I recently attended a talk by Louise Richardson summarizing the key themes of her recent book, What Terrorists Want: Understanding the Enemy, Containing the Threat. Richardson argued that we would be far better off identifying and exploiting the key vulnerabilities of our enemies rather than using raw military force in our efforts to destroy terrorism. She wasn’t saying that using the military was inherently wrong, but rather that HOW we are using it is ineffective.
She used the story of the Peruvian government’s battle with the Shining Path as an example of how we can more intelligently fight our battles. After tens of thousands of deaths in battles between the Peru military and Shining Path rebels, the government of Peru set up a 70 person intelligent unit focused on bringing down the Shining Path. The intelligence unit concluded that the weakness of Shining Path was its highly centralized organizational structure under leader Abimael Guzmán. The unit set off to decapitate the leadership with the assumption that after it did, the movement would collapse. After much research on Guzman, it was discovered that he had psoriasis that required treatment with a certain skin cream. Once the unit had that information, they set out to find out where the cream was sold and tracked Guzman down. After Guzman was arrested, the Shining Path collapsed.
The morale of Richardson’s story was to understand the enemy’s Achilles’ heel and then use a very focused approach to bring him down.
New medium, same old advertising model.
It’s critical to understand that today’s online advertising business grew out of the traditional media world (broadcast, radio, print). In that world, supply was concentrated in the hands of a few media companies. Prices were very high as a result of limited supply and the cost of producing a campaign, managing performance, and producing analytics to measure performance were (and still are) very expensive. Media companies had high-price direct sales organizations that had to meet minimum revenue thresholds per deal in order to operate profitably. And “measurement” was brand awareness studies conducted in a few markets based on surveys before and after a campaign. The bottom line — only firms with extremely large discretionary marketing budgets could afford brand advertising. Basically, the Fortune 1000.
Internet advertising extended the traditional media model despite the clear differences in the medium. Supply was no longer concentrated in the hands of a few firms, but rather thousands of firms. The internet is interactive, so it was now possible to track at a far more granular level (first click-through rate and now very detailed purchase data by having the advertiser drop a pixel on their checkout page to track conversion). And we finally had a way to make the experience self-service, so that the costs of hiring an agency to develop the creative, buy the media, and run the brand awareness studies could be made self-service. Yet, most sites used the same old model. Why? That’s the way it was done before and there was a massive industry that had a great deal to lose from change.
What is the Achilles’ heel of the agency-media complex?
The existing infrastructure of the industry is not well-suited to serve everyone other than the Fortune 1000. There are about 5.7 million firms with payroll in the United States. While the top 890 firms (based on employment) account for 30% of U.S. payroll, there is a significant business to be done with the rest of the 5.7MM firms not in the top 890 firms.
The problem is, it’s hard to serve and reach those firms. And that’s the opportunity. Why is Google AdWords so powerful? Sure, search is incredible in that a user tells you what she is looking for right now. But note these key characteristics about Google Adwords:
1. It is self-service.
You don’t need to know or call anyone to get started. Just go here and get started. You will need a credit card.
2. The creative costs are virtually zero.
You don’t need a creative team to do a huge amount of research for you to test messaging. You can run as many experiments as you would like and AdWords will basically run extremely cheap A/B tests for you. In fact, this is how I test messaging for key marketing messages cost effectively. And since everything is text, you don’t need a visual design team.
3. The advertising aesthetic matches the current page aesthetic.
Google’s advertisements look and feel like search results. This is less jarring to a user, especially when the content is also relevant.
4. Pay for performance = Self-funding marketing.
You only pay when someone clicks on your advertisement. And Google provides free analytics that is integrated into the Google campaign management system.
Why does advertising take such a big hit during recessions? When big companies make cuts, they cut discretionary spending. Since Fortune 1000 marketing is discretionary (it’s too hard to understand the link between advertising spend and revenue), it gets cut. Small businesses dependent on cash flow have tiny discretionary budgets. But with Google AdWords, you can figure out ROI much more easily. So with each dollar spent, you can spend more. Now that’s viral marketing. All of a sudden, millions of businesses which relied on an annual Yellow Pages advertisements and word of mouth are big marketers.
