Posted by & filed under PPCChat.

Here is a complete screencap of this week’s PPCChat session, hosted by Julie F Bacchini. The discussion centered around Navah Hopkins’s article on overcoming bias in PPC to enhance ad performance.

Q1: Smart versus manual bidding – what is your position on which strategy to use and under what circumstances?

Typically I lean on smart bidding. This is based on working in large accounts likely to get sufficient data to be worth it, and to not want my team spending time fiddling vs strategizing. For smaller accounts or small localized campaigns – say a new store opening where signals won’t build fast enough for smart – manual can work really well. @JuliaVyse

Now ask me if my clients prefer to pay for what they see as ‘busy work’ or if they like to see LOTS in the change history. the answer will annoy me. @JuliaVyse

I firmly believe Smart bidding is only for established (90+ day old) accounts. That said, auto bidding (max clicks and target impression share) are totally fair to use in the early days. Bid caps a re a mustManual bidding is great for folks who will take an active role in managing their accounts and will never meet the 30+ (ideally 60+) conversions in a 30 day period. @navahhopkins

I love this question. Because philosophically, I really like the idea of smart bidding and taking advantage of all the extras it offers advertisers. But in reality, I have found it to be extremely frustrating for advertisers that are in the low conversion volume level (fewer than 50 per month). So, I spend time trying to find ways to get them the volume and return/cost per acquisition they need. @NeptuneMoon

For new & small budget accounts: Manual bidding/eCPC to begin with Smart bidding after conversion data has come in at desired levels consistently for a couple weeks. For existing campaigns/client accounts
Smart bidding is usually set up when we take over so we continue with minor tweaks.
For select tests on search, we still opt for manual bidding to test waters. @alimehdimukadam

My thoughts are:

  1. A new account or one that has very little data I will usually begin with manual bidding. This is to get enough data to hopefully go with automated bidding at some point.
  2. If the account has a lot of data, I will 100% use smart bidding, because that will do better for my brand than me watching and changing bids all the time.
  3. If a campaign is not doing well, sometimes I will go back to manual bidding to see if we can get it back in line. @Ichasse

B2B and Smart bidding are not yet friends. come on technology, find an answer! @JuliaVyse

Yeah, B2B is a whole different animal. @Ichasse

@JuliaVyse  I think the issue is B2B brands struggle with privacy compliance that goes with sharing true values for customers (or the tech integration is too messy so folks give up). Curious if you see it differently. @navahhopkins

The smart bidding tools are heavily tuned toward ecommerce. And I get it, that is an easier formula. But it does get tiring having to “igure out how to try to make ecomm centric features work for lead generation. @NeptuneMoon

B2B you really have to make sure smart bidding is getting down funnel signals so it knows the leads are good. @revaminkoff

This is one of the biases I think it’s worth confronting – ROAS and POAS don’t have to exclusively belong to e-commerce. however it does require more compliance management to get those numbers from B2B orgs. @navahhopkins

@navahhopkins  I also think there is a big disconnect between what platforms want and recommend relative to sharing customer and sales data with them and what companies are willing, able or comfortable doing (for a whole host of reasons). That is a big hurdle without any easy answers. @NeptuneMoon

@navahhopkins  I think some of it is privacy compliance, and I’m generally in that camp. You can’t just break Canadian privacy laws because Google or Microsoft call it a best practice. On the other hand, there is a habit of treating pretty light info as Pentagon level secure and use the legal team as an excuse to not share anything at all. As usual there’s a happy medium. My Gov and Pharma clients need to be extra careful and I support that. My fast food and retail clients…probably can share a little to get a little.@JuliaVyse

I am smart all the way, especially since I am building the campaigns and teaching business owners to manage at this point. I also have had it work SO well simply because I have a strategy which is crazy. @runnerkik

I think another interesting aspect to this is businesses also think essentially – if I share sales and/or profit data with Google, will that influence what we pay for clicks. There is a reticence there on that front that I don’t often see talked about. Think surge pricing but for ads. @NeptuneMoon

All bidding is “smart” – even CPC is eCPC. @ferkungamaboobo

warning – this is from someone who doesnt do account management anymore – I think the place for manual is small brands and highly competitive and volatile industries. smart bidding/bidding strategies for areas that have a lot of data and a stable industry. @TheMarketingAnu

