If humans can't perfect real time advertising, maybe a machine would be better

By James Robinson , written on February 12, 2014

From The News Desk

Real time advertising is hard. The moment marketers are trying hard to capture is often gone before they’ve even put their arms out to try. At last year’s Super Bowl Oreo lucked into 15,000 retweets worth of free brand exposure with an ad made on the spot and sent out during the halftime blackout. For this Super Bowl, brand representatives had their phones ready like an old-timey duel but the most viral story out of the day were the spelling mistakes in JC Penney’s Twitter feed because it was cold and the person running the account had mittens on.

Taykey, an Israeli-founded advertising company that moved into the US three years ago and now has a handful of offices across the globe, concedes that real time advertising has its complexities but thinks that the machine can take out a lot of the work involved.

When founder and CEO Amit Avner began the company, Taykey offered brands a dashboard that tracked for them what their target audience were talking about in at any given moment. Avner says that the information was interesting to clients but impossible to take action on.

“The conversation changes 50, 60, 100 times a day. The data was relevant for just a few hours. And when you take a company like Unilver, with 200 brands and 200 audiences, there’s too much for it ever to be practical,” Avner says.

So instead, Taykey took the same 50,000 sources of data - the Facebook and Twitter firehose, news and publishing sites, message boards - and came at the problem from another angle. Its second iteration monitored what a brand’s audience was talking about in real time and bought ads next to the things that mattered to that debate. When the conversation changes, the ads move with it.

For an example, Avner tells me that late last year when Justin Bieber announced a tour of South America one of Taykey’s brands was trying to reach 16-year old girls, a major Bieber powerbase. Buying ad space before Justin Bieber video clips is expensive, but Taykey’s monitoring system saw that for a brief space of time 16-year old girls across the world were taking an intense interest in places like Columbia and Rio De Janeiro and placed ads in relevant spots for as long as the spike lasted.

“We want to really appeal to big, big brands to take some of their TV money and spend it online. TV is a billboard. You want to be next to music, you buy American Idol. Online is a fragmented audience, if you follow where the fragments go you can create a TV-like experience where you’re always relevant,” Avner says.

“We think that a certain percent of video, mobile and social budgets should be adaptive.”

From an advertiser standpoint, Avner says it’s an easy system to learn. Taykey offers different services for social, display and video ads. A brand chooses its target group and in which streams it wants to chase its customers. Taykey’s automated system places ads wherever that audience is engaged at any point in time and reports back to its clients about the 200 things their audience was talking about the most on any given day.

“We’re saying these are the ads we want to show and this is our audience and we’re placing bets on it,” Avner says.

“And as a bonus, there’s also no privacy concerns. We’re just looking at the information people publically broadcast. There’s no tracking involved.”

My immediate suspicion upon hearing this was that a brand could find out what its audience was talking about, but then finding the actual audience could be like looking for a needle in a haystack. Going after five million young Pepsi drinkers by targeting the hundreds of million pieces of content about Justin Bieber on the Internet is still a longshot.

Avner concedes that it’s not exact and that it is like trying to find the needle but in a smaller, more relevant haystack. Taykey’s automated system allows them the benefit of a constantly shifting target and he thinks the method works. The company has seen click through rates that are fifty percent better that the market average. Users watching a Taykey-placed ad on YouTube are 25 percent less likely to click on to the video early. The company was recently one of 20 vendors used by a major international client and a post-campaign survey showed it easily outperformed the other 19 in brand recall and intent to buy.

Taykey is working with Amex, Paramount, Coca-Cola, Dell, Sony and Unilever, Avner boast. He says the company has experienced 400 percent growth each year and is doing “eight digits” worth of business. Its approach to searching out an audience by interest and not tracking a user individually makes for a nice break from the industry norm of tracking and the aspect of automation and strong results it has achieved so far obviously allow it to capitalize quickly on its insights. But despite its growth and success, simply amalgamating public information and moving quickly will be very replicable by other agencies. If this space grows, Avner and his company will have to innovate quickly to stay ahead.

For Avner, his goal with the company is to show people that real time advertising doesn’t only have to take place on days like the Super Bowl. Marketers have previously gone crazy there because so many people are tuned into just one thing. But Avner thinks that when we’re all plugged in to the Internet, everyday is a Super Bowl-level event.

[Imagesource: Thinkstock]