62 Responses to “Publishers Are Killing Web Advertising's Potential With Misguided Pricing”

  1. JT Batson

    Jim, thanks for bringing the industry’s attention to many of the issues publishers are facing today. While I don’t agree with everything you say within this article, I do agree that we need to buck up as an industry and regain control of the value of our medium.

    Pricing is a serious problem for publishers. Much of this comes from a lack of technology to allow publishes to manage their inventory and business. Publishers have very little visibility in to their overall pricing, inventory value and cost of sales. Publishers have little to no control over 3rd parties selling their inventory and are generally lacking in terms of ad product segmentation. This mainly stems from the fact that publishers manage their business using “ad servers” built in 1996.

  2. Tom,
    I thought that DART's Ad Exchange would provide that "arbitrage". It seems to me having one (or several) central auction points for advertisers to seek out media and pay a reasonable price for it.

  3. Rick makes some excellent points when he says the agencies need to do their part. The problem is that agencies don't want to deal with 80 invoices per month buying ads at small sites. They want to make one buy and have done with it so they can go do lunch. Which makes me think there's an excellent arbitrage opportunity.

    Also, when it comes to the new business model for media, I like t call it a "Heinz 57" business model. You have to manage many revenues streams. Similarly with the agencies, publishers don't want to do that.

  4. David O.
    I agree – there is more value in other offline media. But that is why the author (and I) are obssessed with price. Where else can you purchase an ad for a $4 CPM? At this price, a .5% click rate is reasonably priced…even if you consider brand value equals zero (which I do not).

  5. Jim McCarthy – generally, I agree with you. Advertising is really kind of a weird business and probably shouldn't exist.

    When it first came into being, it actually WAS useful – providing information, letting people know of the existence of certain products, etc.

    Over time, however, it has become a very different animal. It's about entertainment many times, it's about ego other times, it's about who-knows-what most of the time.

    But it should exist at some level, otherwise there is no way to have the breadth and depth of entertainment and information that we currently have.

    Certainly, we can do with less of entertainment like the Kardashians and other ridiculous "reality"…that would probably HELP advertising. But there aren't alot of options available, besides advertising, which allow for the levels of choice and quality (or what passes for it) that currently exist.

    Banner advertising "doesn't work" is also saying Print Ads "don't work" and is also saying Radio Ads "don't work", etc. If advertising doesn't work – then it should disappear.
    It doesn't, because at some level it does work. The old saying half of my ad budget is wasted I just don't know which half has been assigned to several people. Regardless of who said it – it remains true.

    It's trying to figure out how to reduce this waste that keeps GOOD advertisers in business. To that end, many have opted for solutions which they deem "valuable" – such as the click or the response (for DR). Others have opted for "brand recognition".

    Regardless of the choice of measurement – it's clear advertising has a value and it's probably true that every method of measurement is useful to some degree. That is why I eschew CPC advertising – it devalues the brand building portion and customer retention. Why should I only get paid for one "sales lead" when each lead creates multiple purchase opportunities?

    In the end, I think all publishers (including myself) would love to see a revenue stream that could keep us in business but also get us out of the ad business. What is really strange is that it's almost impossible to get out of the ad business. Even subscription based businesses get into it – because the subscription creates a "qualified audience" – someone who is desirable and willing to interact.

    HBO may not serve ads in a fashion you and I recognize, but they certainly do have them. So does PBS. Fact is, a revenue stream is a revenue stream – and if you can have more than one, you will certainly try to get it.

    I find it interesting that you admit that banners have ad/brand value but they don't work. I'm not sure, aside from being facetious, what you're trying to say. Certainly, it's funny that advertising is a weird and sometimes dysfunctional business.

    But it is a business – and if it's run well (in some ways that the author of this article outlines), it can be very profitable for publishers and very useful for advertisers.

  6. David O.

    "banners, regardless of click rates, have value. A message is still being sent and received… the direct model discounts the other more valuable aspects of advertising: return sales
    brand awareness, information dissemination"

    True, but how much value ? Offline offerings offer a much greater value in the areas of brand awareness and information dissemination.

  7. Michael McNamara

    With all due respect to Mr. Spanfeller, it appears he is making a valid argument for pricing integrity, but his foundation is misplaced.

    Mr. Bcool was generous to synopsize "premium publishers should not offer remnant to networks and clicks are a poor measure of online ad effectiveness." This is crucial to understanding addressing the problem that Mr. Spanfeller is offering for discussion. Mr. Dietz (who I believe could partner successfully with Mr. Bcool) truly exposes the root of this problem with regard to the premium publishers; "use it’s [R/F] capability to qualify those with time visible, page position, share of voice, etc." Taken in reverse order, the premise that selling "remnant" becomes a non issue and premium publishers can return comfortably to the business of doing business.

