told The Drum that the company hopes to counter any concerns from consumers around its highly-targeted advertising with heightened relevance in the consumer offerings and by promoting the program’s benefits on its site.

"Targeting based on life stage and seasonal hooks is nothing new in the industry, but our ‘eBay Advanced Targeting’ will be based on actual behavior on the site, rather than demographic assumptions" she told netimperative.com. "By looking at search and purchase history, we can help brands step away from stereotyping shoppers, and greatly improve their relevance and results."

How effective do you think predictive targeting tools based on online shopping behavior will be? Will consumers likely find such targeting more relevant to their needs or intrusive?

Please practice The RetailWire Golden Rule when submitting your comments.
Braintrust
"I think predictive targeting will be hugely effective. A simple improvement on this model would be to be upfront with consumers about targeting and ask them during checkout if they would like to see more items like what they are purchasing."
"Woe be to the retailer who uses Big Data to predict future purchases and makes the wrong offer to a prospective consumer! And even if predictive offers are accurate, there is the "creep out factor.""
"Be ready, eBay will know when you need a new blue button-down Oxford shirt. You do still wear those, don’t you?"

Join the Discussion!

6 Comments on "EBay predicts parents’ purchases"


Sort by:   newest | oldest | most voted
Zel Bianco
Guest
3 years 9 months ago

I think predictive targeting will be hugely effective. A simple improvement on this model would be to be upfront with consumers about targeting and ask them during checkout if they would like to see more items like what they are purchasing (this could also ask them if they are buying a gift — surely you could target new aunts and uncles in a similar way as their parents, but you might tend towards recommending loud and obnoxious toys!)

This prevents offending consumers who would find predictive targeting intrusive while also giving you the opportunity to collect more data and sell more relevant products.

Chris Petersen, PhD.
Guest
3 years 9 months ago

Woe be to the retailer who uses Big Data to predict future purchases and makes the wrong offer to a prospective consumer!

The problem with Big Data predictions is that they are based upon “data points over several years and create correlations between buying and browsing behaviors.” While these correlations are based upon online behavior, they are still assumptions based upon “predictive relationships.” These assumptions are not infallible at the individual consumer level. And even if predictive offers are accurate, there is the “creep out factor.”

Using similar technology, Target was able to make highly relevant “expectant” offers to a teenager before her family knew she was pregnant. How many customers will be comfortable with eBay targeting them as “expectants” based upon their browsing behavior?

Ben Ball
Guest
3 years 9 months ago

Online is the first shopping environment where researchers have been able to practically and reliably link vast amounts of purchase data with individuals (or at least individual IP addresses). The predictive patterns should be fairly easy to discern and apply in this environment. It has never been complicated. It has simply always been too complex due to disparate systems, incomplete records and data that could not be accurately linked.

Be ready, eBay will know when you need a new blue button-down Oxford shirt. You do still wear those, don’t you?

Dan Raftery
Guest
3 years 9 months ago

Predictive consumption tools are finally getting to the point where supply chain partners can use them to drive real efficiency improvements. Here I include noticeable improvements for consumers, many of whom will welcome the help. Smart of eBay U.K. to target new mums, who are probably the most receptive to the support.

Sure they will make mistakes and there will be some very vocal privacy advocates. Those can be managed. With loyalty being defined as frequency these days, this makes a lot of sense for any retailer with the resources.

Martin Mehalchin
Guest
3 years 9 months ago

If this is executed well it could be very effective in generating lift. But eBay had better be confident in the accuracy of the tool. Parenthood is an extremely personal and emotional part of people’s lives, so any misfires with this program could unleash an avalanche of negative reactions and PR. Chris is spot on to remind us of the Target example from a few years ago.

Arie Shpanya
Guest
3 years 9 months ago

I think this is a great idea — making life easier (and of course boosting sales) for new parents who don’t have much free time. Whether it is successful or not will depend on how it is executed.

wpDiscuz
Braintrust
"I think predictive targeting will be hugely effective. A simple improvement on this model would be to be upfront with consumers about targeting and ask them during checkout if they would like to see more items like what they are purchasing."
"Woe be to the retailer who uses Big Data to predict future purchases and makes the wrong offer to a prospective consumer! And even if predictive offers are accurate, there is the "creep out factor.""
"Be ready, eBay will know when you need a new blue button-down Oxford shirt. You do still wear those, don’t you?"

Take Our Instant Poll

How appealing do you think offers derived by predictive targeting tools will be to consumers?

View Results

Loading ... Loading ...