In 2007, Google earned $16.6 billion in revenue.
While Google is impressive, there are major issues.
There has been too little competition, so it seems like innovation has slowed in recent years. And there are some real issues. Google’s ad systems work well with search, but non-search based advertising is pretty dismal (e.g., content matching). Google acknowledged as much on a recent conference call about their disappointing performance on MySpace. Many of my friends working on content and social media startups have been extremely disappointed with the performance of AdSense on their sites.
While the “traditional” professionals overplay the “search marketing is great for direct marketing, but we are about brand advertising” argument, there is uncertainty about evaluating its impact on “considered purchases.” However, this is as true for existing brand advertising methods are it is for Google or Yahoo! search marketing.
New opportunities.
First, any new business should obviously be self-service. That’s not a business opportunity, but rather a basic requirement. However, the other three points may expose new business opportunities.
1. Eliminate creative costs.
Spot Runner applied this idea to TV advertising. Brilliant. There is an opportunity to do the same for the web (I’m sure there are plenty of examples already out there). Imagine a template system that allowed an advertiser to use some simply GUI tool to build an interactive display advertisement. And then A/B test variations to build a few final advertisements.
2. Have the advertising aesthetic match the current page aesthetic.
The average display advertisement ignores the context that a user is in at the time they consume the advertisement. That jarring experience may actually result in negative brand value (no data on this, just a personal feeling). So how might we fix this?
First, allow for advertisements to dynamically adjust their aesthetic to match the CONTEXT of the page where it shows (e.g., grab colors from the current page to form a pleasant color palette, adjust your fonts). Next, extract information to adjust CONTENT if possible. In addition to content on that page, perhaps exploring links TO and FROM that page (and so on) will help you infer more ABOUT that page. Then adjust your advertisement content accordingly. TechMeme meets advertising.
3. Better measurement.
It’s amazing to me that better measurement systems do not exist for “brand” advertising. If advertisers could understand what works and what doesn’t work, spending would be much larger. According to Morgan Stanley, the combined broadcast ad spend ($44B in 2007), cable TV ($27B in 2007), internet ($21B in 2007), radio ($20B in 2007), and outdoor ($7B) accounts for a healthy $112 billion. Impressive until you contrast that with the $287 billion from direct telephone ($110B), promotions ($116), and direct mail ($61B). And that’s excluding the $16B on Yellow Pages and $46 billion on newspapers / classifieds which arguably belong in the second category. The difference between those categories is accountability.
Make brand advertising measurable (and thereby, accountable) and you will have a huge business. For the online bit, integrating things you can track like promotions will improve near-term purchases. For considered purchases (things that are infrequent purchases), surely there must be a better model (inference models based on long-term purchase data)?
16 responses so far ↓
Scott Fitchet // June 4, 2008 at 3:20 pm |
Do you still have a comment feed?
Some of your previous posts had readable comment threads but they get buried quickly with the blog format. I don’t think a message board is a good idea, but maybe some of the newer feed readers make historical thread tracking/following/commenting easier.
(If not then maybe they should!)
Oliver's Stuff » It’s amazing to me that better measurement systems do not exist for... // June 4, 2008 at 4:06 pm |
[...] belong in the second category. The difference between those categories is accountability. — Reinventing online advertising: Exploiting the Achilles’ heel of the agency-media complex « Laser… View the forum thread.blog comments powered by Disqus June 4th, 2008 / Tags: measure, [...]
Oliver Thylmann // June 4, 2008 at 4:08 pm |
Great article, still need to digest, but biggest problem is the following:
If you having a measurable brand advertising system, then this is only a huge business if it provides higher revenues for the publishers than the unmeasurable part. And that is where the problem comes in.
This is actually what MS is trying to do with their long term tracking (you might have bought via a search on google but you saw the branding ad before on cnn.com)
Mike Speiser // June 4, 2008 at 6:29 pm |
You can get the post feed here:
feed://laserlike.com/feed/
And the comment feed here:
feed://laserlike.com/comments/feed/
darrylxxx // June 4, 2008 at 9:45 pm |
Great article, many interesting points, really appreciate it. Darryl
John Furrier // June 4, 2008 at 10:00 pm |
Great to see a post that is so relevant and provocative. It is this kind of thinking that is important to embrace because the future is unwritten and most social media and advertsing programs are shooting in the wrong direction. Your post certainly provides the direction.