    For the rest of the publishers that are bound by the 80%/16% consumer universe, Mr. Spanfeller's assertion that a big part of the problem is the notion of "remnant" ad units are killing web advertising's potential with misguided pricing is misplaced and perpetuates the lazy and thoughtless pricing models that are the the true danger to profitability from web advertising.

    Harsh? yes, but to summarize the earlier points; don't offer remnant ad units if you don't have them (Bcool) and use "right metrics" (Dietz), it is really that simple. While Mr. Spanfeller is genuine in his attempt to help us isolate the dangers that pricing models pose to profitability, he has unwittingly fallen prey to the true root of the problem which is not "remnant ad units" but rather publishers should take care to define and stick to pricing models that contains measurable value.

    By introducing and defining remnant ad units as he has, he is suggesting that publishers ignore the pursuit of substantive pricing models and instead waste energy battling semantics with ad networks looking for a deal. In doing so he is creating the invisible hand he speaks about with skepticism.

    1) Remnant ad units did not originate with the airline industry, “back of the book” of magazines or on late-night TV (the latter 2 are more appropriately termed as "direct response units). Remnant ad units came to the fore in the newspaper industry, shortly followed by print publishing; all well before the Write Brothers took flight. In fact, remnant ad units may very well be an example of a truly well crafted pricing model. Remnant ad units were sold by publishers who had a last minute ad cancellation and instead of enduring the cost of repaginating the publication they sought to offset a potential high and unacceptable operating cost by aggressively seeking out trusted and reliable advertisers to fill the space. This practice may be a concept that is inconceivable to those who have no experience outside the web industry, but believe me when I say remnant ad units helped publishers maintain their bottom line without leaving advertising dollars on the table (thank you Sir J Walter Thompson).

    2) While remnant ad units may exist in today's web industry (newsletters, emails, etc.), the reality is that there is no scarcity online, and the countless unsold impressions that are going to waste are just that — they are not being sold. It is the publishers who are package those impressions as "remnant" and therefore the publishers who are creating the invisible hands which Mr. Spanfeller accurately pointed out is fundamentally driving pricing down across the web and changing the metrics these large inventory buyers are using to negotiate. Remnant this, or remnant that, and like terms are not killing web advertising's potential with misguided pricing. It is the publishers who a assigning terms helter-skelter like "remnant" to sell inventory that could and should have been packaged more responsibly and sold accordingly.

    Media companies and the so called big ad networks understand these differences, they are well practiced in finding and exploiting the amateur pricing models that todays web industry thinks are so clever. Discounting is a slippery slope, but lazy and thoughtless pricing models are tantamount to throwing bacon grease on that hillside.

    If you tell the media buyer that you have a remnant, you are telling them if they buy it, they are helping to offset an otherwise lost opportunity cost. If in fact that sale does not offset an otherwise unavoidable expense, then you are simply promoting gimmickry which will certainly kill the web industry's potential to efficiently profit from web advertising.

  8. Jim McCarthy

    Rick, you're right. Lots of stuff has been tried, and yes, banners do have branding/advertising value.

    You're absolutely right about the Internet and advertising-driven businesses: not very much has worked, and you can now add "selling display advertising" to the list of things that don't work. I'd suggest strongly that entrepreneurs stay very far away from advertising-based business models for that reason.

    There's simply no law of nature that says the advertising business (in any form we'd recognize) must exist.

  9. I think most publishers HAVE moved into the "other things to do" realm. Saying they haven't ignores the variety of ways they have tried to increase revenues:
    content integration
    integrated text
    cross-platform integration

    Fundamentally, we are having this conversation because banners, regardless of click rates, have value. A message is still being sent and received. Click rates are as valuable as the phone calls DR clients get on TV – thus, while they have a place in the online (and TV) world, the direct model discounts the other more valuable aspects of advertising:
    return sales
    brand awareness
    information dissemination

    If all we did on TV or all we did online was price ourselves based on "response", then there is no model that would work to keep the content providers in business. Google's various forays into new revenue work due to size – not because they are fundamentally "BETTER" models to work with. Google is also very quick to exit revenue models which don't work – a luxury that is also afforded them due to size.

    Small sites cannot take risks on new revenue models. I challenge anyone posting here to show me a variety of small sites which have successfully implemented new revenue models after living on ad revenues. They are few and far between – usually existing in niches and nether regions of the internet.

