- Online advertisers should be able to monetize their ads more effectively in the future. This will be a huge tailwind.
- Scale will be a moat.
- Youtube has a bright future ahead of it.
- I suspect that television advertising and ad agencies will face serious headwinds.
In online advertising, one of the most important things is the ability to target ads accurately. Search advertising monetizes incredibly well because:
- The search query provides excellent information about what the consumer is interested in buying. If the search query contains the two words “web hosting”, there is a good chance that the consumer is about to buy web hosting. If the consumer is searching for information about mesothelioma, it is likely that they may want the services of a lawyer for an asbestos lawsuit.
- People use search engines to research their purchasing decisions (e.g. reviews, to find the lowest price, etc.). This is the best time to advertise to a consumer.
Display advertising (banner and text ads on a website) does not have the ability to target ads so accurately. It does not monetize as well as search advertising. Google’s 10-Ks suggests that search advertising generates significantly more ad dollars than display advertising.
In display advertising, most ad targeting was traditionally performed by analyzing the contents of a website. If a webpage contains a lot of words relating to web hosting, the ad network will likely show ads relating to web hosting. A new and different approach is ad retargeting. To the consumer, it appears as if ads “follow them around”. After they visit a particular website (that specifically installs a tracking pixel to enable retargeting), the ad network will understand that the user is likely interested in buying a particular product/service and will continually display ads for that product/service whenever the consumer visits unrelated websites. I believe that this technique greatly improves the monetization of display advertising.
Big data and analytics have allowed retailers to discover interesting things about its customers. The New York Times has an excellent article about how Target figures out that its customers are pregnant. Target uses this information to send relevant coupons to its pregnant customers. There are privacy concerns to doing this. One father accidentally discovered that his daughter was pregnant after noticing that Target was mailing pregnancy advertising to his daughter. Target now throws in unrelated advertising so that its customers won’t know that Target knows that they are likely pregnant.
In the future, I suspect that Google will find ways to do similar things. Based on a user’s search queries, Google can likely determine a wealth of information about its users. I suspect that it will find new ways of improve its ad monetization based on these insights.
Google users can look at what Google thinks about their interests by visiting https://www.google.com/settings/ads. Google currently guesses:
- Gender. This is based on a user’s Google+ profile, so Google will guess incorrectly if the user lies about his/her gender (which I have done).
- Age demographic. Again, based on a user’s Google+ profile.
- Languages. For some reason, Google did not correctly guess my language.
- Interests, based on websites visited. Google is pretty good at guessing my interests.
Google acquired this technology via its DoubleClick acquisition in 2007. To be fair, this was a very long time ago and it seems like this technology hasn’t done that much.
Scale will be a moat
Advertisers will spend more time optimizing their campaigns for ad platforms that drive the most traffic. Rational advertisers will do this because they want to maximize their overall returns. Time spent on optimizing Google search ads will likely yield ~9X the benefits of optimizing their search ads on Bing (assuming 9:1 ratio of market share). Because of this dynamic, some small advertisers don’t even bother with Bing. The focus on Google means that advertisers spend their ad dollars more effectively and see a higher ROI on their ads. Because of this, they will bid higher on ads to generate more traffic and to maximize their profit. Because Google is the largest search engine, it will receive higher CPM rates for its ads.
On the search side, scale isn’t that strong a moat because search users simply want to use the best search engine. If another company makes a better search engine, then Google will lose market share.
On the display advertising side, scale is a much stronger moat. The customers of ad networks are content creators who want to monetize their webpages effectively. Because Google Adsense is the largest display network, it can offer content creators the highest CPM rates. This helps Google maintain its #1 status in display advertising. I think that it will be very difficult for competitors to displace Google.
Where display advertising sits on the monetization totem pole
On the other hand, display advertising is one of the least effective ways of monetizing a website. Some bloggers will say how much money they make from Google Adsense (or other display advertising) versus other monetization options. Where there are multiple revenue streams, Google Adsense is often the least effective monetization option. Generally speaking:
- Selling a product or service directly will generate the most money.