I’m been yelling from the twittertops and my blog ( http://furrier.org )about the future of online advertising. I’ll add this to the list of posts to watch.
Thanks
Rolly Rouse // June 4, 2008 at 10:57 pm |
Awesome post.
I’d suggest an addition to your excellent list of opportunities: make advertising more useful and welcome to consumers.
In the long evolution of advertising, a key underlying shift is from intrusive push to consensual pull. That is from mass advertising that’s unwanted and wastes your time to responsive advertising. To ads that are helpful and valuable to you personally (and I don’t mean behavioral targeting).
Google took us all a big step forward in this evolution. You type in what you are looking for. You get ads that match.
As you point out, Google’s ads and their free, “organic” content complement one another. Their ads really do help you find what you need (although most people I know claim they never click on them).
It’s important to note that Google didn’t create the foundation for its ad revolution by focusing on advertising. It did it by focusing on the consumer.
Google is, and has always been, intensely passionate about the consumer experience. That’s a big part of why they continue to produce breakthrough results.
Integrating new kinds of user experiences with new kinds of advertising (more targeted, more measurable, and more welcome) remains a vein of opportunity that’s barely tapped.
Self-service, radically-simplified tools for ad creation, and better measurement are – as you point out – all critical ingredients. But the new opportunities extent far beyond display ads.
Indeed, they involve rethinking the structure and presentation of the content itself. Content that’s needed to support both the advertising and a great underlying consumer experience.
Consumers are already crying out for new innovations in how ads and content work together.
Geoff Whitlock // June 5, 2008 at 1:15 am |
Great article, I agree with many points – But I will highlight one in particular.
The loss of creative in advertising.
The western world is built on creative advertising. I am not saying that this is right, however, we willingly created an Ad Culture.
Our side of the planet and some of the other, generally go to advertising and marketing to discover everything and answer all our questions.
To give that a comparison, many countries rely on their gods or thier governments for messaging and information. Well, advertising is our god (at least the god that speaks to us all).
In many, if not most cases, creative advertising is needed to get the message across to the right audience. Google ads have nice websites with well designed shopping carts behind them. The ads themselves need to be creatively crafted to cut through the barrage in almost every category.
There are millions of simple creative advertisements that have been made for next to nothing that have vaulted companies to heights never expected both online and offline.
Remove the high costs for creative, not the creative element itself.
http://www.lifecaptureinc.com/blog
jeremyliew // June 6, 2008 at 4:16 am |
Mike,
In an ideal world I would agree with you, but there is just so much friction and inertia in the ad world that some of the timelines on this advice might be longer than we’d imagine.
Take self service for example. It seems obvious, but Google (who know more about how to make this work than anyone) have failed to get self service ad sales to work for local ads, radio ads, TV ads, basically ever product extension that they have tried. Even pay per call (as close to pay per click as you can get) hasn’t been able to get advertisers to self serve at scale.
I think the issue is tied to measurement and pay for performance. For online search advertising, because users are pretty deep in their buying process, you can easily track all the way through to a transaction. For anything further up the purchase consideration funnel it gets impossible to track an advertising instance to the action transaction, and at that point you don’t have pay for performance, so self service breaks down.
Your point on make brand marketing measurable is the right answer but you’ve skipped over the difficult bit of this process. There are entrenched key players in the ad industry who actively don’t want this to happen, and you need an industry consensus around any acceptable standard. Therein lies the problem
Mike Speiser // June 6, 2008 at 5:41 am |
I understand the challenge, but I don’t buy this argument:
“For anything further up the purchase consideration funnel it gets impossible to track an advertising instance to the action transaction, and at that point you don’t have pay for performance, so self service breaks down.”