    Implementing any new revenue models will require making deals with those you'd hope not to – such as Google. Why should I implement a new model, allow Google to take a portion, and hope that it will generate as much or more revenue than my old one? Either way it's a risk, and even if it works, Google's getting rich off me, something I'm not interested in generally.

    We're having this conversation because it's a good one to have. It discusses the primary goal of most people who are reading this article – reaching an audience with an advertising message in a price efficient method.
    It also discusses the fact that alternatives MAY exist – but are only suitable for certain players on the web.

    There is no cognitive dissonance that I see at work. I would strive daily to implement a new method if I knew it would generate the minimal levels of revenue that I currently make.

    Sadly, the web is overloaded with overpromise and underdeliver stories on revenue generation. In that respect, it remains the Wild West.

  10. ed dunn

    Jim McCarthy

    <i>"the question for the industry is not how to price it, but how to find something else to do."</i>

    Common sense indeed..however, this thread demonstrates the cognitive dissonance among "old school" offline ad execs with a "horseless carriage" menatlity towards marketing and monetization in the Web domain.

    Google, Inc. already demonstrated technically and financially contextual advertising science is scalable and defunct Alta Vista and others already demonstrated banner ads can't scale.

    Why is we having this conversation?

  11. Rajeev Goel

    My company, PubMatic, works with both ad networks and premium publishers, and I don't know anybody that disputes, on either side, that direct sold inventory is a better deal for the publisher.

    However, in many cases, campaigns purchased through ad networks have fundamentally different objectives and therefore should be purchased differently.

    I'll point to the article I wrote for Ad Age on Aug 14th in response to OPA's ad effectiveness study which also failed to mention the fundamental differences between the two types of inventory.


  12. Jim McCarthy

    The core issue here is that online display advertising is, for the most part, fundamentally ineffective.

    It may not be all about clicks, but clicks should be some indication of whether ads are having an impact, and with CTRs as low as .05% now, it's a pretty good indicator that no one is looking. (I don't know if Jim's "low education" data is true, but if so, it's not helping support the argument that this stuff works.)

    I would even go so far as to say that nearly all online display advertising is now remnant-grade, and the question for the industry is not how to price it, but how to find something else to do.

    Here's a further explication of that idea:


  13. George John

    I'm confused about the fervor with which Spanfeller and others are proselytizing the anti-ad-network religion.

    One of the reasons I started a display ad network was that I believed good content deserves to be well-monetized, and in our democracy an informed citizenry requires not just a free press but a profitable press. How can we get to a point where an important article about healthcare or foreign policy can be monetized as well as an article reviewing the latest digital camera?

    What publishers need is not to be Luddites and reject technologies that could help them extract more value from their audiences, but rather to embrace them and collaborate to build even better monetization solutions that the publishers desperately need. Online audiences are just too heterogeneous for a manual sales and campaign management process to capture and extract all that value — it requires technology.

    Spanfeller makes a remarkable claim that invalidates the hard work of anyone who's ever held a job as an online media planner: "On the web, ads generally perform the same regardless of when or where they are viewed."

    Hopefully the Forbes.com salesforce didn't believe this, otherwise it would make for a pretty comical sales pitch: "Please advertise on Forbes.com, your ad will generally perform the same here as anywhere else."

    I work for an ad network, Rocket Fuel Inc, where three of our customers (a restaurant chain, a clothing brand, and a resort hotel) have all told us in the last week that our ad network is driving the best performance of any buy in their campaigns. By one advertiser's own measurement we are driving over $10 in sales for every $1 they spend with us. Their other partners aren't delivering these results, and the only difference is that Rocket Fuel uses technology to select exactly "when and where" the ads should be shown for maximum effect.

  14. I agree with the premise of pricing to demand – in fact, while I worked at AOL years ago, I was shocked at their pricing models which were premised on the company's own "perceived value" as opposed to "real value". This, of course, led to Yahoo's price destruction during the busted bubble of 2001-2003 and the subsequent marginalization of AOL as a publishing product.

    Where AOL stands today is of little consequence in the grand scheme of things, but what is important is that their pricing model remains standard to many publishers.

    The basic premise of "pricing to demand" is essential regardless of whether you have limited (network TV), somewhat limited (Cable TV), or unlimited (Newspaper) inventory. I have worked in 2 of the 3 industries, and now I'm in online. I have never worked at a newspaper or magazine, but I do know alot about economics – supply and demand. How you price something will determine whether or not you create consistent, long term benefits for yourself and your clients. That, in return, will make you profitable in the long run.

