Nervve’s vis­ual-search tools open up a new fron­tier for ad­ver­tis­ers

▶ Its soft­ware helps put a price on spon­sor­ships ▶ ▶ “We can scan an en­tire sea­son of Yan­kees games in un­der two min­utes”

Bloomberg Businessweek (North America) - - Contents - �Ira Boud­way

Ad­ver­tis­ing is com­ing to NBA team jer­seys next year. In April the league’s own­ers voted to al­low 2.5-inch-by2.5-inch spon­sor­ship patches on the front left shoul­der of game uni­forms. It’s novel real es­tate: No other ma­jor U.S. team sport per­mits such dis­plays, though soc­cer clubs have sold spon­sor­ships on shirts for decades. To fig­ure out how much ex­po­sure the new patches will get and how valu­able they are, some NBA teams are turn­ing to Nerve Tech­nolo­gies, a Buf­falo-based com­pany whose vis­ual search soft­ware is used by U.S. in­tel­li­gence agen­cies.

Nervve moved into sports last July, when it linked up with Wasser­man, a tal­ent agency and mar­ket­ing con­sul­tant that spe­cial­izes in sports. Ear­lier this year, ac­cord­ing to Amy Brooks, the NBA’S head of team mar­ket­ing and busi­ness op­er­a­tions, the league asked Wasser­man/nervve and about 10 other con­sul­tants to demon­strate how they could help teams eval­u­ate spon­sor patches patches. At the All All-star Star Game in Fe­bru­ary, both teams wore Kia lo­gos on their shoul­ders. It was, in part, a test run. Af­ter team own­ers ap­proved leaguewide patch sales, the NBA rec­om­mended two of the con­sul­tants, Wasser­man/nervve and Repu­com, a global sports-re­search com­pany. “Nervve re­ally stood out from the speed per­spec­tive,” says Brooks. A dozen teams, in­clud­ing the Cleve­land Cava­liers and Or­lando Magic, are work­ing with Wasser­man/nervve to an­a­lyze the mar­ket for the patches.

Nervve’s soft­ware can track how of­ten, where, how large, and for how long a spon­sor’s logo ap­pears to peo­ple watch­ing the game on TV or any other screen. The soft­ware can scan an hour’s worth of footage in five sec­onds or less. “On a sin­gle server, we can scan an en­tire sea­son of Yan­kees games in un­der two min­utes for a par­tic­u­lar brand,” says co-founder and Chief Ex­ec­u­tive Of­fi­cer Thomas Slowe.

Be­foree Nervve Nervve, says Zack Su­gar­man Su­gar­man, Wasser­man’s head of prop­er­ties, his an­a­lysts sat for hours watch­ing sport­ing events and log­ging when and where spon­sor lo­gos ap­peared on sta­dium signs, blimps, hats, and so on. “We’ve been do­ing mea­sure­ment and eval­u­a­tion for years,” says Su­gar­man. “Nervve just lets us do it more ac­cu­rately, faster, and at larger scale.” Wasser­man uses Nervve’s raw data to help leagues, teams, and ad­ver­tis­ers de­ter­mine the vis­i­bil­ity and worth of in-game ads—and where to place them.

Slowe has been study­ing ma­chine learn­ing since his un­der­grad­u­ate days at Rut­gers Uni­ver­sity. Af­ter re­ceiv­ing a master’s from MIT’S Me­dia Lab and serv­ing as a re­search sci­en­tist at Eric­s­son, he worked at a mo­tion-de­tec­tion startup in North­ern Vir­ginia. In 2003 he founded a com­pany that an­a­lyzed fa­cial ex­pres­sions.

With co-founder Ja­cob Goell­ner, Slowe started work­ing on the

tech­nol­ogy in 2011 for what be­came Nervve, aim­ing to serve U.S. in­tel­li­gence agen­cies. From his pre­vi­ous ven­tures, he knew the in­tel­li­gence com­mu­nity was buried in sur­veil­lance footage. “An an­a­lyst will get a ter­abyte drive of a hun­dred or a thou­sand hours of video on their desk each morn­ing,” he says. The au­to­mated sys­tems avail­able to scan for threats typ­i­cally re­quired rooms full of servers and op­er­a­tors with ad­vanced de­grees. Nervve cre­ated a drag-and­drop in­ter­face that al­lows a min­i­mally trained user to find ob­jects quickly and ac­cu­rately, with­out a lot of hard­ware.

The com­pany spent two years fig­ur­ing out how to iso­late pat­terns in pix­els with­out re­ly­ing on the long strings of yes or no ques­tions usu­ally found in ma­chine-vi­sion al­go­rithms. Slowe uses the metaphor of a jig­saw puz­zle: In­stead of pick­ing up pieces one at a time, Nervve scans a jumble all at once and makes guesses, leav­ing yes or no de­ci­sions to the very end.

Late in 2012, Nervve be­gan test­ing the sys­tem for what Slowe calls “a lot of three-let­ter agen­cies,” with­out spec­i­fy­ing them. In-q-tel (IQT), the in­vest­ment arm of U.S. in­tel­li­gence agen­cies, be­came one of Nervve’s back­ers in 2014; it hasn’t said how much it’s in­vested. Nervve li­censes its soft­ware to gov­ern­ment con­trac­tors, who run the searches them­selves. Slowe says the gov­ern­ment uses the tech­nol­ogy, for in­stance, to look at video col­lected by drones.

Once the gov­ern­ment busi­ness was up and run­ning, Slowe turned his at­ten­tion to the pri­vate sec­tor and be­gan so­lic­it­ing me­dia ex­ec­u­tives. Wasser­man saw the po­ten­tial value right away. The agency, says Su­gar­man, has al­ready used Nervve’s tech­nol­ogy to an­a­lyze brand ex­po­sure in foot­ball, base­ball, Nas­car, and Ken­tucky Derby tele­casts.

The first NBA patch sale came on May 16, when the Philadel­phia 76ers an­nounced that ticket seller Stubhub would add its logo to the team’s jer­seys. (Wasser­man doesn’t work with the 76ers.) Stubhub will pay about $5 mil­lion per year for the space, ac­cord­ing to an NBA of­fi­cial with knowl­edge of the deal. Su­gar­man ex­pects some teams to charge three times as much.

The bot­tom line Sports teams and leagues are us­ing Nervve’s im­age-recog­ni­tion soft­ware to as­sess the value of in-game ad­ver­tis­ing.

Newspapers in English

Newspapers from Canada

© PressReader. All rights reserved.