The constant semantic changes are so real! and do not help clients wrap their heads around this stuff. They have more to do with their day than memorize more Google labels for activity. @JuliaVyse

I think it’s important too to de-jargonify — you can say something like “look, this is what it’s called in the platform, but I’m going to call it X because that’s a word a human would use.”This hit home for me with a question on video ads I got: “my CPV on Facebook is so much better than on YouTube, why is that?” @ferkungamaboobo

So much experience and insight in these answers. Y’all really know how to use the platform to achieve the best possible results for your clients. However, this should serve as a word of caution to any advertiser (particularly SMBs) trying to do this themselves that Google Ads is hard. @robert_brady

Q2: Performance Max campaigns, both generally and relative to brand terms – what is your position on using PMax and under what circumstances?

We now hold PMax office hours with one of our clients who cannot deal with the lack of accountability in performance changes YoY. So while it has some cool bells and whistles, I wouldn’t say it’s saved us a lot of time or made things easier. @JuliaVyse

My main struggle with PMax is that it mixes 2 (or more, depending on your definitions) channels. @ferkungamaboobo

I have complex thoughts on PMax…It represents a smushing together of awareness and search (direct intent) advertising and that is so foreign to not just us as PPCers but also businesses. Businesses have long been trained to have different expectations for awareness advertising and more high/direct intent advertising. And we did a lot of that training of them! Now along comes PMax and it says “forget all that!” And that is not a minor undertaking. It also muddies waters that clients are used to being clear and separate. @NeptuneMoon

From what I have heard, these are going to be the main campaigns for Google in the future. We may not have some of the campaigns we are used to having at some point. Having this in mind, it is in our brand’s best interest to find ways to make them work. Google has gotten rid of a lot of older stuff (match types, ETAs, etc….).Google is also leaning heavily into AI and what that will bring and PMax campaigns are better designed to take advantage of that. I say all this and out the other side of my mouth, I will say I am not a very big fan of them for B2B currently. @Ichasse

Accounts spending less than $5k per month likely won’t have enough conversions to meet the 60+ needed conversions in a 30 day period. Aside from that – I actually love PMax because it takes the bias out of it. That said, you need to do the work of splitting out asset groups and adding negatives/exclusions. @navahhopkins

Did anyone submit to Search Honours? All the categories were AI-focused, which = PMax adoption.@JuliaVyse

 @ferkungamaboobo the whole point is the mixing so we don’t ignore a channel that could serve us. I agree we need to rethink how we process metrics because of this. @navahhopkins

The client expectation question is so real. Like give me the tool, sure. But also give me a way to explain why the engagement rate did that. or why conversions did that. or which asset “worked” best. these are not questions we can easily answer, and it’s in a platform where everything supposedly is measurable. @JuliaVyse

 @Ichasse I think we all have seen the writing on the wall that PMax will be the only campaign type in the not-too-distant future. Google has not been opaque about that, IMO. @NeptuneMoon

The platforms don’t want us asking those kinds of questions any longer though @JuliaVyse and that is a HUGE shift. It is what it is but it will fall on us to help rightfully confused clients through all of these systemic changes.@NeptuneMoon

@navahhopkins I hear you, but I fundamentally don’t think Google is trying to eliminate bias with this tool. They’re finding a way to charge CPMs on search IMO. @JuliaVyse

Yeah, the explanation part, the analysis part is what I’m struggling with. The user path is so different between channels and even networks (Gmail ads vs YouTube vs Search) — what does it mean to group them together? You’re doing that top-level spend analysis on other data anyway. @ferkungamaboobo

@NeptuneMoon yeah, we have to work with the tools or we will be left behind. I have friends who refuse to use it still and they are seeing worse results in the regular campaigns and struggling. Google is taking our toys away if you don’t use the new ones.They control the platform and eventually, we will only have the choices they want us to have (like it or not). I see a lot of advantages to PMax and I also understand the frustrations of the brands who are used to crazy returns on smaller spends. @Ichasse

We can agree to disagree – PMax came about as the monopoly talks really picked up. My “tin hat” moment, is that PMax represents a way for Google to say we support all channels not just search (because they did get tagged for a monopoly in search text advertising but not search advertising) @navahhopkins

Mixed feelings about pMAX. Yet to see it work really well for smaller accounts. It’s more like a hit & miss, though we have dialled it in for a few Ecom accounts. Excluding – brand search. That’s the approach I am taking for now. @alimehdimukadam

