Facebook is hoping to prove its value to marketers with a new partnership with Dunnhumby. The tie-up aims to make it easier to connect ad exposure on Facebook directly to in-store and online sales at an anonymised household level.
Alex North, head of partnerships at Facebook in EMEA says: “The ambition with this partnership is to help advertisers understand the true business of advertising on Facebook in a granular, anonymised and secure way.
“One of the key benefits is that the solution will help advertisers understand their Facebook campaign metrics – what is working and what is working less well – and optimise those campaigns based on an objective understanding of whether their campaigns are driving sales.”
How the partnership will work
Historically, Dunnhumby has been able to measure and optimise the effect of media and promotions in the retail space and use that to personalise communications around shopper preferences. The most notable example is for Tesco Clubcard, where the supermarket bases the coupons and deals that get sent out to its loyalty members on their shopping habits.
The aim now is to move outside the retail environment and into brand marketing. The deal will link Dunnhumby data on 17 million shoppers in the UK with Facebook’s user base – the 37 million people that log in and use the service monthly. It is working with Acxiom to provide a “safe haven” in the middle of the data sets to ensure information remains anonymous.
It works by using a “test and control” methodology where one group is exposed to advertising on Facebook and the other is not. Given that all other parameters are equal, brands should be able to tell what impact a campaign on Facebook is having on sales.
Brands will still be able to measure engagement – for example likes or shares. But they will now also be able to monitor sales metrics including units sold, visitor numbers, whether the campaign brought in new customers or is driving frequency of existing numbers. They will also get lifestyle and life stage data – so demographics and whether customers are a family or single.
Adam Smith, head of media strategy at Dunnhumby, explains: “In a nutshell what brands will be able to do from a media measurement perspective for the first time is connect ad exposure on Facebook directly to sales in-store and online at a household anonymised level.
“It will make a link between what people see on their mobile device and the impact that has in-store on sales behaviour over time.”
Adam Smith, head of media strategy, Dunnhumby
How brands can use the partnership
The partnership has been in the works for more than two years and in that time the technology has been piloted on 10 brands at eight major FMCG clients. It is now rolling out so that any brand that sells through Tesco can use it.
Dunnhumby says it has so far been used to help develop creative best practice, to work out audience responsiveness to advertising and frequency of exposure, as well as sales and ROI.
While Facebook cannot share which brands have trialled the technology, examples include soft drinks, laundry detergent and beer brands. For example, one beer brand tested six different creative executions on mobile to discover which one best drove sales.
Another test for a detergent brand found that advertising on Facebook drove behaviour change so more people went down the aisle in-store. But once there a competing brand saw a sales hike because it had a deeper promotion, showing the importance of marrying Facebook campaigns with offers and deals.
Finally, an ice cream brand launched NPD testing on two different target audiences and found that one group was four times more responsive than the other. However while the messaging pitched to a young, female, upmarket audience the data actually showed a “huge sales response” from older age groups.
North explains: “This is a very large sample that brands can cut by an manner of factors – frequency, the creative, category purchasing and understanding sales uplift on other brand in the category. It is a very granular measure of the impact of advertising and will show the nuances that impact sales.”
Dunnhumby believes there are benefits even if the results are negative. For example if there is no sales boost a brand will know that their creative or targeting is not working and can adapt their strategy.
North concludes: “It is important for brands to learn what works and what doesn’t so that when they invest they are doing so in ways that will drive real sales impact.”