ellipsis flag icon-blogicon-check icon-comments icon-email icon-error icon-facebook icon-follow-comment icon-googleicon-hamburger icon-imedia-blog icon-imediaicon-instagramicon-left-arrow icon-linked-in icon-linked icon-linkedin icon-multi-page-view icon-person icon-print icon-right-arrow icon-save icon-searchicon-share-arrow icon-single-page-view icon-tag icon-twitter icon-unfollow icon-upload icon-valid icon-video-play icon-views icon-website icon-youtubelogo-imedia-white logo-imedia logo-mediaWhite review-star thumbs_down thumbs_up

The ad exchange revolution: Crucial insights

The ad exchange revolution: Crucial insights Josh Dreller

We're arguably two to three years into the exchange buying revolution. I say arguably because buying online display ads via a real-time, auction-based model has been around in several forms for years. However, it was only as recently as 2009 that demand-side platform (DSP) became a household term in this industry.

So, how fast is this channel growing? Fast. In the eMarketer report, "Methods Used by US Ad Agencies to Buy Online Advertising, Q1 & Q2 2011," 10.6 percent of the agencies surveyed were buying ad inventory via a DSP in the first quarter of this year. By the second quarter, that number doubled to 19.4 percent. I've seen other reports showing that almost half of all online display budgets will be purchased this way by the end of next year.

If that's not worthy of "revolution" status, I don't know what is.

But fast success has bred deep growing pains. Audience buying has risen so quickly that buyers are scrambling to get their act together. The demand has outpaced the infrastructure, training, and human expertise needed to take full advantage of its potential. In terms of hiring talent, if you have one full year on a DSP, you're pretty valuable. Two years? You're an expert. Three years? You're a guru. Whoa.

Ask any exchange buyer who is actually in the trenches for an agency trading desk or in-house buying team, and they'll tell you that there's still a lot to learn with audience buying. Standard best practices that have been virtually set in stone by now with other channels such as search and email are still yet to be determined. What this means is that we're a little over a year away from this channel becoming a dominant force in the digital industry and only a few pockets of folks have really figured out how to approach this thing right.

Ultimately, there have been a lot of expectations that have built this tremendous demand, but the jury's still out whether or not we're actually there yet. I decided to tap into three prominent players in the exchange ecosystem in order to do a deeper dive into the promises of this revolution and how they feel things are netted out so far. This panel includes members from each of the three major parts to audience buying: the demand side, the supply side, and the data that powers both.

  • From the demand side: Zach Coelius, CEO of Triggit

  • From the supply side: Jay Sears, general manager of PulsePoint Exchange

  • From data side: Russell Glass, CEO at Bizo

Coelius, Sears, and Glass have not only lived through this revolution but have been key players in its growth. How you negotiate through the true value, the misinformation, and the noise will help determine your success in this area. It's more important than ever to set your practice on solid ground and sift the clutter. 

The following are the nine major value propositions that this channel has presented. Have the promises been fulfilled or is there still more work to do? What's the future of these expectations? Let's see what our experts had to say.

Promise 1: Buying audiences is better than buying sites.

Glass: Over time, buying audiences will become standard operating procedure for marketers, and there will be no such thing as "buying sites." There will certainly be money spent on the big brands, but all buys will ultimately have 100 percent targeted audience composition on the site that the advertiser is looking to advertise on.

Sears: First content was cool. Then audience was cool. And then content was not cool -- content was the kid getting beat up in the lunch room. Now content and audience are together again. Both are cool, and they are cooler together. We can now have audience, in context. The best of both worlds, with scale brought to you by real-time bidding.

Coelius: The fight between buying audience vs. buying sites is very simply a false choice. Sites matter. Audience matters.  Disregarding one or the other is huge mistake. One of the reasons why the exchange and DSP market has seen such tremendous growth over the last years is that it is finally possible to do both at scale. A DSP can now provide full transparency and control of where the ads will run while concurrently leveraging audience data at scale.

Promise 2: It doesn't matter where the inventory is, just who it's being served to.

Coelius: The "where" matters. The quality of the media placement has a tremendous impact on the efficacy of the ad. It doesn't matter how beautiful the targeting is if the ad runs below the fold on a scrapper site running an impression fraud scam.

Glass: Inventory control will become more important, and solutions are already beginning to expose attributes like "below the fold," "banner engagement," or "political content," etc. This will help the situation, and bid prices will be affected by these inputs. However, this will only further drive the notion that the inventory does matter, and ultimately pricing for the impression will take that into account. The better the inventory, the more the advertiser will bid for the impression.

Sears: Somehow some people think serving out-of-context ads in digital is OK. It's not. Unless you are betting your career on belly fat ads, your client puts value on environment. It's audience plus content. Over the next three years, look for an industry definition and taxonomy around quality -- what are the common measuring sticks of quality that will enhance the liquidity and trading of digital impressions. The taxonomy will by lead by IAB, and the delivery of the measuring sticks will be a battle between companies like Trust Metrics, verifiers going upstream like AdSafe, ad servers that should have done this years ago, and other new emerging entrants.

Promise 3: Data is the new currency.

Coelius: Data by itself is worthless. Leveraging data successfully requires experience, technology, and a willingness to experiment. Marketers who are able to get over that hurdle will crush their competitors who don't.

Sears: We always had data. Data per se is not really new. What is new is how dynamically data can be applied -- in real time, at the impression level. There are "three legs of the data stool" -- first-party data from the advertiser, third-party data from various data partners, and "supply provided data" from the exchanges -- what programmatic buyers see in the RTB bid call (see "A Smart Pipe Manifesto"). Marketers should have a strategy in each of these three data areas.

