In a new column on AdExchanger, I write a bit about why the promise of inventory centralization has been slower than expected. I think you should read all of it, but a few takeaways/bits of advice that I thought were worth pointing out (paraphrased):

Prove that you can make more money selling stuff in a market. Show us data that proves how much more money publishers can make in a market environment than throwing remnant inventory into a daisy chain. Real data please. I know it’s out there.

Don’t scare publishers by saying stupid things. Despite what the TV tells you, nobody wants to hear they saved a lot of money on car insurance. The same is true for online media. Be smarter about using the benefit of small pockets of test publisher data to make smarter direct buys from large publishers.

Do the simple things first. See if the inventory is worth buying before doing an RTB integration with it. Let me create campaigns without having to do some part (any part) of this setup manually.

Think API before press releases. Before hyping up vaporware, rather spend those resources building connections into your systems.

I think we can take a leaf out of the API books of companies like Facebook and Twitter. Create ecosystems around your data and advertising, and you will receive the rich rewards of remaining relevant in a complex (sometimes quite odd) marketplace.