The internet offers unprecedented insight into consumer behavior, but not all the information is useful. The key to a successful campaign is having access to intent data.

Some claim there is too much data out there and the problem is marketers don't know how to use it. This is a very dangerous half-truth, like going out to the ocean without any fresh water; there may be a lot of water around you, but none of it is drinkable.

There is a lot of data, the most valuable kind on the internet being intent data. The internet is the only medium where shoppers leave reliable and accessible signs of their intent to purchase. Searchers spend 5 percent of their time searching, yet marketers chase these consumers with almost 50 percent of their budgets. Why? Because if you want to sell, why push on a rope? Why not politely respond only to those who are asking for your product or service? Then your ad becomes an answer and not an intrusion. Purchase intent powers search marketing and marketers love it because it works.

So are marketers swimming in intent data? Not at all, and here is the dilemma with targeting today: the creators and owners of intent data aren't the ones who own the traffic and inventory. The largest retailers, travel sites and auto sites have data that is more specific than what is gathered via search. An auto site knows a consumer's preferences in make, model and zip code. Travel sites know departure and destination cities, along with dates and preferences. This is more information than even Google knows. The problem is that the best data sources have very poor incentives to share that information with marketers. 

Let's go back to our analogy. Yes, marketers are surrounded by data (ocean water), but we're short of the truly indicative intent data (drinkable water). Most targeting players and networks are forced to scale by mixing what little intent data they have with non-intent data. They do this by adding "interest data," which assumes that if you read an article on cars, you must be in the market to purchase one. To further complicate the matter, many networks add fancy algorithms to deduce intent, which is necessary if you are guessing based on an article.

On the other hand, real intent data stands out on its own. If a person asks for a Honda dealership in New Jersey, it is clear what they want. The problem lies in mixing intent and interest data. Once you play that game, you gain volume but lose the quality audience that is genuinely interested in what you're selling.

Let's put this into context. Say you're running a marketing campaign for JetRed Airline's summer promotion. JetRed is a new airline and only offers routes to the top 10 most traveled cities. Now, the most traveled destination makes up only 4 percent of the total traffic on travel sites, with the top 10 destinations accounting for about 25 percent. JetRed's cream-of-the-crop audience would therefore be that quarter of the consumer population. However, most behavioral providers don't package their travel segments that way. Instead, they dilute them in two distinct ways.

The first dilution comes from the fact that most providers won't segment by destination city. Because of this, JetRed may end up buying a segment that includes the 75 percent of the population they really don't want to reach. The second dilution occurs when providers bundle in people who aren't in-market at all, but rather anyone who ever read a travel blog. These two successive dilutions may give you five to 10 times the segment size, but that comes with an almost proportionate loss in accuracy. Does that really create value for the marketer?

So are we really drowning in data? Yes, but all the wrong types. Be a smart marketer, ensure your campaigns are reaching audiences who have given intent signals and oh, bring a bottle of filtered water the next time youre out at sea.

Omar Tawakol is CEO of BlueKai.