- Driving sales to somebody else who sells something is the next best thing. A content creator can put affiliate links so that they receive a commission for each sale that they generate.
- Sponsorships / direct relationships with advertisers can also make money.
- It used to be that selling links to help others game search engines would also be an effective way of monetizing a website. I don’t know if this is still a thing.
The reason why many website owners choose to run display advertising is because the other monetization options take a lot of time and effort. Because Adsense takes virtually no time to setup, getting paid a lot of money for very little work is not a bad thing.
For video content, content distribution companies have already figured out that making users pay for content is the most profitable way to monetize content. When a Hollywood movie is first released, it is monetized through movie ticket sales. Then it goes to home video, video on demand, and other forms of users paying for content. After that, a movie might appear on cable (supported by subscription fees + advertising) or on a broadcast network (supported by advertising).
Traditional television advertising
Nielsen measures the demographics of each television show. Based on this, advertisers can target ads based on products and services that the demographic would be interested in. Advertisers would also understand that higher-income individuals have greater purchasing power and are more valuable.
The problem I see with television advertising is that targeting is poor. Niche advertising will never work. A mesothelioma/asbestos lawyer cannot effectively advertise on TV whereas they can effectively advertise online with search and display ads. As well, a lot of ads will reach people who aren’t interested in that particular product or service.
On top of that, traditional advertising is extremely difficult to track. In the online world, advertisers can get a very good idea as to their campaign’s return on investment. If the advertiser sells their product or service online, they can directly measure how ad clicks translate into completed purchases. This is incredibly powerful. It allows advertisers to bid aggressively on ads to maximize their profit. They can rapidly figure out what works and what doesn’t, which increases the effectiveness of their campaigns.
Television advertisers are left guessing as to their ROI. My personal suspicion is that many corporations overpay for advertising in the same way that publicly-traded companies overpay for advice from investment bankers or waste money on consultants. Many corporations may have negative ROI on their TV ads. Their behaviour can persist because they simply aren’t sure what their ROI is. It is easy for a CEO to piss away money on wasteful advertising especially when ad agencies have a vested interest in encouraging them to do so. During recessions, television advertisers cut back on their advertising budgets. If they knew what their ROI was, I doubt that they would lower their advertising spending to the degree that they do.
My theory is that online advertising will be extremely compelling compared to traditional advertising. Once companies figure out that they can track their advertising’s effectiveness, I think that they will reconsider the effectiveness of television advertising. They may (or may not) begin to reconsider the effectiveness of a traditional ad agency’s services. Ad agencies will not be able to get away with hocus pocus hand-waving about how they will supposedly make money for their clients. If customers start paying attention to results rather than an ad agency’s mystique and reputation, the agencies’ business models may suffer. Of course, there are some people who can truly create value through advertising. However, these people can leave and start their own firms. Eventually corporations will realize that an ad agency is an intermediary that doesn’t create a lot of value. Corporations can hire talented individuals directly. Because shareholders in an ad agency don’t own the people, such a shift could hurt the economics of publicly-traded ad agencies. (Of course, I could be very wrong about this.)
To learn more about ad agencies, you can read books such as David Oglivly’s Confessions of an Advertising Man. He talks about the importance of advertising that makes money for the client and how he likes people with a direct response background (because in direct response, you quickly learn what works and what doesn’t). From reading his book, I got the sense that tracking the effectiveness of traditional advertising is incredibly difficult (except in direct response).
What actually works in advertising
Online advertising allows advertisers to quickly figure out what is and isn’t working. There is a huge wealth of information available as advertisers can track a user’s clicks and see how they traverse the advertiser’s website. As well, advertisers can split-test their ads and test new ways of improving their ads and their landing pages.
Online advertisers have figured out a lot of surprising things about advertising that aren’t obvious:
- (Fake) independent reviews. The core idea is that “independent” review sites have more credibility with consumers than somebody trying to do a hard sell. On top of this, an effective review site will tend to oversimplify differences between products into a number or star rating. This is because users prefer simple answers to complicated buying decisions.
- Simple messages tend to work better than complicated messages. One unique selling proposition is enough.