Promise 4: The ability to highly optimize exchange buys makes them superior to less optimizable direct-site buys.

Sears: If you are looking for something that is repeatable, at scale, then yes, exchange buys are the way to go. If you are looking for something that is intentionally not repeatable and not scalable -- but deeply customized to produce impact at lower scale -- then direct site buys may be your answer. And one can complement the other.

Glass: This is certainly true, and we've seen this again and again. (By the way, ad networks have the exact same advantage as exchange buys vs. direct-site buys.)

Coelius: Optimization is easy to say and hard to do. Smart marketers should be very skeptical of those who talk the talk with claims of magic algorithms, and instead they should test constantly to see which DSPs can actually walk the walk and truly perform.

Promise 5: Cutting out middle men and buying in an auction-based model leads to lower CPMs.

Glass: Not necessarily. I think the auction ultimately is a more efficient way to expose value. Some impressions will be worth more and some less, but once the auction becomes efficient through more bidders and the consistent use of data, we should see impressions clearing close to actual market value. Ultimately that benefits all players in the ecosystem.

Coelius: At the moment prices on the exchanges are unnaturally cheap, and marketers who have learned to leverage DSPs are able to see tremendous return on advertising spend. Yet the reason why prices are cheap is that the good inventory is mixed in with worthless junk. Without a smart buying platform, it is easy to lose your shirt buying supposed cheap inventory.

Sears: Stop worrying about the money other people are making and worry about your own ROI. Look at media spend as an investment and not a cost management game, and you will stop putting impediments in front of you and your client. This should be a race to the top, not a race to the bottom.

Promise 6:  Exchange buying give you a better chance to target niche audiences than site buys.

Sears: Yes, I'd agree. The emergence of real-time bidding (RTB) allows for the real-time assembly of audience and content at scales not previously achievable. Think of RTB as your own personal, virtual web property, customized just for you. Within the next three years, a client will be able to describe the virtual web property they want to purchase -- all aspects of audience, content, creative capabilities -- and will be able to purchase this on a spot or guaranteed basis. You want to buy guaranteed audience, content in an upfront market across the massive fragmentation that makes up the internet? Done. Not done like it is today with high operational cost, but done in a way that is technology driven but still requires highly skilled executives to design, execute, and modify the strategy for excellence.

Coelius: Scale matters. Without scale, niche audiences are pretty much worthless.

Glass: Largely false. I believe that highly endemic sites make it easier to find highly niche audiences than exchanges. There are some data providers (like Bizo) that can help to scale access to niche audiences within a specific vertical, but endemic sites will generally continue to have high value when it comes to finding certain highly niche audiences.

Promise 7: Targeting is king.

Coelius: Targeting does rock.

Sears: Unless you over-do it. Then the huddled masses will demand scale and overthrow the king.

Glass: Targeting will become more and more important, and ultimately those who have the most valuable and unique access to data will control media value and spend. Google, Facebook, LinkedIn, etc., are all trying to figure out how to have their tentacles (JavaScript, pixels, or similar) on as many pages as possible to see as many uniques as often as possible. This will ultimately drive huge value, and companies that can give them more and unique tentacles will be likely acquisition targets. 

This is becoming more and more true. In addition to the value of "buying audiences" discussed above, as more impressions are bought via auction, targeting becomes critical to know what to bid for the impression. Thus, ultimately targeting will become king, and those who possess better targeting and more data will have an advantage over those who don't.

Promise 8: Major workflow and cost efficiencies are gained by running all of your media through a trading desk.

Sears: There should be. However, too many buys still use trading desks and DSPs for the same line items they allocate to ad networks and single sites. The savings and efficiency is missed unless you manage overall spend in a more dynamic and cohesive way.

Glass: Coming from a company that operates a trading desk and a traditional ad network model, I can say that there are definitely clear benefits to run all media through a trading desk from a workflow perspective. I think that the cost efficiencies are still less of a certainty, and we see that the ad network model with site transparency and control often delivers significantly more ROI than the trading desk model. Additionally, given the lack of transparency, marketers don't always know if they are hitting all audiences. Over time, as more and more media becomes transparent on the exchanges and the overall quality increases, the trading desks should be able to compete more effectively with networks.

Coelius: Some day all digital media will be traded in an automated fashion, and smart marketers who understand this are moving quickly to develop a new skill set.

Promise 9: Auction-based exchanges offer an even playing field on which smaller advertisers and agencies can compete.

Glass: False. Not today and likely not ever. Those agencies that have the resources to mandate volume and scale have the advantage today because exchanges and bidding are such a new concept.  There simply aren't many (any??) small advertisers or agencies out there that know how to effectively leverage the exchanges, and even some of the trading desks still don't really know what they're doing. Just like on Wall Street, the smaller players tend to get left out in the cold, and the high-volume insiders generally have a significant advantage.

Coelius: At the moment, the playing field looks much fairer than it's even been. Small players can now get access to the world-class tools that previously only the big boys and girls could play with.

Sears: Yes, agree. Same great taste, less filling. Even playing field -- just invest in your people and bring your "A" game.

Josh Dreller is VP of media technology and analytics at Fuor Digital.

On Twitter? Follow Josh Dreller at @mediatechguy. Follow iMedia Connection at @iMediaTweet.

As a media technologist fluent in the use of leading industry systems, Josh Dreller stays abreast of cutting edge digital marketing and measurement tools to maximize the effect of digital media on client goals. He has achieved platform certification...

View full biography


to leave comments.