Posted by & filed under PPCChat.

Ginny Marvin was invited by host Julie F Bacchini to join this week’s PPCChat session where Ginny helped PPCers to find answers to most of the announcements & updates made in the Google Marketing Livestream event which was held on May 27th, 2021.

Here is the screencap

Q1: If you don’t use automation, is Google Ads still using machine learning to try to show your ads to the best potential customers? What advantages do you get from using the automated options?

For bidding, only the 5 Smart Bidding strategies use machine learning to optimize for conversions or conversion value in every auction and factor in multiple contextual signals. @adsliaison

In general, automation allows you to take advantage of real-time optimizations based on numerous signals — including your own data, inputs & goals — that just isn’t possible to do manually at scale. @adsliaison

Obviously a little more transparency on some of these signals would help. I’m sure ML will help us – but I think we can help the ML along too. @Galliguez

I think additional reporting on “responsive” ad units is a great start. Knowing CTR and conversion data on ad copy combinations could help businesses in other areas like email, page copy, etc. and I don’t think it degrades privacy to report it. @robert_brady

I am big on automation, but I am bigger on human oversight and intervention. @JonKagan

I’d say Google is likely using the auction over machine learning. Highest manual CPCs + whatever is left of quality score metrics. I’m not certain how much importance is still put on QS in the manual world. @sonika_chandra

The advantages begin with time back on bidding and ad copy variation creation and continue through increased data to make strategic decisions. @sonika_chandra

PPCers Responses & Queries Regarding Automation

Can we expect more updates in the future which would help budget control while using GA automation? @AhmerZeshan

100%! – Ginny is their anything in the pipeline to stop overspend per campaign? @andreacruz92

Would really love insights on how much budget shifts can help or hurt automation. We all have different schools of thought on “jumpstarting” the learning period. Would love your take! @navahf

Thanks for this, Ginny. Can I ask how the pandemic has affected these signals (if at all)? e.g – is GAds learning “faster” than pre-pandemic? @AzeemDigital

Q2: Google keeps talking about “partnering” with advertisers and then talks about pushing automation on us. How is that partnering?

This is really important & I want to spend a bit of time on it. I get that the automation acceleration can feel like losing control & precision, but the aim is always to deliver more advertiser value, at scale. @adsliaison

Computing power has increased exponentially & ML systems can now use 10s, 100s of thousands of signals at levels that aren’t possible manually. @adsliaison

Obviously a little more transparency on some of these signals would help. I’m sure ML will help us – but I think we can help the ML along too. @Galliguez

YES! I really, really want a bad match reporting mechanism so I can at least help teach the machines to be better faster for my accounts! I have been suggesting for this for YEARS. @NeptuneMoon

This would be super helpful. @adsliaison is there any chance of this coming? I feel like this would help both Google and the advertiser. @amaliaefowler

I think additional reporting on “responsive” ad units is a great start. Knowing CTR and conversion data on ad copy combinations could help businesses in other areas like email, page copy, etc. and I don’t think it degrades privacy to report it. @robert_brady

W/ all the complexity in customer journeys today, automation is really the only way to be able to anticipate behaviors & find converting customers at scale. It’s a mindset shift about what to focus on & what’s going to move the needle. @adsliaison

We now need to understand how to work w/ the systems & how to use new controls & levers to speed up learning & guide the ML w/ your business data & inputs on what you value, what’s working, etc. @adsliaison

Automation is also going to be critical as the privacy landscape evolves and we use modeling to fill in gaps where consented first-party data isn’t available. @adsliaison

All that said, this is a partnership & we are listening to feedback and investing in several areas the teams know are important: twitter.com/adsliaison/sta… @adsliaison

I always feel partnering means the brand, and to hell with whoever is there along the way. @JonKagan

IMO, I think most advertisers aren’t afraid of automation acceleration as “losing control & precision” but rather the potential removal of bidding options as the real loss. Especially when we have less data than ever before. @gregfinn

Exactly. I am not afraid of automation. I’m not afraid of testing it out and implementing it. I feel like the lack of communication and control around some of these things makes it difficult to feel confident in Google’s transparency and communication. @amaliaefowler

