TC50: DataXu Optimizes Ad Bidding, Buying Across Exchange Platforms In Real-Time

DataXu is a Boston-based startup founded by a couple of tech entrepreneurs and MIT alums who are committed to making waves in the online advertising landscape by debuting the first real-time ad optimization system working across exchange platforms from the likes of Google, Microsoft and Yahoo.

Essentially, the aim is to bring more power to advertisers – rather than publishers – who are looking to increase the ROI of their online ad campaigns by making it easier for them to make fast decisions based on qualitative data and act on them virtually in real-time. Ad exchanges serve to float unsold and/or undervalued inventory in a pool to be bid upon by advertisers. On the marketplaces, sellers get guarantees about the impressions that will be sold for the highest bid price above the specified reserve, and buyers can use the exchange to supplement their online campaigns with low-cost impressions.

The DataXu platform values, bid manages and buys ads on an impression-by-impression basis, across the major ad exchanges and based on smart algorithms. The platform is said to be capable of processing hundreds of thousands of “ad decisions” a second, each returned in under 100 milliseconds, through automated, campaign-specific algorithms.

Founded in 2007, DataXu has raised $7.8 million to date from venture capital firms Atlas Venture and Flybridge Capital. The startup has been running its system in private beta on Yahoo inventory, is about to add Google’s upcoming Ad Exchange platform – which they say is in fact launching in the next few weeks and will be processing “hundreds of billions of dollars” a day – to the mix and has just added Havas to its roster.

Expert panel Q&A:

Q – Marissa Mayer: On a technology level, it looks impressive. My questions is: are you targetting people?

Ad buyers can build their own data profiles, so you can tweak it to fit your core audience. The Internet is becoming more dynamic, and what we’re doing fundamentally is make decisions quickly, change campaigns in real-time and learn from past behavior.

Q – Paul Graham: What’s the rocket science behind it, the core engine?

A: Our system is designed to find the features that matters for brand, really custom. Advertising is not a one-size-fits-all, you need dynamic, intelligent algorithms.

Q – Tony Hsieh: We’ve dealt with third-party pixels at Zappos, and it causes problems. How do you deal with that?

A: As soon as we can tie data together, we can work, so it doesn’t have to be pixel ads.

Q – Marc Andreessen: What’s your sales model?

A: we can paid on a CPM basis, like an ad server, but a percentage on the lift.



Extra coverage:
TC50: DataXu optimizes ad campaigns in real-time VentureBeat.
TechCrunch50: DataXu a Real Time Ad Optimization System Trends Updates.
DataXu Bringing Sophisticated Buying Strategies With Real-Time Bidding Says CEO Baker
DataXu Optimizes Ad Buying Across Multiple Platforms Pulse2.