Facebook undervalued in last-click attribution model: Report

New Kenshoo research finds social networking platform is undervalued by up to 30 per cent in a last ad attribution model against other types of custoemr attribution techniques

Marketers who use last ad attribution to measure Facebook’s brand impact as part of a multi-channel campaign are devaluing the social network’s impact and making flawed marketing decisions, a new report claims.

According to the Quantifying the Impact of Multi-touch Attribution research paper from digital marketing technology company, Kenshoo, the last ad model undervalues Facebook advertising by 12-30 per cent compared to five other industry standard alternative attribution models: First only, prefer first, divide equally, prefer last and U-Shaped.

While criticism of last ad attribution has grown in recent years, it remains widely used by marketers globally. There has also been insufficient data to backup the flaws in such an approach, the Kenshoo report stated.

Kenshoo analysed campaign performance data across its clients managing ads across multiple digital marketing channels between March and May. The company employed a cost-per-acquisition metric across all advertisers, which was calculated by aggregating all recorded conversions and dividing that number in the total media cost of the ads which drove them.

Kenshoo’s clients include Expedia, Hitwise, John Lewis, Starcom MediaVest Group, Tesco and Travelocity.

The research found Facebook was undervalued by 30 per cent using a last ad technique compared with a first only approach, where credit is only given to the first customer interaction. It was also undervalued by 20 per cent in a last ad model when contrasted with a prefer first approach, where the first customer touch gets the majority of the credit, and each following interaction is credited less in a rules-based, linear way.

Looking at Facebook’s worth using last-click attribution was also undervalued by 16 per cent against a divide equally technique, where the value of each touch is equal to the rest; and 15 per cent undervalued compared with a u-shaped approach, which uses an 80:20 rule to give the first and last click majority weighting and the middle clicks less credit.

A prefer last approach gives the last click most credit, followed in a linear fashion by the second-last touch. Against this example, Facebook was undervalued by 12 per cent in a last ad approach.

“Marketers relying on last ad attribution to optimise their campaigns could be making million-dollar decisions based on flawed data; so they can expect to achieve flawed results,” said Josh Dreller, director of marketing research at Kenshoo. “Multi-touch techniques better reflect the reality of consumer behaviour and enable marketers to improve budget allocation and campaign performance.”

Kenshoo’s findings echo those of a review of Google US Analytics ecommerce tracking accounts earlier this year, which claimed social media was the second most used channel after display ads in the earlier part of the purchasing process.

It is also backed up by the findings of a recent Aggregate Knowledge report, which found social continues to deliver the highest quality users.

Follow CMO on Twitter: @CMOAustralia, take part in the CMO Australia conversation on LinkedIn: CMO Australia, or join us on Facebook: https://www.facebook.com/CMOAustralia

Signup to CMO’s new email newsletter to receive your weekly dose of targeted content for the modern marketing chief.

Join the newsletter!

Or

Sign up to gain exclusive access to email subscriptions, event invitations, competitions, giveaways, and much more.

Membership is free, and your security and privacy remain protected. View our privacy policy before signing up.

Error: Please check your email address.
Show Comments

Blog Posts

Is customer segmentation dead?

Ginni Rometty, the CEO of IBM, announced the death of customer segmentation five years ago saying, "The shift is to go from the segment to the individual. She might have been a bit premature for most marketers, but if customer segmentation isn't dead yet, it's definitely on life support.

Richard Taylor

Senior digital strategist, Spinach

How people buy brands

Andrew Ehrenberg was a giant in the field of marketing science. He believed scientific methods could reveal law-like patterns of how people buy. In this post, I summarise one of Ehrenberg’s most important discoveries and its implications on how people buy brands.

Kyle Ross

Strategist, TRP

Is artificial intelligence riddled with bias?

The purpose of Artificial Intelligence (AI) has always been to replace the menial and repetitive tasks we do each day in every sector, so that we can concentrate on doing what we do best. Saving time and money has certainly been a decent outcome as AI infiltrates the business landscape, however, now we are starting to see problems that cause major issues in practice.

Katja Forbes

Founder and chief, sfyte

This is so cool & Innovative . A Milestone will be created by this.

Digital Marketing Courses

AANA, IAB and MFA chiefs detail first cross-industry digital advertising practices

Read more

“2019 will be the year brands leverage their social capital with consumers to help drive sales, answer questions, and act on the brand’s ...

Engenius

Predictions: 9 digital marketing trends for 2019

Read more

At the deeper levels of artificial intelligence, computing machines make all kinds of correlations among whatever data is available to th...

Fraction Tech

Is artificial intelligence riddled with bias? - Customer Design - CMO Australia

Read more

https://myiplookup.com/ - find your ip address and location information in our main page. Also there are many ip tools you can use : IP L...

savefrom

iSelect outlines new approach to arrest ineffective marketing as its reports full-year results

Read more

https://myiplookup.com/ - this website will allow you to View Alexa Ranking and graph Check http headers of a website, tool to compare te...

savefrom

The Star's first CMO creates all-new marketing team

Read more

Latest Podcast

More podcasts

Sign in