PPCers Asks Regarding Machine Learning

Would be great if you could speak to how much humans “ruin” or “mess up” the “learning part of machine learning. We’re often told we can’t touch a thing…but if the thing is spending 2x-3x what we gave it for daily budget, it’s a bit tough to square away with clients. @navahf

I think machine learning has flipped the “start running ads and get information back quickly” mantra that we have all lived with/by for MANY years… Expectations have to be different with automation and ML involved. @NeptuneMoon

does the shift mean some of the modeling models are being opened up for smaller accounts? For a lot in automotive, their first dance with Google modeling is how they never reach the threshold for in store visits, which doesn’t give confidence the new stuff will apply. @cdawg2610

My question here is simple: why aren’t you including feedback or signals from advertisers, who actually knows the business? @DigitalSamIAm

Q3: For advertisers with smaller budgets and/or smaller volumes of conversion activity, automation still struggles. What is Google doing to address this and what suggestions are there for making automation work for these types of accounts?

Great question. Smart bidding still might not always be the best option for smaller budget or lower volume accounts. To evaluate results accurately, you may need to look at longer time periods w/ at least 30 conversions, such as a month or longer (50 conv for tROAS). @adsliaison

And yes, as alternatives, Smart and/or Local campaigns might also be worth another look. @adsliaison

Thank you for confirming this. Smart bidding STILL needs at least 30 conversions in a month to run! Don’t get peer pressured to turn it on until you meet the data threshold! @navahf

This is why it often does not work for B2B unless we set up low-value, high-volume microconversions. @beyondthepaid

Experts Need Little More Information

So, does this mean that smaller volume or smaller budget accounts will not be able to have machine learning working on their accounts (per answers to Q1) if they don’t go automated bidding? The volume levels needed are not easy for all to achieve? @NeptuneMoon

Or manual? @amaliaefowler

Q4: How exactly do we go all in on first party data? Is it just uploading customer match lists? Will you make importing offline conversions and linking CRMs to G Ads much, much, much easier if we need to do this as well? @markbclicky

For many biz, developing a 1st-party data strategy is a BIG shift & will require intentional, strategic investment to create new ways to establish customer relationships OR making use of data you have but don’t yet act on. @adsliaison

Customer Match let’s you upload online & offline customer data — and it’s now available to nearly all businesses — announced at GML last week. @adsliaison

In addition to the Salesforce integration, you can pull in CRM data with Zapier. More on that here: support.google.com/google-ads/ans… @adsliaison

And you may find this article helpful as well thinkwithgoogle.com/marketing-stra… @adsliaison

Q5: There weren’t any updates about keywords, match types, ETAs going away as some of us have heard from reps. Should we take that as no major changes are happening in those areas this year?

Well, w/ RSAs now the default, for example, that’s a clear sign to focus on them. I’ll resurface this reference, too: twitter.com/adsliaison/sta… @adsliaison

RSAs, like anything new, do take some experimenting & understanding of how they do & don’t work. The two descriptions should be unique & work together, for example. @adsliaison

And the other thing I’d point out is that broad match has been overhauled to deliver more relevant searches by incorporating signals like landing pages, keywords in ad groups, and more. @adsliaison

Questions & Opinions From PPCers

Is this true for broad match ONLY if you’re using a smart bidding strategy? @NeptuneMoon

Is there any mechanism coming for us to report when these matches are not accurate / miss the intent completely? @amaliaefowler

This. This is a HUGE issue. When you hide STRs AND push automation AND don’t provide a feedback loop, the only group you’re helping is your shareholders. @DigitalSamIAm

I mean good, but this does feel similar to the exact overhaul, which basically just tasked us with a race to negative everything it would pick up. are match types just going away once these different overhauls end up sending the same signals? @JuliaVyse

You will take your marginally above breakeven performance and you will like it! @CJSlattery

Shouldn’t that have always been a part of making sure that keywords are most relevant to ads & LP’s (regardless of match type)? Not sure why that would be a new thing? @jdb426

