A Brief Interview with Michael Yavonditte, Quigo

Way back in 2004, I spoke to Michael Yavonditte, the CEO of Quigo. In that post, I noted: According to Yavonditte, Quigo has perfected a relevancy algorithm that does AdSense one better – far better, to paraphrase his words. Quigo is focusing on picking off the high-brand-value publishers who…


Way back in 2004, I spoke to Michael Yavonditte, the CEO of Quigo. In that post, I noted:

According to Yavonditte, Quigo has perfected a relevancy algorithm that does AdSense one better – far better, to paraphrase his words. Quigo is focusing on picking off the high-brand-value publishers who use AdSense but are looking for a network solution that pays them more for what they believe is significantly better inventory than the lowest common denominator AdSense approach.

Well, it’s been a long three years, but Michael recently landed Time Inc. as a customer, in a high profile partnership that has shined the spotlight on the (sort of) startup. So I asked him for a Searchblog Brief Interview. The results are below.

So, first question. What do you think motivated Time Inc. to work with Quigo over Google?

A few reasons: Quigo will make Time Warner more money over the next 3-years. They can leverage their dominant media brands to command top
prices from advertisers and begin to “bundle” performance-based advertising with more traditional forms such as banner ads. 

Time’s senior management team spent a lot of time with Quigo’s team over the past 6-months. It’s a real partnership based on economic and product
development considerations. They want to sell their advertisers into the program — an important yet misunderstood part of our business model,
and not an area I wish to spend too much time on here. They want to control their pricing and relationships. Our system let’s them do both.
They want to collaborate to build new ad formats. And, they want world-class yield management. Yield is one of our core competencies.

We package all this up like no one else in the world today and deliver within a white label environment. The macro trends favor our approach here.

When you say “macro trends”, can you elaborate?

Sure. We’re seeing more and more publishers beginning to use our “you sell, we sell” model. This means that they are starting to value this aspect of our model much more than 2 years ago. Many publishers are beginning to embrace transparency. It means different things to advertisers and publishers. Publishers like to know what their pages are really worth. They’re more interested than before in auctioning off premium placements — not remnant placements. This was unheard of 2-3 years ago. This is precisely what Quigo does: we create hyper-transparent marketplaces for each placement on each site. This movement will only get bigger and such a trend benefits Quigo.

Aren’t publishers’ worried about losing their premium CPMs to an auction that pushes it toward remnant pricing?

I don’t think this will happen. On quality sites there will always be a pool of inventory that a publisher, with a sales force, can sell at high rates. Then there are the placements that sit between premium and remnant. They’re usually filled with graphical ads, and often purchased by hybrids. Hybrids are advertisers that care about direct response and building brands. They’re typically the brokerage houses, PC makers, banks, insurance companies, etc. With the right business model and auction approach, a company like Quigo can make such placements into terrific marketplaces.
We have many “text-only” placements that yield better than traditional CPM placements. It doesn’t always work this way but we’re getting better and better at yield and targeting. As such, the net value of each of our placements increases substantially over time. Publishers that still sell through bi-lateral negotiation can, in many cases, do better. This is art and science colliding.

I have heard from analysts that publishers are seeing higher eCPMs using Quigo than with Google. Care to tell me how and why?

We have very solid eCPM’s and feel like we can compete with anyone within the branded or premium segment.
Our results have gotten progressively better because of our relentless focus on yield optimization algorithms and some very craftly tweaks to our business model. We’ve pioneered the “art & science” of site-specific, page-specific auctions. How do you price each placement? Which placements are important? When should we use semantic targeting vs. psychographic targeting? What type of ad format works best? We’ve been doing this from the beginning and it’s really paying off for our customers.
I give the Time, Inc. folks a lot of credit. They actually rolled up their sleeves and dug into our model and tried to validate much of this through a significant amount of dialogue between us. They know much of what we know and they chose us because they truly believe we’ll make them more money.

Does this scale? I mean, if what you say is true, why isn’t everyone using Quigo?

Can we scale? People have been asking us this question from the beginning. We’d like for all premium and branded publishers to work with us, but it’s not quite that easy. We compete with some of the fiercest and most sophisticated competitors in all of business. Breaking through takes time, effort and luck. The Time, Inc. partnership puts AdSonar in the top tier of ad servers in the U.S. with more than 20 billions ads served per month. This is up from 6 billion/month in 2006.

We launched Time’s properties in less than 2 weeks and only hired one new employee in anticipation of the launch. We achieved 99.8% ad coverage in those two weeks. As has been reported, we’re estimating that this partnership will generate over $100M in three years. We’re obviously quite proud of what we’ve achieved to date but always strive to get better.

Could Quigo take on tons of new clients like Time Inc right now?

I wish there were tons of customers like Time Inc. in terms of brand, quality and scale. The answer to your question: Yes.

Thanks Michael, and good luck!

Leave a Reply

Your email address will not be published. Required fields are marked *