Sig­nif­i­cant shifts toward more automa­tion affect­ed near­ly every aspect of the paid search land­scape in 2018, and there were oth­er impor­tant devel­op­ments that will have a last­ing impact in the new year.

Major changes in SEM made waves through­out 2018 and are redefin­ing near­ly every aspect of paid search mar­ket­ing. There were a num­ber of momen­tous shifts, near­ly all with a com­mon thread of more automa­tion and machine learn­ing.

We iden­ti­fied nine big areas of change in 2018 that will shape the way paid search mar­keters work in the year ahead (and beyond).

1. Google Ads: New brand name, new UI

If you still call it AdWords, you’re not alone, but Google Ads is catch­ing on. The name change from AdWords to Google Ads is indica­tive of the fact that key­word selec­tion plays a less­er role in paid search mar­ket­ing than even a year ago, but more broad­ly the name change reflects the platform’s growth from one cre­at­ed for text ads to one that now includes dozens of ad for­mats across Search, YouTube, Gmail, Maps and a net­work of part­ner sites and apps.

For­mer head of ads and com­merce, Srid­har Ramaswamy, said dur­ing the announce­ment that the AdWords name con­not­ed key­words and search, “It’s basi­cal­ly a slight cog­ni­tive dis­so­nance to all the oth­er great things that we are doing in terms of both the for­mat and sur­faces these ads can show.”

Aaron Levy, direc­tor of paid search at Elite SEM, put it this way in a talk at SMX East: “Key­words are an old data lev­el. We have many more ways of tar­get­ing now. AdWords is now called Google Ads for a rea­son.”

With the new inter­face, we lost some things such as Dis­play Plan­ner and gained a YouTube reach plan­ner, notes and the abil­i­ty to make changes from the Overview page in addi­tion to oth­er new fea­turs. If you’re still not total­ly onboard, here are some help­ful tips.

Why is the UI change sig­nif­i­cant? “The UI itself isn’t the impact, but what the new design means,” said Levy via email this week. He sees the new UI as the change that will have the biggest influ­ence in 2019. “The new UI is more of a com­mand cen­ter than a dash­board — it looks strik­ing­ly sim­i­lar to the artist for­mer­ly known as Dou­bleClick. There are a mil­lion new levers and a mil­lion new ways to seg­ment data, which to me says we’re final­ly mov­ing away from key­words and text ads as the ‘core’ levers of SEM, espe­cial­ly with Google blow­ing up match types. There are a ton of sec­tions of the UI that cur­rent­ly only have one or a few fea­tures (see: advanced bid strate­gies), which def­i­nite­ly were not put there by acci­dent.”

2. Responsive Search Ads

We first report­ed on the big­ger, more auto­mat­ed ads in May when they were still in beta. Respon­sive search ads RSAs) are part of the con­tin­u­um of let­ting machine learn­ing mod­els do the work of ad cre­ative opti­miza­tion. Some of the ini­tia­tives that have come before it: dynam­ic search ads, auto­mat­ed ad sug­ges­tions (for­mer­ly Ads Added by Google) and Google’s efforts over the past year to get adver­tis­ers to give up man­u­al A/B test­ing and add at least three ads per ad group.

From our per­spec­tive, the intro­duc­tion of Respon­sive Search Ads (RSAs) was by far the biggest devel­op­ment of 2018,” said 360i VP, Search Mar­ket­ing Prac­tice Lead Jason Hart­ley. The impor­tance of RSAs, said Hart­ley is less about per­for­mance (though he said they’ve seen inter­est­ing results), and more about the bal­ance of automa­tion and con­trol that RSAs offer. “In 2019, we expect the equa­tion to shift fur­ther toward auto­mat­ed cre­ative and away from the man­u­al con­trol we’ve become accus­tomed to, with mod­u­lar ads becom­ing the default unit in the major search engines.”

An ad strength indi­ca­tor and some­what more exten­sive report­ing for RSAs were intro­duced in August. That same month, Google said it would soon roll out RSAs to more lan­guages and in the mean­time extend the extra char­ac­ter ben­e­fits of RSAs to text ads for every­one. This fall, Bing added sup­port for the third head­line and sec­ond descrip­tion in text ads, includ­ing the abil­i­ty to import the longer ads from Google.

3. Exact Match becomes “exactish match”

As with ads, more machine learn­ing was inject­ed into key­word-to-query match­ing in 2018 with the inclu­sion of same mean­ing words in close vari­ants of exact match key­words. The match type lost its lit­er­al mean­ing (the com­mu­ni­ty vot­ed to dub it “exac­tish match” in our renam­ing con­test ear­li­er this month) and forced mar­keters to rethink how they use match types alto­geth­er.