@navahhopkins I do think you’re onto something with PMax being at least in part a preemptive step against data and monopoly claims/issues. @NeptuneMoon

fwiw, I use PMax a LOT and it’s not a tool we’re going to leave behind. I just wish it was a little more upgraded. @JuliaVyse

I just find it so odd since the DoubleClick acquisition is much more broadly actionable. Or rather, the impact of 20 years of that acquisition or whatever. @ferkungamaboobo

I appreciate the improvements made to PMax and honestly, I do love the campaign type. That said, some people outright shouldn’t use it: low conversion volume, low search volume, etc. Also, Microsoft’s PMax is super interesting. @navahhopkins

Scripts work well to see some insights, better than nothing. @alimehdimukadam

They look at their lack of video as a negative, but it’s a essentially a pass for all the practitioners who are afraid of video. @navahhopkins

@navahhopkins I need to get someone from Microsoft Ads on a chat this fall…@NeptuneMoon

Do you want me to ask? I can try to pull in our friends over at Optmyzr. @navahhopkins

That would be great! @NeptuneMoon

Oof, Microsoft’s inventory is uh, inoptimal. It’s a real challenge for them in display in general. @ferkungamaboobo

@ferkungamaboobo If I could get Xbox placements in Microsoft PMax, I would SPEND. SO. MUCH. @JuliaVyse

I don’t know, Microsoft has for many brands actually done much better (% improvement-wise) vs. Google this year for a lot of brands. Even folks I talk with are seeing some really big improvements. @Ichasse

Microsoft Audience network is 100% cheaper than Google inventory. The core issue is that Audience Network goes to MSN and Outlook most of the time and if you don’t set yourself up for remarketing, you’ll be sad. @navahhopkins

Something in the Q3 thread inspired me in terms of a multi-channel example: say I use PMax for a fast food resto. The video & display placements may show up for someone who generally likes fast food and is currently looking around for pizza. But a search ad or map ad would not make any sense for someone who definitely wants a slice when I only offer burgers. The ability to get tight and the way the negatives are applied makes a relatively complex action for what use to be pretty simple IMO.  @JuliaVyse

PMax has been doing very well when we have enough conversion volume to feed it, but we’re still giving it the assets versus letting the AI generate everything. @revaminkoff

Q3: Keyword structures (SKAGs, STAGS, match types,  etc.) – what is your position on which strategy to use and under what circumstances?

I’m a firm believer in STAGs > SKAGs. Which is why I forced myself to write out the defense of SKAGs (because we all need to confront biases). The core of it is I don’t think SKAGs are worth the time investment regarding negatives and budget to support them all. Especially given that close variants let you bid on whichever keyword champion you want. @navahhopkins

Definitely biased in favour of STAG vs SKAG. And now that exact match is basically the old broad match, I’m confident in that approach. @JuliaVyse

I do tight exact match and I keep it to 2 to 3 ad groups with 3 to 5 keywords.  I use match type as a lever. @runnerkik

SKAGs are a great example of something that worked swimmingly in a period of time. But then inevitable changes to keywords happened/are happening and that strategy become a lot less effective and efficient for most. (I’m sure there are some legacy SKAGS still doing well). @NeptuneMoon

STAGs  @alimehdimukadam

It is also clear that Google wants STAGs (search theme ad groups). They repeat it often. @NeptuneMoon

STAGs was new to me when someone asked about them a while ago – it just sounds like a normal campaign. @ferkungamaboobo

I wrote this back in 2019 – I stand by 90% of it (accounting for updates) @navahhopkins https://www.wordstream.com/blog/ws/2019/08/29/single-theme-ad-groups

google's recent match type changes

I do sometimes add a lot of keywords in one group with phrase match to test and check for search terms at a very low bid. Then build out tighter adgroups from that data. @alimehdimukadam

I am a firm believer in testing. Personally, I do favor STAG, but if something else works. I won’t shut down a SKAG campaign if it is doing amazing. Shoot, I still have some very old ad types that Google no longer supports that beat out the new ad types that I won’t shut off. @Ichasse

SKAGs just has always seemed like SO much squeeze for very little juice. @ferkungamaboobo

