Back from the AdMeld Partner Forum

I had the pleasure of attending the AdMeld Partner Forum today at the Time Warner Center. It was an reunion with current and past employee galore. Alums have scattered to the wind and they represented probably a dozen companies.

Twitter was hot with some great nuggets from the panels.

Kicking it off with the keynote was none other than Emily Riley from Forrester Research. Emily is old school, one of the first delivery managers in the company. She helped pioneer the online education business back in the day with Casey Cook and Mike Peralta. Emily is super sharp and delivered the goods about RTB as well as shared some interesting marketer feedback about their expectations of display performance. If you weren’t there you can download a copy of the research here.

Out of all the panels some of my key take-aways are:

1. It’s small, but growing fast. When people mentioned the number of bids being handled in a day or a month we’re still not talking about large numbers. People realize scale is important and this was echoed across many panels. It’s not efficient to sell six impressions to anyone right now.

2. You don’t necessarily need RTB to get what you need. Much of the performance the demand side platforms (DSPs) are driving aren’t really related to naming your price in real-time. It’s the ability to cherry pick the right impressions based on a data set you choose and you know correlates with performance for you.

3. Someone on the agency panel said something to the effect, “CMO’s eye’s glaze over at the discussion of the technology involved with all of this.” This was in response to whether companies would build and/or run their own DSPs. Since many companies change their agencies every few years, doing it all themselves would require them to change out their marketing team. Not sure a wholesale liquidation of the team will be the most efficient. Operators will be needed to run the DSPs for most advertisers.

4. RTB won’t takeover all of the direct sale business anytime soon – it would be too unpredictable for the buyers and sellers. The advertisers need to spend their marketing budgets and show results. A large portion of this needs to be locked in to ensure the results. Leaving it up to the minute by minute moves of auction is not reassuring. You might not spend your budget and reach your audience.

The sellers of media have goals and executives that need to forecast revenue. Minute by minute forecasting doesn’t really work. This will encourage sellers to sell upfront deals to lock in demand and commissions.

5. Attribution is still up in the air and just because we can measure better online with clicks and purchases doesn’t mean we aren’t delivering branding benefits. It’s just easier to overshadow with such measurable results around the bottom of the sales funnel. If we can’t predict every result perfectly there is still need to buy and pray you choose wisely.

6. Publishers are getting smarter about how they sell their inventory. Designing the right channel conflict and understanding what each channel should offer as a feature set to advertisers should start to help bring some order.

Wrapping it up, special thanks to the AdMeld team for putting a great event together.

(originally posted on February 4th, 2010 on