Opin­ions ran the gamut when we checked in with mar­keters in Novem­ber about the impact and how they were man­ag­ing the change, match types and cam­paign orga­ni­za­tion.

Amy Bish­op, founder of Cul­ti­va­tive Mar­ket­ing said, “With the tran­si­tion of exact match to more of an “exac­tish” match, the release of the audi­ence reports in Google Ana­lyt­ics, the increased empha­sis on YouTube tar­get­ing set­tings, it has become more and more clear that PPC pros need to focus on under­stand­ing the full cus­tomer jour­ney.” With that focus and infllux of automa­tion, Bish­op said, the mar­ket­ing dis­ci­plines of under­stand­ing your mar­ket and cus­tomers take prece­dent over man­u­al man­age­ment tasks.

We will like­ly be less and less able to con­tin­ue to rely so heav­i­ly on key­word tar­gets and, instead, need to con­tin­ue to focus on lever­ag­ing demo­graph­ic, fir­mo­graph­ic and engage­ment data to map out cam­paigns to our advan­tage,” said Bish­op.

4. AI-powered insights

This may not strike you as a big impact change, but both Google and Bing have ded­i­cat­ed sig­nif­i­cant resources to devel­op­ing much more robust (read AI-dri­ven) rec­om­men­da­tion engines in their inter­faces. Bing intro­duced a com­pe­ti­tion tab, per­for­mance insights and loca­tion rec­om­men­da­tions that high­light per­for­mance changes and com­pet­i­tive pres­sures, all deliv­ered with machine learn­ing. The amount of sug­ges­tions can get over­whelm­ing at times, but they are gen­er­al­ly light years ahead of the “Raise your bud­get to get more clicks” vari­ety of rec­om­men­da­tions of yore.

David Pann, gen­er­al man­ag­er of glob­al search busi­ness at Microsoft, explained at SMX East this fall that the whole sys­tem changestoo often for peo­ple to keep up. “So what we did is we built an AI infra­struc­ture that is real­ly pow­er­ing all the rec­om­men­da­tions that we are putting up through our user inter­face or through our sales team.”

Google also con­tin­ues to iter­ate on the data visu­al­iza­tions avail­able from the Overviews page. Ahead of the Black Fri­day week­end, it added a card in retail­er accounts show­ing how their cam­paigns were trend­ing com­pared to last year.

The goal is to spend less time down­load­ing and ana­lyz­ing spread­sheets and more time focused on strat­e­gy and cre­ative tac­tics.

5. AI-powered bidding

The man­u­al bid­ding option is now buried below a grow­ing list of machine learn­ing-dri­ven bid­ding strate­gies, includ­ing ECPC. On the Smart bid­ding front, Google intro­duced Tar­get Impres­sion Share, Pay for Con­ver­sions in Dis­play cam­paigns when Tar­get CPA is the bid­ding strat­e­gy, and rolled out Smart Bid­ding for search part­ners. Bing intro­duced Tar­get CPA and Max­i­mize Con­ver­sions bid­ding strate­gies.

Bish­op said of the Pay for Con­ver­sions option, “It’s almost as if Google is giv­ing us some train­ing wheels to test our wings with some of the tar­get­ing options that adver­tis­ers have been leery of. Mar­keters now have a safe­ty net for test­ing new tar­get­ing options with­out the risk of spend­ing a sig­nif­i­cant amount and get­ting no leads in return.”

6. Audiences: LinkedIn data for Bing, MSAN and more

In May, Bing launched the Microsoft Audi­ence Net­work (MSAN), which encom­pass­es native ad inven­to­ry on, Microsoft Out­look and the Microsoft Edge brows­er, as well as syn­di­ca­tion part­ner for what are now called Microsoft Audi­ence Ads. It uses AI for ad deliv­ery opti­miza­tion and uses data from the Microsoft Graph for audi­ence tar­get­ing, includ­ing web and search activ­i­ty, demo­graph­ic and con­sumer behav­ior activ­i­ty, and select LinkedIn pro­file dimen­sions. In Octo­ber, Bing also made LinkedIn cat­e­gories of com­pa­ny, job func­tion and indus­try avail­able for tar­get­ing in search and shop­ping cam­paigns in the U.S. Look for more attrib­ut­es to pos­si­bly become avail­able next year.