@ferkungamaboobo Google is a bit contradictory in their advice these days regarding ad group themes… They want themes that are discrete but not too discrete but also not too broad. It is very Goldilocks right now. With a big dose of “when in doubt, you can always let the AI figure it out for ya…@NeptuneMoon

Yeah – I mean I don’t think I’d separate out attorney/attorneys/lawyer/law firm/lawyers/law firms anymore. despite regionalisms @ferkungamaboobo

No truer words were ever spoken. in both cases. SKAGs are a lot of work for not a ton of return for the effort and Google still has not perfected the themes yet enough to know what works “the best” quite yet. It is on us to test how discrete we can go. @Ichasse

Having swapped geo modifiers match each other is also nice these days. @ferkungamaboobo

To be fair, this push to consolidate started with keyword data getting smushed together. Remember when you could easily see:

  • predicted keyword data for individual terms
  • query data for nearly all terms

With both being significantly decreased, they next step push to just go themed kinda had to be followed. Throw in the constant loosening of match types and here we are. @NeptuneMoon

The thing is – we do still need to do the work to find our keyword champions. Knowing if lawyer or attorney will do better in a market is a critical strategic task for PPC managers I just don’t see in a lot of accounts. @navahhopkins

I’d be happy if I didn’t have to make competitor negative lists for all accounts honestly @navahhopkins (cause they match to so much now). And I would love to still get data that would definitely tell me if lawyer or attorney performed better. That is that hardest part of all this for me, honestly. @NeptuneMoon

Google trends is helpful for that @navahhopkins

Search term N-gram analysis but you run into sample size issues with robustness of data. @ferkungamaboobo

You can get auction prices and heat maps for search terms and topics. @navahhopkins

@navahhopkins Yes and no. Once you get regional, it gets fairly light. @JuliaVyse

Q4: Do you think you have any biases against platforms? If so, which ones and why??

I don’t know if it is a bias as much as resignation that things are not what they were 10 years ago or even 2 years ago right now. And the constant need to figure out “what does that mean in practical terms for the accounts I am responsible for…” @NeptuneMoon

I HATE Reddit. @navahhopkins

I was very much pro-Google. 11/10 with the level of details & controls we had. Now I am at a 7/10. And my liking of Meta as a platform has gone up by a couple points. @alimehdimukadam

Personal bias… I don’t do TikTok. Ain’t no way I am giving that app access to my phone and it’s data. @NeptuneMoon

Programmatic Display. Chumbox Ads. Amazon Display (Search is v fine). Twitter. Reddit. Quora. @ferkungamaboobo

I’ll include Twitter/X in the dislike/disdain category (content and controls) @navahhopkins

I think everyone here knows how I feel about Mark Zuckerberg and his stupid website he wants me to update for him. Not a fan of Criteo and some of the black box folks out there. And twitter!!! used to be my joy. @JuliaVyse

Amazon is trying and I appreciate it @navahhopkins

I’m super curious about Amazon video ads. @ferkungamaboobo

@ferkungamaboobo I’m a fan! they do really well up here. @JuliaVyse

Ohh yes, I love Amazon coz no attribution issues. For stores advertising on Amazon. Gets a lil murky when we add branding ads to the mix but everything is within the platform. @alimehdimukadam

I do in some ways. I know Google is still the best platform out there (currently), but they are getting worse. Amazon for ecom is actually not far from overtaking Google for revenue vs. spend. Microsoft has also been improving over the past year and if they could get the same volume as Google, I think they would be big $$$ for advertisers vs. Google.Facebook, well they overinflate their revenue numbers by quite a bit and not a big fan of their 1 day view, 7-day click window and that is the most conservative you can set it to. @Ichasse

@ferkungamaboobo I sat next to a guy who upped his investment from $ 3 million to $45 million per year on Amazon video ads because of the ridiculous good return – he sells vet supplies. the product is great. but I agree there’s room to grow on the programmatic display side. @navahhopkins

@JuliaVyse Are you not impressed by the renaissance of Zuckerberg? From Senator we run ads to I am cool with a chain in the metaverse. @alimehdimukadam

Here’s a positive bias I’ll declare: I really like LinkedIn. Yes, it’s pricey, but it’s specific, actionable, and they don’t inundate users with ads every scroll. @JuliaVyse

LinkedIn is lovely! @navahhopkins

@alimehdimukadam  not so much impressed as horrified, hoping it sinks into the sea. @JuliaVyse

