Troubles Organizing Data Prevent Advertisers From Using Preferred KPIs

April 20, 2018

This article is included in these additional categories:

Analytics, Automated & MarTech | Brand Metrics | Brand-Related | Customer-Centric | Data-driven

Advertisers believe that sales metrics are the most effective in evaluating a campaign’s success, but instead tend to use traffic metrics more often, according to a study [download page] from Bazaarvoice and Digiday that surveyed almost 250 brands, agencies and publishers on the topic of attribution.

Almost half (48%) of respondents said that they most commonly use traffic metrics, such as average time on page and unique visitors, to evaluate the performance of their digital campaigns. That compares with one-third (33%) who most commonly use sales metrics. (Fewer than one-tenth primarily use brand awareness or lead generation metrics.)

But when asked which metric they believe is most effective in evaluating a campaign’s success, half of respondents chose sales metrics – such as conversion rates or return on ad spend – while only around one-third pointed to traffic metrics.

The main culprit behind this apparent disconnect is a difficulty with organizing data. Indeed, fully 8 in 10 said that organizing data from multiple channels is the biggest limitation preventing them from using the KPIs they consider most effective more consistently. Other hindrances such as technology limitations (43%) and cross-channel marketing (40%) are troublesome to some, but are not as widespread.

Despite not using the KPIs they find most effective, advertisers are highly confident that the KPI that their brand or agency partner most often recommends is best for proving success. Moreover, the vast majority feel that they’re very aligned with their partner on delivering consistent and accurate metrics on that KPI.

Last-Touch Attribution Still the Most Used

The survey reveals that marketers and advertisers continue to gravitate towards last-touch attribution. A plurality (39%) said they use last-touch most often to measure digital campaign performance, while 28% use first-touch and just 11% multi-touch attribution.

Multi-touch attribution is more common when measuring offline campaigns, as it’s used primarily by 28%. That’s still slightly outweighed by the proportion using last-touch (30%) attribution for offline campaigns.

Surprisingly, more marketers said they find first-touch (39%) than last-touch (28%) the most useful for measuring digital campaign performance, with fewer (21%) again pointing to multi-touch attribution. And for offline campaigns last-touch (40%) again gets the nod over multi-touch (28%), though first-touch (15%) attribution isn’t considered among the most useful in this regard.

The findings are interesting given that multiple pieces of research have indicated in the past that while marketers still primarily use last-touch and first-touch attribution, these have long been considered the least effective attribution methods.

The report makes the case that the attribution model used may vary according to the goal. For example, a branding objective may necessitate a focus on first-touch attribution, whereas conversion-based campaigns may instead opt more for a last-touch model. And while a multi-touch model “would incorporate a healthy mix of sales and traffic metrics, eliminating the blind spots that arise from focusing on just one or the other… marketers often don’t have the time to focus on everything at once.”

The full report can be downloaded here.

About the Data: The data is based on a survey of 128 brands and advertisers, 69 agencies and 44 publishers.

Chart-Library-Ad-1

Explore More Articles.

Which Skills Are Important in RevOps?

Which Skills Are Important in RevOps?

9 in 10 RevOps professionals view data analysis skills as being important, a high percentage also don’t believe they need this skill for their job.

Marketing Charts Logo

Stay on the cutting edge of marketing.

Sign up for our free newsletter.

You have Successfully Subscribed!

Pin It on Pinterest

Share This