More difficult than direct/CPC advertising (like AdWords), yes. Impossible, no way. With online display advertisements, why can’t you:
1. Drop a pixel on the advertisers checkout page (for any online purchase, for offline collect lead information — e.g., pass on auto lead to a dealer). The pixel will allow you to correlate who has seen what and what actions they took thereafter. In fact, some very smart NYC based firms are buying traffic from large publishers who don’t know how to manage their inventory, dropping pixels on check-out pages, collecting cross-site data through Revenue Science, and selling the traffic at a much higher price. How can these small firms buy cheap traffic (from those who have huge volumes of consumer data) and sell it for a higher price? And do it consistently? True, much of this is focused on shorter-term purchases, but it shows the terrible state of the display advertising industry today.
2. Run bucket tests to get further granularity. Using inference (not perfect. but better than 90% of what’s out there today), you could even figure out what creative is leading to good outcomes versus bad outcomes if you had a large enough audience.
3. When offline media like TV finally joins in (federating your set-top box ID with internet IDs), you can extend your inference models to determine exactly what TV ad led to a purchase. Over time, we will be able to know the probability of a consumer who watched an Audi ad today resulting in the purchase of an Audi in 18 months…
Much of this technology exists today and you could likely increase CPMs by 5-10x by taking advantage of it (just the online bit). The broader problem, as you state, is inertia. And that’s why there are so many small entrepreneurs making millions of dollars feeding on the ridiculous state of the market today.
To solve the problem of inertia, perhaps Google will just go buy a large US-based ad agency, cable company, outdoor vendor, and radio business and show the world how advertising should work.
Traffic-Supercharger // June 6, 2008 at 6:32 am |
It seems a good discussion is going on.
Where I like the stuff on “New medium, same old advertising model” script, because it is quiet helpful for me.
Thank you!
Traffic-Supercharger
Ethan Bauley // June 6, 2008 at 3:05 pm |
Can we reexamine the economic rationale for orthodox “brands” before we get into designing new businesses to support them?
My $0.02:
In industrial media, creating or finding information was expensive. Firms had to manufacture “brands” with packaging, slogans, and ads to communicate expected benefits.
In a world of social media, reliable information is cheap and easy to find. Connected consumers can [and do!] choose to debate the expected benefits among themselves in incredibly rich detail.
To me, the question is: what are the implications for brands in an age of connected consumers?
Is the potential really limited to “dropping pixels” and reselling commodity inventory? Is it limited to increasing the efficiency of creating…the same old tripe (SpotRunner)? Is it about reducing the cognitive load of ugly interstitial display ads?
These are all noble goals and are directionally dead on, but they seem pretty marginal (even though I agree there’s money to be made). But still: does everyone in the Valley really want to be in the business of optimizing mass media advertising strategies?
I’m all for picking the low-hanging fruit, but…
I can’t help but think services like GetSatisfaction or twitter.zappos.com are hinting something much, much bigger.
Thanks for a thought-provoking piece…
- Ethan
Jeff Shultz // June 6, 2008 at 7:06 pm |
The problem with meaningfully and reliably measuring brand advertising is that the timelines for development of brand affinity and purchase intent can be vastly different. The decision to reach for a Coke rather than a Pepsi at the cooler is informed by the cumulative effect of perhaps decades of brand messaging, up to and including the logo on the can itself. You very well may be reaching for that Coke because of a television commercial you saw when you were five.
There is unquestionably an opportunity to better understand the effectiveness of advertising in developing brand affinity. But for many highly-branded product categories, brand affinity (as opposed to demonstrated purchase intent) may be the more meaningful measure of successful advertising.
More (Ad) Money, Mo (Ad) Problems // July 22, 2008 at 6:38 am |
[...] ad buying and the resulting lack of change in the market. That is until I stumbled across this. Leave it to a VC firm to spell things out in dollars and sense. WIthin, Michael Speiser of [...]
hypnosiscourse // March 21, 2009 at 2:36 am |
Data visualization may be becoming more ubiquitous (due to the growth in data available)– but where’s the application to storytelling? Remixing data to increase understanding or engagement may be good for business, and it may make your site the cool internet phenomena of the month, it may even lead to game-changing models for business, science, or civic engagement.
080604 Daily Links for June 04, 2008 Five Links for Recruiters | johnsumser.com: Recruiting News and Views // May 19, 2009 at 4:08 am |
[...] Reinventing online advertising: Exploiting the Achilles’ heel of the agency-media complex Good reminders about the fact that most online media involves generals who are fighting the last war. [...]