Ha, I just struggle with “who is actually on this platform” with LinkedIn. @ferkungamaboobo

I admit I have a strong bias in favor of Microsoft because I believe in their ethics. @navahhopkins

Let’s be real @alimehdimukadam Zuck is kicked back enjoying the hell out of Google’s legal woes and Elon going full Bond villain. It’s his best year ever. @NeptuneMoon

Yeah, LinkedIn is just solid. Not flashy and not terrible. Just a hard-working part of the marketing mix! @Ichasse

Haha yeah, true that! And in my opinion, I think he has really worked on his public speaking skills & comes across more likeable. @alimehdimukadam

I struggle with that a lot with most ad networks. Who is this for? Why does this exist? Why do these ads exist? Who would click on that in the mode they’re in? @ferkungamaboobo

@ferkungamaboobo  that’s the discussion I regularly have with clients. The goal of the platform is to keep people on the platform, so looking to TikTok/ig, etc for ‘traffic building’ feels a little counterproductive. @JuliaVyse

Right – which doesn’t mean it’s a bad call! Just a different one that ye olde impression->click->conversion->rebuy cycle. @ferkungamaboobo

@ferkungamaboobo  As someone who has worked in an ad network, it’s mostly B2B feeds & traffic exchange. Google feeds. Yahoo feeds by side Criteo Pubmatic and the likes just auction it to different buyers by a group of syndicated sellers. It’s a big market. And there is a lot of junk that goes around too.
A good quality network needs to have a lot of traffic filters in place.@alimehdimukadam

I don’t miss yahoo paid search cpc arbitrage. @ferkungamaboobo

The monies that were made off of those….crazy amounts. There would be chargebacks too. @alimehdimukadam

(I do miss Gemini as a platform though, RIP) @ferkungamaboobo

Q5: Do you have a natural trust or distrust of advice or recommendations from platforms? Does it vary by platform and why?

I don’t have a distrust, just a frustration that they seem to ignore the state of the account. I think Google burned the bridges for all the other ad networks so now all ad networks suffer from a strong mistrust. @navahhopkins

I’m not sure I would call it distrust, but I do keep front and center in my mind that ad platforms serve themselves/their shareholders first and foremost. And, that means that things they do, do not always align with what is best for advertisers in general and/or any single account. If you keep that in mind, their behavior can still be super frustrating and sometimes confounding, but it does make sense. @NeptuneMoon

YES! but it’s contextual: almost none of the recommendations from the machines or AMs are relevant to public sector. weirdly LinkedIn and Reddit teams do tend to get it and recommend okay stuff for us. @JuliaVyse

Google’s antitrust testimony helped no ad network on the trust front either! I also think that Google’s recommendations have a long way to go to be helpful for most accounts. If their goal is to have a lot less actual hands-on management of accounts (which feels like a logical conclusion to draw) then it would be nice to see recommendations that were more grounded in the specific account. And maybe that is too dependent on advertisers sharing loads of their internal data with Google? But as it generally stands now, the recommendations are often the same across different accounts. And that hurts credibility. @NeptuneMoon

Most platforms tell you what you need to know in their documentation. The specifics are more murky, but I feel it’s similar to SEO: they’re trying to make the inventory better and work, but sometimes making chicken nuggets is a bit…. less than ideal to be FAR too charitable. I’m more ambivalent with the secret sauce community. @ferkungamaboobo

I think I trust Claude more than a Google sales representative. @alimehdimukadam

The reps for Google are literally paid to make advertisers do things that may not be in their best interest, so yeah there is no trust. @NeptuneMoon You are spot on. Their responsibility is not to the brands, but to shareholders and themselves first. If that means brands get something good too, then great, but if not oh well. @Ichasse

Understand they are doing their jobs but you can’t be pushing auto-apply throughout for everyone without understanding the nuances. @alimehdimukadam

Like, McD’s sauce is just ketchup and mayo, maybe with some pickle juice. You can call that a secret, or call that a (tbh lazy/quick) remoulade. Same with advertising. @ferkungamaboobo

PPCChat Participants

Related Links

Stop wasted ad spend with Karooya

Stop the wasted ad spend. Get more conversions from the same ad budget.

Our customers save over $16 Million per year on Google and Amazon Ads.

Leave a Reply

Your email address will not be published.