    What makes the web interesting is that size can determine value. Part of the current pricing dislocation is the result of agencies putting a premium on "REACH" for online purchases. I would contend that is a valid purchase criteria for TV, but not for online. Trying to get "REACH" online is like fishing in large pond and thinking that you'll probably catch lots of fish if all you do is use the right bait. You may be right from time to time, but you probably won't be.

    As a result, publishers with large sites push their CPMs to the limit to "maximize" their revenue (I'd argue this is limiting their revenue by reducing potential growth of the advertising pie) – making smaller sites compete by keeping their prices lower. This means smaller sites have a harder time reaching revenue levels that are profitable.

    My belief is that if agencies were going about their business well, they would seek out smaller, highly targeted sites and purchase them first – even at a small CPM premium – then purchase the larger sites. Why? Because putting the onus on the larger sites to lower their CPMs is what will drive the agencies' ability to truly achieve the "REACH" they are seeking.

    In addition, it is common sense to realize that smaller, targeted sites are better for clients who are seeking a good result for their ad campaign. I recently ran a campaign for a client seeking a particular audience. I have a number of small, targeted sites meeting their needs. In order to prove my worth, I sold a $2.00 CPM and got a 1% click rate. While I didn't sell on a CPC or guarantee a click rate, any analysis can show that with a 1% click rate has a value of $4-10 on a CPM basis, depending on the client.

    Highly targeted, and very small, sites can provide the kind of response clients are looking for – which can then be "topped out" with low CPM rates from larger reach sites.

    As it stands, pricing models that currently exist are killing the smaller players in the web arena. Larger sites demand minimum budgets ($100k per month, no less) and will not lower their CPMs appreciably to allow advertisers to achieve reach.

    A quick story to finish – I once worked at a cable outlet which had pushed their CPMs to the upper limits. They couldn't figure out why their revenues were falling, because they had "valuable inventory" that was being sold to DR (Cable's version of remnant). I came aboard, lowered prices, reduced DR, and got more sold nationally. At the same time I reduced a 2 year backlog of Audience Deficiency Units which had built up. We saw a 16% increase in revenues that first year – while many on the sales team complained that it wouldn't work.

    Online needs to clean up its act….and pricing is the place to start…but the agencies have to start doing their jobs properly, too.

  15. Jim, great piece with points that really need to be said……publishers have to charge a premium/fair price to make this medium work; and great content will attract quality readers/buyers which deliver back to the advertiser the roi needed….free is really not free , ever!….Wilma

  16. It would seem to me that one of the primary concerns of online publishers is getting agencies and clients to shift more and more of their advertising budgets over to the online space. However, what you suggest is essentially going to these clients and stating "Hey, do you like the lack of clear ROI you're currently getting with TV, radio and print? How about shifting that to online? We too have no clear ROI or definitive metrics for success, and as an added bonus, you don't understand the medium at all!"

    Not especially persuasive. The sort of collusive price floor you're advocating may be feasible at some point, but it seems more attainable when advertisers have shifted more of their budget online, are satisfied that its performance is at least as good as offline, and online publishers aren't so focused on squeezing every last dime out of their page views just to keep the doors open.

  17. Jim Spanfeller

    Thanks all for the comments. Keep them coming.

    BCool, thanks for your wonderful summation. I wish it was as obvious as you suggest. We would all be much better off.

    As to this last comment by John Dietz, I have to say I mostly agree. Although routed in offline thinking, the notion of reach and frequency online is something that I think we will have to get to. Both as a determination of branding value AND as a way to parse internet advertising spends next to offline advertising spends.

    We do indeed have ways to measure branding success (and yes Ian, branding is advertising and thus has vastly differant characteristics then DR, like frequency for one) but they are not adaptable to offline measures.

    And finally the issue is not that publishers do not have data about their readers nor that they do not use that data to help sell their sites. Rather it is when a publisher unknowingly shares that data with an outside vendor and that vendor is completely non transparent to the end user.

  18. John Dietz

    There are certainly some salient points here, but I think what you've missed is understanding of the metrics unique to online advertising that can lead to a better ROI for branding campaigns (since you correctly pointed out the failing of click through rates).

    Once you throw out clicks, and devalue impressions (because an impression is really a measure of the ad server placing the ad, whether the user saw the ad for 30 seconds, 1 second, or 0 seconds because it was below the fold), you have to look at what's left.

    It may seem strange, but I think online advertising should start looking at things like reach and frequency, and use it's capability to qualify those with time visible, page position, share of voice, etc. Once we start looking at the right metrics, branding ROI can be calculated to some degree.

