In an interview this week a frustrated sounding Conrad Bird, the marketer responsible for communications spend at Number 10 Downing Street, said ‘I need ROI not eyeballs to show my ad is effective’.
In his case, where every pound of taxpayers money spent on advertising means a pound less to spend on other public projects, proving ROI is an absolute essential, but marketers across all verticals face the same challenge. How to prove that spending a sum of money will deliver the best possible return. How can the technology industry develop a way to make it easier to prove ROI while delivering results which offer better value for the marketing spend?
Digital advertising is spoilt for choice when it comes to metrics and reporting. In real-time it’s possible to report on viewability, audibility, view through rates, click through rates, on-target impressions and even eye and emotion tracking. In comparison to TV where marketers don’t even know if the viewer was in the room during the ad, digital metrics provided a gigantic step forward for brands.
But while useful for benchmarking and providing a useful source of data, these metrics do not prove ROI. Some, such as view through rates which indicate that users were interested enough in a product, service or message to watch a video to the end, correlate to final revenue numbers. Others, such as click through rate have been widely discredited as being a reliable indicator of sales, across many of these metrics there is no clear link for brands to see the value of their advertising.
The metrics which do provide easy, measurable and actionable insights into ROI are purchase intent, brand awareness, product recall etc. Across the digital advertising industry we do sometimes report on these metrics by partnering with a third party to carry out brand studies, which deliver marketers clear insights into brand metrics and how well the creative is performing. The campaigns are still traded on and optimised towards impressions, clicks or views, but the brand metrics are reported on.
The question that remains - if brands want campaigns which deliver ROI why are we trading on and optimising towards impressions, clicks and views? Why are we not, as an industry, optimising towards the brand metrics like purchase intent which do link to ROI?
Thanks to developments in artificial intelligence the technology exists to deliver on these metrics. By carrying out surveys in the same way as traditional brand studies, but surveying users before and during campaigns, it is possible to identify consumers who are the most likely to deliver uplift in brand metrics. In Bird’s case, users who show the greatest likelihood of engaging with the GREAT Britain campaign would be identified in real-time by the AI algorithms and served the ad.
The campaign isn’t targeted to the user most likely to click on the ad, but the user who is most likely to change their perception of Great Britain, accomplishing the overall campaign objective and consequently delivering ROI. AI is self-learning by nature, incorporating information throughout the campaign and refining the system over time.
By not delivering ads to users who are not going to change their minds and upweighting to people who are, the campaign becomes more efficient, wasting far fewer impressions and saving, in this case the tax payer, money.
Digital advertising provides more insights into a campaign’s success than any other advertising medium. As an industry tech providers need to with media agencies to deliver brands campaign results and performance across the metrics which matter to them, and not just those which are easy for us.
Stephen Upstone is CEO and founder of LoopMe