On last night's Yahoo conference call following its earnings report, CEO Terry Semel commented on postponing the rollout of a new advertising system -- words that sent the stock plummeting about 14 percent after hours (emphasis added):
I would like to discuss the improvements we are making to our search monetization capabilities and the introduction of our completely redesigned platform. As I described last quarter, there are three elements to the new system design: the core data platform, the advertising management applications and the new marketplace design. We have made some strong progress in a couple of key areas.
First, the underlying technology and core data platform was code complete in May, at the time of our Analyst Day. Since then, we have reached out to several hundred third-party agencies and tool providers, and have them accessing our development tools and sandbox to adapt their systems to our new platforms. This is a critical part of a successful launch, as their third-party tools must be upgraded in advance of advertising migration to the new interface.
Second, we recently did a detailed review of the advertising management application for the front end with about 175 of our most valuable search advertisers, and the feedback, I must add, was very positive. As a result, we believe we have the right functionality to help our advertisers drive demand and insight into their market.
As with any significant new endeavor, there has also been a lot of learning over the last two months. First, our comprehensive new platform approach, while complex and time-consuming, is coming together very, very well. We focused upfront on building a new system, and on various infrastructure investments that will enable us to iterate on consumer needs very quickly during the migration process by developing more automated tools and processes.
Second, we have also learned that given the complexity and importance of going to market with an application of this scale, it probably requires a little more time than we originally anticipated for two key steps. The first is the final integration and QA testing process, which is now underway. The second is the timeframe between the final QA signoff and the actual customer launch, given the legacy systems in place that will be operating in parallel.
Therefore, to meet the standards that we believe our clients should expect from us, we think it is prudent to add some extra time to our original estimates for the commercial launch. We now expect to roll out the front end, or advertising management application, in Q4 rather than Q3. We do have an aggressive plan to try to get the marketplace design piece out by year end, but because we want to collect feedback from our advertisers on the timing and cause them as little disruption as possible, for your planning purposes, we think you should plan for sometime in Q1. We believe this is the right decision, in terms of assuring the most successful commercial launch possible.
More and more marketers are coming to understand that the Internet can really help them meet their needs across the full marketing spectrum, from brand awareness, consideration, intent to purchase, trial, all the way through to transaction. To help accomplish this, Yahoo! is making the right investments in audience size, demographics and engagement, as well as important advancements in our advertising technology, to seek to deliver even better results.
And, later on in the Q&A, COO Dan Rosensweig responded to analysts' surprise:
Ben Schachter - UBS
...Also, on the Project Panama [the search advertising system that was delayed], I was wondering if you could talk about specifically what's changed since the Analyst Day, when it seems like you were pretty confident in the timing, to now. Thank you.
...On the question of Project Panama and the timing, Terry did mention today that we are going to move it a quarter away. We are two months further into the process. We're actually extremely pleased with the process. We're extremely pleased with the product, the stability, as you saw from the 175 advertisers who have a chance to see it and comment. We think we have picked the right feature sets and they are extremely pleased with it.
But as we got further along in the process, we wanted to make sure that we did it right. We don't manage the Company for a particular quarter, so we focused on making sure that we did all the necessary testing. We're going through testing now. Things seem to be looking good. But we do, for example, over 20,000 different tests to make sure that these things are right, stable, it's the right advertiser experience, the advertisers get what they expect.
We would rather take the extra time to make sure that we do it right, rather than try to rush into a quarter. This, of course, remains our top priority.
Mark Rowen – Prudential
Thanks, a couple of questions...
Second, if I could just follow up on the last question, are you finding a lot of bugs in the system that's taking longer to fix? It's still not clear to me why you need to push it out by a quarter. Thanks.
I will take the second question first and then go to Sue on the trends in advertising. No, we're not finding a lot of bugs in the system. It's just the reality is we put out our best timeline at the time. We're two months further into it. To get it over the goal line the way we want to, we just think it's going to require a little bit more time to make it the quality of the kinds of products that we normally release. So we would rather do it right and take a little extra time than hurrying it out to make a particular date.
We gave you the best date we could in May, and we are giving you the best date we can now. We're very excited about the prospects of it, and you have seen a lot of response from advertisers who are also very excited about the prospects of it.
So to be specific, we just learned more about the integration, the interdependencies and those things. So it's not about bugs or quality. It's literally just what it takes to get it over the finish line. We want to do it right.