When will Google start giving reporting on combinations beyond impression data? @robert_brady

With RSA being the default, will there be increased reporting on the RSA combinations and performance? (We all really, really want that). @NeptuneMoon

We ever going to get to pin things in 2 places or are we going to continue to get absolute dumpster combos? @CJSlattery

RSAs remain tough for compliance depts to understand. Any tips on that @adsliaison? @beyondthepaid

Was literally just about to ask this question. Might add, how best to counter the “OK, trial RSA’s, but pin this, this and this” objection. @AzeemDigital

Q6: Speaking of keywords and matching… Search query data loss is incredibly frustrating particularly coupled with the push to broad match. Is it ever coming back? Will we at least be able to see converting terms again at some point?

First, this was solely a privacy-driven decision. But the teams know the loss of query data can make it tough to know what’s working, especially for lower volume accounts and been working on solutions…@adsliaison

To give advertisers new query insights while protecting privacy. It is a priority & there will be more to come on this soon. @adsliaison

The why is to ensure user privacy from the search/impression to conversion. @adsliaison

Thoughts From The Experts

I’m curious how it is a privacy decision when the STR is still showing lots of stuff that has 1 impression only? How does Google actually determine what is a privacy issue if something that is a 0 click 1 impression term is allowed to show. @CJSlattery

What is the actual criteria for what gets hidden from the STR? I think some transparency on this would be welcome. We truly have no idea why things are being hidden other than claims of “privacy” without any actual information to back that up. @CJSlattery

Since these are drastic changes in Google Ads, will we ever be able to get all of our search query data back. This is like blind marketing without any benefits. @tweetgirlmem

Fair – but would it not be favourable/beneficial to smaller/low vol accounts to give them access to this data while G works on a suitable solution for them all? Imagine how much business they could be losing without sight of their own data? @AzeemDigital

Especially as now we have to charge smaller advertisers more due to the time it takes for us to do negative keyword research as well. @amaliaefowler

Can anybody elaborate more on how me seeing query data that does not align with specific sessions given the delay in importing into an ad account = a privacy decision? @amaliaefowler

Q7: Many people have heard from their reps that GCLID will not be as we have known starting in the fall. Is this true?

GCLID is/will still be available, but we’re no longer sending GCLID for iOS 14 traffic coming from ads on a handful of Google apps. See more here: support.google.com/google-ads/ans… @adsliaison

Q8: Why are the case studies always about big brands and e-commerce brands? Does Google care about SMBs and B2B? Why is B2B especially always ignored?

Last year, the roadmap was rewritten to help SMBs hit hard by the pandemic (incl. ad credits) & helping them sell online, new local biz features etc., and Pef. Max, lead form extensions, & easier conversion imports are examples of B2B-friendly features. @adsliaison

I’ll pass on the feedback about the case studies. In the meantime, you can check out several SMB partner case studies here: partnertestimonials.withgoogle.com @adsliaison

Q9: Can we get more insight into how Performance Max campaigns will work? Will they operate in the same way smart campaigns work or will we have more control/insights. Are they specific for lead gen, ecom or both? Can we optimise them or is it fully automated?

Performance Max is a goals-based automated campaign type for advertisers of all sizes. The Pef. Max beta is only available for advertisers if your primary objective is to generate leads or to drive online sales without using a product feed. @adsliaison

While there are similarities, retailers and advertisers with physical stores should continue to use Smart Shopping and Local Campaigns to meet their marketing performance objectives as they do today. We will continue innovating on these campaigns in 2021. @adsliaison

Perf Max supports Max Conv w/ optional tCPA and Max Conv Value w/ optional tROAS across Google’s ad inventory. Ad serving across campaigns will continue to be based on Ad Rank and likelihood of conversion. twitter.com/adsliaison/sta… @adsliaison

Oh! speaking of Locals, now that Waze is in the family, are they going to get store visits reporting? right now they’re still on Placed. @JuliaVyse

PPCChat Participants

Related Links

Leave a Reply

Your email address will not be published. Required fields are marked *