Google’s Services Are Worth One Cent Per Day
If Google moved to a paid-for subscription model then you could get everything for about one cent per day. That has major implications for the future of the web.
Google derived 98.8% of total revenues from advertising in 2005. But that figure has been steadily falling ever since and reached a low of 89.5% last year - and we project that it will continue to fall.
Part of the reason has been Google’s attempt to increase paid-for for revenues, for example by increasing the focus on paid-for features offered as part of YouTube.
But the other part of the reason is more worrying, which is that there might be something fundamentally wrong with Google’s ad-supported monetisation model.
Bearing in mind that last year Google accounted for about 44% of all advertising expenditure on the web, then if this is true the implications would be dramatic not just for Google, but all those who are part of the digital advertising economy.
Let’s take a look at some of the key trends that are shaping the internet advertising economy:
- People are using ad blockers to get rid of annoying ads: This problem has become so serious that in February 2015 Google, Amazon, Microsoft and Taboola agreed to quietly pay the German start-up behind AdBlock Plus to stop blocking ads on their websites.
Regardless, it is clear that a significant and growing proportion of the online population is objecting to the quantity, perceived low quality and aggressive nature of the online ads they experience.
- But the rise of mobile will mean even more aggressive ads: This is a serious problem for Google because the small size of the mobile screen means that it is fundamentally harder to monetise using digital ads than is the case when those same ads are served to PC-sized screens.
As mobile accounts for an increasing percentage of online engagement time, companies like Google and Facebook will be forced to respond with even more aggressive ads.
These companies might think that the rich understanding they have of their users will allow them to serve more relevant and less obtrusive ads, but this raises another problem which relates to the use of personal data.
- More ads and better ads means extracting more personal data, but that data may be being used for the wrong purpose: It is by no means clear that the mass scale harvesting and monetisation by advertisers of very detailed user data will be acceptable in the long term.
Many users do not know that ad networks and advertisers are routinely transferring detailed behavioural data between websites they have relationships with, in order to better target their ads.
While it seems that many users do not care about this, provided the use of their data is not too obvious and not too creepy, I sense a growing problem when that data is being used to serve low quality ads, which are in the majority.
The fact that regulators around the world and especially in Europe are beginning to take action against companies like Google and Facebook, should at the very least suggest that something might be wrong here.
There is a very big difference between (i) a service provider harvesting user data to develop new services, add new features and therefore deliver more value and (ii) where data is being used by unknown third parties to create targeted advertising campaigns and so-called native advertising, which is advertising that has been dressed up to look like content.
Google is trying to do both things at the same time and I think this will become a growing problem because the first type of usage is clearly better than the second.
- Google is trapped in a vicious circle where the higher its revenues the more ads the company needs to serve per user achieve the same percentage growth: Google’s overall user base is no longer growing strongly in developed internet markets, which account for the vast majority of ad expenditure.
Therefore, the only way to increase revenues is to increase the average ad revenue per user, which means more ads, more sponsorships, more product placements and more native advertising.
One implication is that the company is being forced to create Google-branded products in specific vertical markets - like shopping, weather, flights and many more - which are then inserted into search results pages to the likely disadvantage of competitors.
These products are monetised using various forms of paid advertising, sponsorship or product placements – which have the effect of damaging the objectivity of Google Search.
While Google might think these products are superior to those provided by rivals, because Google refuses to use its own search algorithm to objectively rank these new products the company has drawn the ire of rivals and has attracted the attention of regulators in the US, and especially in Europe where the company has been accused of illegal behaviour.
Already accounting for 44% of all online ad expenditure, Google increasingly suffers from the law of big numbers, which means that each year the company will have to serve a geometrically-increasing number of ads to a fixed user base.
Plainly, a point must be reached where there will be simply too many ads.
- Ad fraud and ad-stuffing is growing and destroying value: Advertisers will spend a total of USD 160 billion on their digital ad campaigns this year, but at least 10% of that will go to fraudsters and criminal gangs who are using sophisticated measures to trick advertisers to serve ads to pages that are requested by machines.
Separately, as publisher sites seek to grow their revenues, they are adding more and more ad units to their pages. This has led to many ads being served to parts of pages that are never seen, further destroying value.
While advertisers and ad platforms are working to manage these issues, they will remain an important feature of the digital ad market.
- Digital advertising effectiveness is falling: Research shows that as many as 30% of clicks on mobile ads are made in error – due to the small size of the mobile device users accidentally click on ads and then realising their error, they quickly go back to the page they were viewing. But advertisers have to pay for those accidental clicks.
This problem also occurs on sites that ‘trick’ the user to click on ads as they move around the page.
Meanwhile some sites have so many ads they are barely usable or they might partition a story into segments which force the user to click on ‘next’ multiple times, where each click results in more ads.
The collective impact of these problems has been to instil a feeling of cynicism in the minds of internet users when they use ad-supported services: instead of the user being surprised by an ad which is relevant, interesting or just entertaining, a growing number of users have become adept at ignoring the ads and navigating their way around them.
The ad formats that work best tend to be on specific platforms, like specific social media services, or app ecosystems. Rather than a single consolidated online ad market, the problem of falling overall effectiveness has forced the online ad market to segment into a number of major ad platforms each of which has its own ad units and ad products.
