Is Online Advertising on the Way Out?


This week, our Digital and Optimisation Strategist, Michael Fridman noticed an interesting opinion piece on The Conversation, about the death of online advertising. As you can imagine, it’s raised some interesting conversations in the Loves Data office. Read on for Michaels’ response – and defence of – our second love, online advertising.

“Reports of my death have been greatly exaggerated.”  Mark Twain

These days, we all see a lot of advertising and many of us love to hate ads. For those of us working in the online marketing industry, we’re not immune either since we do see a lot of online advertising that’s poorly targeted, exploitative or simply annoying. However, this makes it easy to develop misconceptions when reading articles about online advertising without informed background knowledge.

This forms the source of my disagreement with David Glance’s latest piece in The Conversation, provocatively titled ‘Will ad blocking deal the final death blow to already failing online advertising?

The overall argument is that the advent of ad blockers will pose a serious challenge to companies that rely on advertising for revenue, especially since ad blocking is likely to be more pervasive on mobile devices. It’s certainly true that the display advertising model is not sustainable for a lot of publishers. Whether selling banner ad space on a cost-per-click (CPC), a cost-per-thousand-impressions (CPM) or cost-per-acquisition (CPA) model, it remains a genuine challenge for those publishers to rack up enough clicks (or impressions or acquisitions) to cover their costs. It’s also true that these publishers have been experimenting with a range of other options, including metered paywalls and voluntary subscriptions. Many of these have failed while some have succeeded spectacularly.

Two notable examples are the (recently retired) political blog The Dish by Andrew Sullivan, who was able to use his large audience to raise $333,000 in the first 24 hours of moving to a voluntary subscription; and the educational video brand Crash Course, who were able to move a substantial part of their production costs over to Patreon. This will continue to be an important and developing space in internet culture as the rise of services like Patreon redefine the relationship between content producers and consumers, expanding it beyond the advertising model.

However, the main claim in Glance’s piece that I take issue with is the suggestion that online advertising is failing – having somehow been shown to be both unnoticed and ineffective; and that ad blocking (which is mainly blocking display ads) is likely to have a game-changing impact on (say) Google’s bottom line.

Online advertising for giant networks is not failing and can be seen by the fact that revenues have continued to grow. Example: Google’s (unaudited) Q1 revenues from 2015 show a 14% growth rate from Q1 2014.

Furthermore, the revenues show that most of Google’s revenue (about 75%) still comes from Google properties, with most of it resulting from search advertising. While some ad blockers would block search ads, they are also some of the least intrusive, most relevant and most above-board advertising available online. Any ad blocker that would have a whitelist should be including search ads, and they are also most likely exploring deals with advertising platforms and ad blockers. The high user response rates also testify to the fact that consumers don’t have the same reaction to search ads. Example: many well-managed accounts across diverse industries can attain clickthrough rates of over 10% for non-branded queries.

As to suggestions that even search advertising is ineffective, these have been largely based on some experiments done on eBay’s AdWords accounts, published as ‘Consumer Heterogeneity and Paid Search Effectiveness: A Large Scale Field Experiment’. The full paper is behind a $5 paywall but a description can be found at this Atlantic piece. The paper found that pausing branded keywords did not reduce traffic to eBay; and that pausing non-branded keywords reduced traffic, but did not show a causal decrease in conversion –suggesting that most ads targeted frequent eBay users who were going to buy on eBay anyway.

While there are a number of issues with the paper’s applicability, the main one is that eBay is a very atypical advertiser. Not many advertisers can claim the same scale and scope: in terms of size (100 million keywords is a lot); the perceived low quality of its ads (as reported by a number of industry commentators); and the highly specific behaviour of its repeat customers. While it’s true that there are large advertisers who may not be using their online budgets optimally, there is absolutely no reason to believe that advertising is ineffective for a more typically sized business (or a large business whose campaigns are run with the required expertise).

It’s true that a lot of the standard advertising model (ie. display advertising) would need to be changed for things to become sustainable. And certainly ad blockers may accelerate other more direct forms of revenue including payment for service, as Glance’s article suggested. But if anything, the trends in ad blockers show that the industry needs to get a lot better at advertising as a whole. A lot of the time, the industry is treated as if the only players are the platform providers (such as Google or Facebook), with advertisers being passive participants, simply submitting ad inventory. But as the eBay AdWords experiment shows, context matters, and expertise matters. It is the application of more expertise that’s likely to make advertising more sustainable and balance users’ desires not to be annoyed with creators’ need to be paid and businesses’ need to obtain ROI.