Ban branded search through trademark law
Branded search is an IP problem pretending to be a marketing opportunity, writes Mutiny's Henry Innis.
I saw a post by a fellow colleague in the econometrics space, Dr Grace Kite, targeting brand search. In it she lays out the case (eloquently) against branded search. She suggests brands are largely bidding against their own trademarks.
Grace is right.
And of course regulators have completely missed this in their ACCC digital review, which confused the issue further. Instead of focusing on Big Media’s rent-seeking arguments against Big Tech, they should have looked at the real issues in how the technology is sucking money out of the ad market.
There are two major issues related to the money drain for marketers: firstly, how technology companies are forcing advertisers to pay for their own trademarks and secondly how tech companies have leveraged their dominant position in advertising to create distortions in the measurement market. Both are linked (and the latter is why we started Mutiny).
Forcing advertisers to pay for their trademarks is important, simply because it sucks money out from more conventional advertising (in a way that is parasitic, moreso than innovative). Much of branded search is defensive and therefore preventative, and it’s hard to ethically justify when used to directly bid against trademarks and brands someone has created and protected.
If you accept discoverability on the web is monopolized through search, then it stands to reason someone shouldn’t be more discoverable on your own IP than you are because they pay the monopoly. That is a precedent we shouldn’t want, and I believe most people wouldn’t believe is fair.
We deal with this all the time with logos. You can’t make a logo super similar to Cadbury, with the same purple logo, and place it next to Cadbury on the shopping shelf. It’s trademark infringement. Yet the same rules we’d apply on a shopping shelf just don’t seem to carry through on search, despite it being the shopping shelf of the internet.
Branded search can, in the right context, deliver incremental sales. But I’d say it’s in some cases, not all, and in some cases mostly because it is contributing to demand generated elsewhere from advertising.
It facilitates demand rather than creates it.
The why, as Grace says so clearly, is simple. If you generate demand for a product, or for that matter a category, and you don’t advertise in branded search you leave yourselves vulnerable to competitive poaching. You lose out particularly if your competitor bids against your terms.
Let’s look at the numbers involved
First, some numbers. Within WarChest I’d estimate (varying on category) some 30% – 60% of budgets are spent on branded or competitor search. It’s a huge part of Google’s revenue – convincing people to bid against each other and on each other’s (largely) intellectual property.
Taking an enterprise sized marketing organization’s search budget being at $1m – $2m in Australia, that can mean some $1.2m being spent in this area.
I’d estimate there are probably 1,000 organisations averaging more than $1m in search spend across Australia. Maybe more. It’s very rare we see someone without it.
That’s around $600,000,000 annually spent in branded search.
Anyone reading this would probably realise this dwarfs what Big Tech is likely being forced to pay Big Media under the poorly executed ACCC regulation.
It’s a huge economic problem.
By our estimation, branded search is a mixed bag for effectiveness (I’m basing this off the over $50m in search spend we have in-platform over a number of years).
Branded search can work in consideration categories. When we combine branded search or competitive search (which works better), it can perform some 40% – 50% less than generic terms, despite gaining the lion’s share of spend.
Branded search isn’t just taking money out of the market. It’s also not driving the big returns marketers believe it is.
So what should we do about it?
If you’ve read this far, you’ll realise the key beliefs that I hold in this area:
- Branded search can be effective, but it’s ethically wrong for marketers and Big Tech to be doing. Bidding on someone else’s IP is akin to bidding on their hard work and poaching at the end, something we stop people doing with misleading logos but seem to allow in the search space.
- Financially it’s a problem. Whilst marketers spend in branded search, there’s evidence it drives less incremental revenue and performance than other search terms in econometric modeling. I’m sure I’m not the only person who’s picked this up (Willem Paling during his time at IAG was extremely vocal on search).
To solve this challenge we need to differently about the regulation around search engines.
Firstly, it’s clear that we should regulate advertising on straight up brand keywords. You shouldn’t be allowed to advertise on “Bupa” or “HCF” by way of example. If I am placing ads with inciting offers as the first ranked result on Bupa, it’s a clear subversion of the user’s intent, particularly if a competitor is bidding. Again, to use the shopping analogy, we wouldn’t let the same thing happen on a shopping shelf.
Secondly, there should by regulation around high intent terms. The simplest example is “buy Bupa”, which again, should be regulated. I’m sure there are 5-6 terms in similar veins that a common distribution style terms that are being mined for immense dollars by Google and Bing.
The impact of removing this kind of advertising would much better for publishers than the ACCC legislation. We would free up significant ad dollars away from brands essentially defending trademarks, and that money likely goes into ad products that persuade and influence purchase decisions.
It also benefits another Google division massively – that of YouTube – which benefits from search data, but does so in a different context (serving video ads against content, rather than high intent areas).
And it would get us away from defending a brand’s IP by gaming search algorithms, and instead focus us on what matters — creating great marketing for great products.
Henry Innis is a managing partner at Mutiny which runs WarChest, a SaaS analytics platform that solves all of these annoying questions like “does branded search work?” and “should I invest more in this channel?” He can be found on Twitter, LinkedIn generally being opinionated and quoting data at anyone who’ll listen.
Interesting piece and hard to not agree. 2 things to add:
1/ trademarks used to be able to be restricted by Google, 14+ years ago. Brands who were Google clients could specify to Google a group of businesses banned from bidding on trademarks. Was very easy and meant most money was spent on category terms. This ended around 2008 around the GFC, brand terms became fair game.
2/ you may find most businesses (or many) buy brand terms not as protection but as a form of “best practice” that response rates are meant to be better if you have #1 paid return and organic return. This was pushed heavily by search engines in the mid 2010’s and endorsed by agencies and clients. There’s a lot of info around publicly making these claims. Brand terms are also a good way to juice perceived roi as when the results are aggregated it can make the campaign seem more effective (as brand terms have high response and seemingly low cost)
But yes agree, they are effectively a tax where you now rent results from search engines. Made worse by mobile where majority of above the fold/in screen content on results return is advertising.
User ID not verified.