Companies in the UK, France and Germany are more likely to decrease (46%) than increase (28%) spending across digital channels as a result of attribution, per results from an Econsultancy report [download page] produced in association with AdRoll. However, this masks distinct regional differences, with the opposite pattern seen in UK.
Indeed, UK company respondents were more likely to say that attribution has resulted in increased (36%) than decreased (22%) spending across channels.
There were several instances in which the UK results were found to be quite different than the results in France and Germany. This may be partly due to the survey methodology, in which third party panels were used to supplement only the French and German samples.
Nevertheless, averaging out the responses across countries offers some insight into which digital channels have seen increases and decreases in budget as a result of attribution.
Paid search has been one of the primary beneficiaries, with a majority (51%) of respondents saying that it has seen increased budgets due to attribution, twice the share (25%) saying that budgets have reduced due to attribution.
Social media budgets have also seen sizable gains due to attribution. Respondents were roughly 4 times more likely to say they had increased than decreased both social media marketing budgets (47% vs. 12%) and social media advertising budgets (40% vs. 11%).
By contrast, more respondents had decreased (54%) than increased (40%) their display budgets as a result of attribution, with affiliate marketing also seeing more decreases (41%) than increases (33%).
The extent to which attribution affects digital budgets also depends, clearly, on whether those channels are included in attribution models. For company respondents, email (71%), display advertising (64%), and content marketing (58%) are the most commonly-included, though a majority also include paid search (55%), social media marketing (54%) and affiliate marketing (52%).
By comparison, mobile apps (25%) and video (24%) appear to be rarely included in attribution models.
While almost three-quarters of company respondents believe that a perfect attribution model is impossible, they clearly favor some over others. The most popular models used are last-click (48%) and first-click (47%), with post-click (35%) slightly trailing. No other model cracked 25% adoption, including algorithmic (23%), first-touch (23%), last-touch (19%), and view-through (9%).
While last-click was the most commonly used model, it’s seen as one of the least effective: just 29% of users perceive it to be very effective. Things don’t seem to change very much: in 2012, a survey from Econsultancy and Google Analytics found that last-click was both the most popular and the least effective attribution model in use.
The most effective models, per respondents to the latest study, were algorithmic (59% rating very effective) and post-click (53%), with few deeming view-through (32%), linear (31%), and time decay (21%) to be highly effective.
About the Data The report is based on a survey of 590 respondents in the UK, France and Germany, of whom 75% are in-house marketing professionals and the remaining 25% supply-side respondents including agencies, consultants and vendors.
The retail (18%), financial services and insurance (12%) and technology (11%) sectors were the most heavily represented. Respondents were fairly evenly split between B2C marketing (36%), B2B marketing (39%) and B2B and B2C equally (25%).
The majority (51%) of company respondents come from companies with more than Â£150 million in annual revenues.