  19. Thanks for taking several hundred words to state the obvious: premium publishers should not offer remnant to networks and clicks are a poor measure of online ad effectiveness. Groundbreaking!

  20. Cord Silverstein

    Mr. Spanfeller, you raise a number of interesting points, but I have to disagree with your premise. Ad prices are dropping because they have not shown real value to the buyers. Online advertising has not been able to show true ROI.

    I also believe that your "study" on purchases through ad networks is not a real apple to apple comparison. Yes, by purchasing through an ad network, an advertiser can pay less, but I also believe the other factor is the customer service the advertiser receives from the network. Publishers want online advertisers to pay premium prices than they should also provide premium service to the buyers. We constantly run into communication and technical issues when buying directly from a publisher which we do not have with the ad networks.

    I think the bottom line is that to get where you want to go, publishers are going to have to make greater investments in technology and personnel to allow advertisers to better integrate and engage with their readers. When publishers begin to do this, they will prove their value and be able to charge what you believe they are worth.

  21. Citing an industry average of clicks is a useless stat…when creatives and offers begin to appeal to one's demographic audience, then they will click as well…

  22. Jason Kelly

    While I agree with a majority of points that Mr. Spanfeller is making about publisher pricing methodology, I completely disagree with his comparison to the airline industry…

    "The fact that we’re relying on methods developed by an industry (the airline business) that has to date not made any money in the aggregate is scary to say the least."

    Having spent over 12 years in the airline industry managing various yield mgmt. groups and now working within online media at a large publisher, I can attest to a number of things that we can in fact learn from an industry that has evolved over the past 20+ years to a place that is defined by:

    common data practices and technology platforms, demand segmentation and revenue maximization, efficient marketplaces with numerous distribution channels that promote the effective buying and selling of seats, the list goes on…

    This evolution required a significant amount of investment in technology, people and vision on behalf of all suppliers (aka airlines or in this case — publishers) in addition to having to adapt their business to address the demise of the travel agency and the rise of Online Travel Agencies which can be directly compared to the advent of Ad Networks in online media challenging the relationship balance between consumer, buyer and seller.

    Again, Mr. Spanfeller is indeed correct about the fact that the airline industry has not made money in aggregate, but I also don't believe that consumers would rather go back to having to pick up the phone to call a travel agent in order research and book travel…

    Evolution is underway (and fast) and requires all of us to look forward/backward and across industries as we continue to adapt for success.

  23. Ian Betteridge

    Tom says: "And rolling back to “demand creation” isn’t going to happen. Why would advertisers agree to a pricing model which would cost them more money with a worse ROI."

    Because it doesn't actually have a "worse" return on ROI?

    The problem is that online advertising measures one thing: instant sales. It's great at that, which is why it's so seductive. But that's also the most promiscuous form of sale, as it doesn't build any long-term relationship with the brand. For some brands, that doesn't matter – but for others, it matters a lot.

  24. Mr Spanfeller doesn't like the demographics revealed by the metrics therefore publishers shouldn't price according to what the metrics reveal. Unbelievable.

    And rolling back to "demand creation" isn't going to happen. Why would advertisers agree to a pricing model which would cost them more money with a worse ROI.

    He has written a good analysis of why ad pricing is in the dumps but he hasn't a ghost of a chance of bolting the barn door – the chickens have flown the coop (mashup metaphor #97).

  25. ed dunn

    <i>For example, we now know that 16% of web users generate 80% of clicks and that this 16% represents the lower income and education segments of the total user base. Do we really want to be held accountable as an industry by metrics generated by the lowest common denominator and a minority of users to boot? I can’t think of too many successful models using these types of metrics.</i>

    Shouldn't this be a clue that web advertising is being adopted mostly by the uneducated and gullible?

    In my opinion, the true misguided notion is still trying to make web advertising legit after 10+ years instead of grasping the fact web is a different animal and a new approch to marketing science must come into play. Zima banner ads are played out….

    Mommy bloggers writing narratives on a product/service is more potent than a banner ad. Discussions on a social network about an upcoming movie is more potent than a banner ad. The ability to scan content and return contextual referrals is more powerful than a banner ad. The ability to tie a product into a twitter #subject tag at the moment is more powerful than a banner ad.

    Contextual analysis, social trending patterns, business intelligence – again, we are in a new paradigm…

  26. Matt Mantey

    When publishers start actually selling audiences and their tendencies (publishers do have "vast amounts of information about their user base" that expires unused every second) instead of continuing to sell the adjacency angle for premiums, their remnant inventory won't be an issue.