If you adver­tise in the EU, you’ve like­ly been deal­ing with the reper­cus­sions of GDPR and rethink­ing how you work with audi­ences. Maria Cor­co­ran, world­wide search mar­ket­ing media man­ag­er at Adobe Sys­tems, said GDPR had the biggest impact for her EMEA team and caused them to com­plete­ly piv­ot their han­dling of both email and audi­ences. Apple’s Intel­li­gent Track­ing Pre­ven­tion (ITP) and ver­sion 2 of Google’s glob­al site tag caused their audi­ences sizes to shrink by 5 to 10% on aver­age, said Cor­co­ran. And she sees that con­tin­u­ing into 2019.

The changes in audi­ences with MSAN in Bing, in-mar­ket for Google, expan­sions and improve­ments in sim­i­lar audi­ences are all immense­ly impact­ful,” said Cor­co­ran. “So the ques­tion becomes: how do we get more effec­tive uti­liz­ing audi­ence options such as in-mar­ket, detailed demo­graph­ics, YouTube view­ers as an audi­ence for search, LinkedIn and MSAN.”

7. Universal automation vs. hybrid management

Uni­ver­sal automa­tion is the term I’m using to refer to cam­paigns that are near­ly entire­ly auto­mat­ed — tak­ing a cue from Uni­ver­sal App cam­paigns, which was the first cam­paign type to have automa­tion woven through­out from tar­get­ing to bid­ding. Uni­ver­sal automa­tion was a major theme this year.

Google intro­duced goal-opti­mized Shop­ping cam­paigns, Smart Cam­paigns for small busi­ness­es and Local cam­paigns for dri­ving in-store traf­fic.

With Smart Cam­paigns, every­thing from ad cre­ation, audi­ence tar­get­ing, ad deliv­ery across Google chan­nels — and soon land­ing page cre­ation — are auto­mat­ed based on the advertiser’s goal.

Goal-opti­mized Shop­ping cam­paigns employ machine learn­ing to auto­mat­i­cal­ly opti­mize ad deliv­ery to achieve the defined con­ver­sion goal val­ue, such as rev­enue or return on ad spend (ROAS). It also com­bines dynam­ic remar­ket­ing and stan­dard Shop­ping in one cam­paign to deliv­er an ad across Google prop­er­ties and the Google Dis­play Net­work.

For Local cam­paigns, adver­tis­ers set a bud­get, and the ads are gen­er­at­ed auto­mat­i­cal­ly based on ad cre­ative ele­ments from the adver­tis­er and their loca­tion exten­sions. Google auto­mat­i­cal­ly opti­mizes ad deliv­ery across Search, YouTube, Maps and web­sites and apps in its ad net­works.

These uni­ver­sal automa­tion cam­paign types are designed pri­mar­i­ly for small­er, resource-strapped busi­ness­es. For more sophis­ti­cat­ed adver­tis­ers, we can see the engines trend­ing toward an automa­tion-heavy approach that still pro­vides some lev­els of con­trol for mar­keters.

Hart­ley refers to this as a hybrid approach. Based on the tra­jec­to­ry of automa­tion in ads and exact match, Hart­ley spec­u­lates, “This hybrid approach has the poten­tial to spell the end of man­u­al con­trol of match types (except neg­a­tives), bid mod­i­fiers, geo­tar­get­ing and oth­er tools we use for cam­paign man­age­ment. This doesn’t mean peo­ple will be erased from the equations—and cer­tain­ly not in 2019—but it will require SEM spe­cial­ists to adopt a very dif­fer­ent mind­set. We have to embrace machine learn­ing as a tool that helps us find oppor­tu­ni­ties that are imper­cep­ti­ble through tra­di­tion­al tech­niques. That’s been option­al in the past. In 2019? It’ll be manda­to­ry.”

8. New inventory locations and surfaces

There were a num­ber of new sur­faces for ad for­mats that opened up in 2018. Bing’s native ads extend­ing across the Microsoft Audi­ence Net­work, as men­tioned above, is just one exam­ple.

In Novem­ber, Google made AMP Sto­ry ads avail­able to all pub­lish­ers, and there are now more than 100 ad tech ven­dors with AMP inte­gra­tions.

Hotel deals from Google Search are now fea­tured on the Ben­e­fits tab of the new Google One cloud stor­age app (for Android only at this point. Expect to see more pro­mo­tion­al oppor­tu­ni­ties. Google said more pro­mo­tions from the Google Store and Google Express ben­e­fits and more will even­tu­al­ly be fea­tured on the tab.

Google has also been test­ing native ads in the Dis­cov­er Feed (orig­i­nal­ly Google Now) on the front page of the Google app. It’s a lim­it­ed whitelist test for now, but could open up more native inven­to­ry in a promi­nent loca­tion.

