One of my favorite com­pa­nies is Visa. It’s right up there with Gillette, Coca-Cola and Honda as an exam­ple of a com­pany sell­ing prod­ucts that peo­ple want and need, day in and day out. After all, who doesn’t need a credit card? What would you think, though, if I told you that Visa was launch­ing an opt-out fea­ture whereby Visa’s algo­rithms would look across its cus­tomers’ credit card accounts to find cus­tomers with unspent credit lim­its? Visa would then auto­mat­i­cally buy prod­ucts and ser­vices for those cus­tomers (based on their pre­vi­ous buy­ing habits) with the goal of spend­ing the remain­der of the credit?

Obvi­ously, Visa would never do this. After all, they’re a publicly-traded com­pany and are grow­ing just fine. Tak­ing such a move would be cor­po­rate sui­cide, as con­sumers would migrate in droves to Mas­ter­Card, Amer­i­can Express and Dis­cover Card.

As strange as it may seem, Google – whose cus­tomer base has no sim­i­lar alter­na­tive — is beta-testing a fea­ture in AdWords that does just that. “Auto­matic Match­ing”, as it’s called, is a fea­ture whereby Google’s AdWords sys­tem finds accounts where bud­gets aren’t being fully spent and then auto­mat­i­cally matches ads in those accounts to addi­tional “poten­tially rel­e­vant” queries in order to spend 100% of the budget.

If you’re get­ting ready to label me a Google-hater, hold off on unleash­ing the hounds. I think Google is a pro-consumer com­pany whose exis­tence has allowed more of the world’s peo­ple to do more things with infor­ma­tion than any other entity in human his­tory. Google has also been a boon to adver­tis­ers, intro­duc­ing a mar­ket­ing mer­i­toc­racy that rewards adver­tis­ers for ful­fill­ing con­sumer intent and pun­ishes com­pa­nies who run afoul of the searcher’s goal. Look­ing more broadly, Google’s appli­ca­tion of scal­able math and automa­tion to online mar­ket­ing has set the stage for a future in which all of online & offline adver­tis­ing – the last major inef­fi­cient mar­ket – will become effi­cient, thereby dri­ving the global econ­omy for­ward and improv­ing lives world­wide in the process. Free at last, free at last.

Homage to the ‘Plex hav­ing been made, “auto­matic match­ing”, which is cur­rently in its sec­ond beta phase, is frankly an SEM lazi­ness tax and smart mar­keters should, well, not be lazy.

Match types are dif­fer­ent types of ad-to-query match­ing that adver­tis­ers can use to con­trol how nar­rowly or broadly their ads are exposed to traf­fic. Typ­i­cally, adver­tis­ers make use of mul­ti­ple match types (“exact”, “phrase”, “broad” and “neg­a­tive match”). “Exact match” ensures an ad shows only when the exact key­word or key­word phrase is searched for; “phrase match” shows ads for the exact phrase as well as vari­a­tions of it; and “broad match” is the broad­est option as the name implies, match­ing your ad with the exact keyword(s) as well as the­mat­i­cally related queries at Google’s leisure. Lastly, “neg­a­tive match” is a way to specif­i­cally exclude key­words within “phrase” or “broad match”, a rel­e­vancy sculpt­ing tool of sorts.

The right match type strat­egy for the ROI-focused, metrics-driven adver­tiser is to use “broad match” as a research tool to find new key­words to add to your cam­paign that you might have over­looked in your pri­mar­ily “exact” and “phrase match” cam­paigns. Because wholly rely­ing on “broad match” will result in tons of non-converting clicks, “broad match” should be used as a farm team of sorts, whose bet­ter play­ers become “phrase” or “exact match” key­words and whose dogs become neg­a­tive key­words, thereby sculpt­ing “broad match” so that it works to the advertiser’s busi­ness goals and not nec­es­sar­ily Google’s.

Back to “auto­matic match”, now. Adver­tis­ers typ­i­cally set higher daily or monthly AdWords bud­gets than they want or expect to spend, sim­ply because set­ting a bud­get equal to what they actu­ally *expect* to spend would result in Google throt­tling back on deliv­ery of their ad as the adver­tiser approaches the bud­get limit. So what you have is a sce­nario where a strong double-digit per­cent­age of Google’s cus­tomers cur­rently have higher bud­gets set than what they expect to spend, and an auto­matic match­ing beta that wrong assumes the adver­tis­ers actu­ally *want* to spend all that budget.

Were Google able to give assur­ances that they would try to be ROI-relevant first, and budget-conscious sec­ond, “auto­matic match” might have value. I’ll even go so far as to say that in some cases it might actu­ally deliver higher vol­umes of transactions/leads/whatever within adver­tiser ROI con­straints. But AdWords is a yield opti­miza­tion sys­tem work­ing on Google’s behalf, and it appears as though spend­ing the advertiser’s bud­get will be the pri­mary goal, and seman­tic rel­e­vancy (and by no means ROI-relevancy) sec­ondary to that.

Keep in mind that even should you opt out of “auto­matic match”, you will likely still feel its effects. If the his­tory of AdWords new fea­ture roll­outs is any indi­ca­tion, 30–40% of adver­tis­ers go with what­ever new fea­tures Google sug­gests. Were 30–40% of Google’s adver­tis­ers to use “auto­matic match”, I’d expect the increase in cov­er­age and com­pe­ti­tion in Google’s ad space to result in key­word infla­tion as bad as the ris­ing gaso­line and food costs we are now experiencing.

To do bat­tle with the yield opti­miza­tion sys­tems we call AdWords, YSM and AdCen­ter, adver­tis­ers need inde­pen­dent SEM and ana­lyt­ics tech­nol­ogy that works to improve their yield and to meet their busi­ness goals. Auto­matic Match will do nei­ther, and so my sug­ges­tion to you – get seri­ous about SEM tech­nol­ogy and ana­lyt­ics, or face the SEM lazi­ness tax later this year.