06 May 2014 | Blog
Visible World recently introduced a new product, HighYieldTM, a TV sell-side platform that allows inventory holders to create and manage their own programmatic TV sales platform. HighYieldTM is designed to work alongside existing sales teams, an alternative way of selling ads rather than a replacement for the traditional way of doing things. So premium inventory will continue to be sold as units by salespeople, but some inventory will be converted to impressions and sold as audiences through an automated system. But how can an inventory holder decide when to put inventory into the system?
This may be a new product, but we’ve been thinking about this for a while. We built an optimization engine for ad sales, called the Dynamic Yield Management System, or DYMS, to help inventory holders optimize their pricing strategy. We could go into all the math—it’s pretty cool—but today we’re going to take a different approach.
We’re going to talk about cows.
Imagine you’re a farmer, and you have a herd of cows to sell.
There are also many different parts to a cow—ribs, brisket, tenderloin and so on—which can each be sold separately. But some parts are worth more than others, and many buyers only want specific parts.
Naturally, you want to sell your cow for the greatest amount, whether by selling the entire cow whole or dividing it into parts. Some butchers only want to buy the best parts of your cow. Some butchers will buy almost anything. Some are willing to buy the whole cow, even the parts they don’t need as much, to get a good price on the parts they want. But if you get stuck with parts you can’t sell, you’ll have to sell those to the dog food maker. And that pays very little.
With so many potential buyers, you want to sell the cow in a marketplace to get the best possible prices. There’s just one catch: if you don’t sell the cow in five days, the cow dies.
And no one wants to buy a dead cow. So anything unsold by then will never be sold.
There are a lot of factors involved in how a cow is priced. Farmers and butchers both have the ability to go to other markets if they don’t like the pricing they’re getting, but that takes precious time to do. Farmers don’t want to risk their cows dying, and the more time it takes a butcher to fill any one order, the fewer orders they can fill and less money they can make. The visibility differs—farmers have a better idea of what they’re charging for all their different cows. But everyone has a pretty good idea of what various cow parts have cost in the past.
So what does this all have to do with selling ad inventory?
The farmer is the MSO or network selling ad inventory. The cow is the ad slot. Some shows are inherently more valuable than others—big cows, whose parts are all more valuable than small cows. And some audience segments watching a specific show—like some cuts of meat—are generally more desirable than others.
But just like some recipes require skirt steak instead of filet mignon, some people have uses for non-prime sections. So all of the cow—or inventory—can be valuable when sold to the right person.
And for a good cow, or ad slot, some buyers are willing to pay a premium to lock in their claim. Tenderloin, or the most exciting demographics, may be worth more than the average price for the entire thing. But if the less desirable segments sell for too little (or worse, don’t sell at all), a cow or ad slot may be worth more as a whole than as the sum of its parts.
Butchers are media agencies. They need to fill their clients’ requests for ad time, with the required demographics, for the lowest price they can. Sometimes, to get the specific inventory they need quickly, they are willing to buy remnants they don’t need as well.
Everyone knows what their own prices have been in the past. And everyone is free to shop around for the best deal. (Though the ones selling inventory know more about the state of today’s current market than the ones buying.) But once a program airs, you can’t sell or buy that ad time anymore—it’s gone. So everyone also has a time limit. And the more time they spend shopping around, the more likely it is that they’ll fail to sell all the time or fill all of their orders.
Cows and ads need to be priced to make everyone happy. Sellers look to sell as much of their cow as they can, for the best price. Buyers also need a good price, but they need to fill their clients’ specific orders as well.
It’s possible to stitch together remnants to make a new “cow”. But if you just stitch random pieces together, you can get stuck with something that doesn’t make a lot of sense, and isn’t very flexible.
It takes a lot of skill and planning to aggregate impressions to form usable audiences.
So, given what we know, what’s the best way today to sell a cow?
Finding an optimal strategy and price that a farmer should seek to sell cows (or an inventory holder should seek to sell ad space) is not an easy task. One approach is through bottom-up, Agent-based Computational Economics (ACE), the “computational study of economic processes modeled as dynamic systems of interacting agents”.
Finding a solution enables us to help not only our hypothetical farmer, but also to solve real world problems. A particularly complicated problem to address is whether or not someone selling ad inventory should segment the audience for an ad slot, or only sell it as a whole. As new inventory continually becomes available, the complexity only increases.
Visible World’s Dynamic Yield Management System takes the guesswork out of deciding whether or not to segment inventory for the optimum price. HighYield layers in data so that aggregated inventory can be split up into audiences that are valuable to advertisers. Visible World’s optimization tools make carve management easy, protecting the value of premium inventory and adding new value to nonpremium. It’s the kind of automated decision making that makes programmatic ad sales not only possible but effective. Selling ads for the best price is complicated. But the workflow doesn’t have to be.