Advertisers are fighting back against these adverse trends with new techniques, like retargeting or switching focus to video advertising, or both. But these options are even more dependent on rich user data which some regulators think should not be used for such purposes.
- Ad-supported webspam is growing and polluting the internet: The widespread prevalence of web content and the easy processes offered by most ad networks make it easy for people to develop parasitic businesses which aim to extract value from the online ad ecosystem while contributing minimal or even no value.
Millions of spammy sites fall into this category and degrade the quality the web experience. All of these websites are funded by advertising.
Any objective review of the above trends would lead to the conclusion that at the very least, Google’s ad-supported business model faces some significant challenges.
But if we dig more deeply we find a fundamental reason why Google’s advertising model is not just challenged, but economically unsustainable.
The use of an ad-supported model to monetise the sort of advanced web services that Google delivers will ultimately limit how fast those services can be improved and the absolute value they can deliver.
Here’s the supporting rationale:
As the digital economy develops, users are spending an increasing amount of time with online media and a decreasing amount of time with offline media.
Because advertisers need to place their messages on the media their target audience is engaging with, this means that advertisers are allocating proportionately more of their advertising budgets towards online media and proportionately less towards offline media.
So far so good - nothing controversial. But let’s now think about how this process plays out in the long term:
We can be certain that, eventually, the flow of ad dollars from offline media to online media must stabilise at some percentage. We can argue about what this percentage might be but for the purposes of this article it doesn’t matter: the absolute level could be 55%, 75% or even 100%.
All that is important is that we accept that there must be a saturation level.
So what happens to the growth of the ad-supported online services industry after this saturation level has been obtained?
The answer is the rate of annual growth of the digital advertising market must fall to equal the rate of growth of the economy (i.e. the rate GDP growth), which is perhaps 4% per year.
The reason for this is that companies set their advertising expenditure as a fixed percentage of sales. Because the whole economy is composed of all companies and all consumers then the annual growth of the total advertising market tracks economic growth which we can assume to be c. 4% per year.
This means that when the digital ad market stabilises, the annual growth of the whole market will fall back to 4% per year and at this point the ad supported online services industry will be a mature industry.
To drive the point home this means that Google, for example, will be a mature company and revenue growth will have slowed to 4% per year.
However, this conclusion is inconsistent with our vision for how the online services industry will develop in the future: we foresee an explosion of innovation and the introduction of an entirely new category of value-added services oftentimes enabled by artificial intelligence. And we can already see the first signs of such services coming through in the form of Google Now, Apple Siri and Microsoft Cortana.
It seems clear that the incremental value delivered by the this new category of services and the annual growth of the markets they will enable, will far exceed what a c. 4% annual growth in revenue will support.
We should be looking for 20%+ per year in some segments of this new market, not 4% per year. Remember that Google already accounts for 44% of total digital ad expenditure so it is not as if Google could – through innovation – accelerate its growth to 20% per year, because it already accounts for too large a proportion of the total market.
Hence, if Google sticks with its ad-supported model in the long run then this will not just result in a service portfolio that is so loaded with ads that users seem likely at some point to defect to rivals, but the company’s rate of investment in new services will slow down which will accelerate the flow of users to rivals.
Taking all of the above factors into account, the conclusion is that ad-funded online service providers like Google will have to find a way of introducing a direct payment model and that this model must become the primary source of revenue.
This has implications for the whole web, which I’ll return to in a future article.
To be clear, the conclusion is that Google will have to shift from being funded almost exclusively by advertising to a new model where the primary source of revenue will be premium subscriptions.
In the figure below I have looked at Google’s actual advertising revenues for the last three years for the US and the rest of the world and compared those figures with the number of individuals who actively use Google’s services.
I then divided Google’s achieved advertising revenues by the number of active users of Google services to derive an equivalent monthly service charge which would return the same total revenue figures:
As can be seen in the above table, on a worldwide basis all of Google’s services could be theoretically provided completely ad-free for the approximate retail price of about one cent per day.
Specifically in terms of the US, the monthly service charge would be about $10 per month.
Just to be clear, this would include YouTube, Google Search, Gmail, Google Shopping, Google Finance – all of Google’s services – as they currently exist, but with no ads.
It is very interesting that this hypothetical service price is, for the US, similar to the prices that apply for subscription-based online music (e.g. Spotify) and subscription-based online television (e.g. Netflix).
I suggest that it is at least probable that this is not a coincidence.
Perhaps each of Google’s 100s of services is analogous to one of the many content items that are included as part of a subscription-based service like Netflix. Perhaps Google’s role in the future will be to ‘open up’ its platform to third party developers who can then create and distribute their own niche services to Google’s users: the more users that are attracted by an expanding number of services, the more developers will be attracted, and the more developers the more users and so on.
The idea of Google shifting from an ad-supported model to a paid subscription model seems like science fiction, not only because of the damage that would be done to thousands of companies whose businesses could not operate without Google AdWords, but also because of the impact on advertisers who would have to stand and watch as a huge door slowly closes shut, sealing them off from their target audiences.
But in the technology industry three things are clear: science fiction is the destination, 'never' is a very long time and the word ’impossible’ suggests an opportunity, if not for Google – then for someone else.