Con­nect­ed TV adver­tis­ing is grow­ing rapid­ly, and YouTube says more peo­ple are stream­ing it on their tele­vi­sions. This fall, “TV screens” became the lat­est device type for video and dis­play cam­paigns in Google Ads. Eli­gi­ble dis­play and video cam­paigns are run­ning on TV screens by default now, and can be man­aged with bid mod­i­fiers. Expect to con­tin­ue hear­ing much more about con­nect­ed TV adver­tis­ing in 2019.

9. Amazon advertising comes into its own; Google’s counter-strategy comes into focus

The Google-Ama­zon face off start­ed with where con­sumers start their prod­uct search­es. Now it’s extend­ing to search and oth­er adver­tis­ing. Ama­zon was declared the third-largest dig­i­tal ad sell­er in the US by eMar­keter in Sep­tem­ber, (far) behind Google and Face­book. Its ad rev­enue is expect­ed to increase by 50 per­cent per year through 2020, which would put its mar­ket share at 7.0 per­cent, up from 4 per­cent cur­rent­ly. From a mar­ket share per­spec­tive, that may not sound impres­sive, but con­sid­er some ana­lysts expect Amazon’s ad busi­ness to sur­pass that of AWS by 2021.

In our Ama­zon Adver­tis­ing Fore­cast 2019, 80 per­cent of Ama­zon adver­tis­ers said they plan to increase spend in 2019, with 30 per­cent say­ing they’ll shift some bud­get from search. Agen­cies are tak­ing note of the growth and build­ing up their Ama­zon adver­tis­ing prac­tices. Elite SEM acquired CPC Strat­e­gy pri­mar­i­ly for its Ama­zon prac­tice and bid­ding tech­nol­o­gy. Merkle built its own bid­ding plat­form for Ama­zon spon­sored brand ads. Just two exam­ples.

I think the rise of Ama­zon as a major play­er in the paid search space is the biggest devel­op­ment of 2018 that will con­tin­ue to impact adver­tis­ers through­out the next year, both in terms of bud­get allo­ca­tions and cam­paign man­age­ment,” said Andy Tay­lor, asso­ciate direc­tor of research at Merkle. “With ad for­mats that oper­ate dif­fer­ent­ly than those offered through Google or Bing Ads, adver­tis­ers are devel­op­ing dis­tinct best prac­tices for the plat­form.” Tay­lor said that though Amazon’s man­age­ment and report­ing capa­bil­i­ties are still well behind Google’s, he expects to see progress in the com­ing year.

I think the argu­ment that Ama­zon is steal­ing bud­get direct­ly from Google has been a bit overblown, said Tay­lor, “but there’s no deny­ing that Ama­zon has become a paid search force very quick­ly and is here to stay.”

Google’s strat­e­gy to counter Amazon’s encroach­ment into prod­uct search is to part­ner with retail­ers. It debuted Shop­ping Actions in March to address three key chal­lenges on the e‑commerce front: (1) how to make mobile shop­ping from its prop­er­ties like Search faster; (2) how to main­tain mar­ket share for prod­uct search in a splin­ter­ing mobile land­scape of apps and dig­i­tal assis­tants; and (3) how to com­pete against Ama­zon.

Shop­ping Actions runs across Search, any Google Assis­tant-enabled devices, Google Express (which includes Shop­ping ads that now fea­ture blue shop­ping tags instead of para­chutes), and fea­tures a uni­ver­sal shop­ping cart and a Google-host­ed check­out when users save their pay­ment infor­ma­tion in their Google accounts. It launched with 40 retail­ers, includ­ing ear­ly part­ners Wal­mart and Tar­get. The pro­gram now includes more than 300 retail­ers, big and small.

On the omnichan­nel front, at SMX Advanced, Google announced the launch of local cat­a­log ads for Dis­play, extend­ed affil­i­ate exten­sions to YouTube and more. At SMX East, Google shared more options for mes­sage exten­sions and store vis­its.

Looking ahead

In review­ing last year’s recap of the year in paid search, the over­ar­ch­ing trend was that “arti­fi­cial intel­li­gence and machine learn­ing under­pinned near­ly every paid search update.” Sound famil­iar? The dif­fer­ence this year is that the shifts were even more pro­nounced and per­va­sive.

2018 is kind of a turn­ing point in terms of SEM I think,” said Levy. “Key­words aren’t gone, but they’ve been great­ly depre­ci­at­ed. Methinks SEM in 2019 is going to be much clos­er to mod­ern pro­gram­mat­ic than it is to the ‘enhanced’ era.”