- Affinity. Consumers have slightly greater trust for people who are from their city. Because a user’s IP address reveals his/her location, a landing page can say that the fictitious blogger who “created” the site is from the consumer’s city. An ethical application of a user’s location data is to have the ad specifically mention a company address if it is close to the user.
- One weird trick to _____ (e.g. lose weight, get a six pack, etc.). These ads grab users’ attentions and gain some credibility because they don’t do a hard sell. Sleazy marketers use this format because it works.
- Is ____ a scam? / The truth about _____ exposed. This is a variation on #2. The ad will suggest that a certain product or service might seem like a scam but really isn’t a scam.
- As mentioned by (insert name of credible news outlet). This builds credibility.
Looking at the big picture, it is obvious that there are many ways to approach advertising. Online advertisers can quickly iterate through their ideas to optimize their ad campaigns. Doing so allows them to bid more aggressively on advertising. I believe that this will drive higher CPMs for online advertising over traditional television advertising.
Obtrusiveness of ads
Youtube allows users to skip some of its ads. In my opinion, this technique creates value. If you look at search advertising, what you see is that a tiny handful of search queries make almost all of the money. A query for “web hosting” will typically pull up a search result page flooded with ads, whereas most typical search queries on Google don’t have any ads. (Google intentionally chooses to show less low-value ads to deliver a superior user experience to hopefully gain more search users from its competitors.) Allowing users to skip ads doesn’t hurt advertisers much and delivers a better user experience.
In the future, ad networks may even figure out what ads users do and don’t like. Based on whether or not a user skips an ad, an ad network can figure out which ads a user is likely to watch based on other users’ behaviour. I believe we have the statistical tools to do this. For example, Netflix and Amazon (and many other websites) currently have effective recommendation engines. Perhaps the same ideas can be used to serve more relevant ads.
All online advertisers have this advantage over traditional television advertising.
This makes sense because advertising is a means to an end (selling a product or service to consumers).
In my opinion, Google has stumbled across a phenomenon that will (sometime in the future) be incredibly profitable. Because Youtube is the leader in what it does, its popularity is a wonderful moat that will protect Youtube from competition.
In the past few years, ad targeting on Youtube hasn’t been very good. I’ve seen ads in French while watching English-language videos on Youtube. Google clearly has the technology to fix this problem. Its Chrome browser and search engine have the technology to detect the language of a webpage. When Google starts putting serious effort into monetization, I think that the economics of Youtube will improve dramatically. We know from Google’s other products (Google Analytics, Adwords, Adsense, and Google affiliate links in search results) that Google has a lot of sophistication in ad monetization.
- Improving ad monetization will encourage more content creators to make Youtube videos. More and better content will translate into higher viewership. This will be a virtuous cycle.
- The cost of delivering video to users will come down as technology improves. Eventually Youtube will not need to increase bandwidth past 4K because it will have hit the limits of human perception (see my post “What happens when technology is good enough?”). This will improve margins.
- As a video distribution platform, Youtube is more compelling that broadcast or cable. In the future, Youtube will have the capability for 4K resolution and the possibility for quality that is superior to HD cable. More importantly, Youtube is an on-demand platform that allows users to watch video whenever they want without having to program a PVR. And lastly, Youtube offers access to ‘more channels’ than traditional broadcast and cable. This gives users access to a vast universe of content. Youtube (and other online video platforms) will ultimately have a technical advantage.
- More consumers will have high-speed Internet.
- The ‘television’ world may reach an inflection point where Youtube monetizes better than traditional advertising-supported TV. This may cause a dramatic increase in Youtube viewership as content shifts onto Youtube.
How much will Youtube be worth in the future? I don’t know. I suspect that Google’s search advertising business will always be more profitable than its display advertising businesses.
An investing perspective
For display ad networks, I would be most interested in owning the #1 platforms in display advertising. For online advertising, that is Adsense and Youtube. I’m less interested in owning smaller players as they are vulnerable to competition. Therefore smaller players deserve lower earnings multiples.
*Disclosure: No position in GOOG. I may or may not buy its common shares or call options in the future.