{"id":24691,"date":"2024-08-14T23:46:35","date_gmt":"2024-08-14T18:16:35","guid":{"rendered":"https:\/\/www.karooya.com\/blog\/?p=24691"},"modified":"2024-08-14T23:52:14","modified_gmt":"2024-08-14T18:22:14","slug":"ppcchat-strategic-tactical-biases-in-ppc","status":"publish","type":"post","link":"https:\/\/www.karooya.com\/blog\/ppcchat-strategic-tactical-biases-in-ppc\/","title":{"rendered":"PPCChat | Strategic & Tactical Biases in PPC | 13th August 2024"},"content":{"rendered":"\n

Here is a complete screencap of this week\u2019s PPCChat session, hosted by Julie F Bacchini<\/a>. The discussion centered around Navah Hopkins<\/a>\u2019s article on overcoming bias in PPC to enhance ad performance<\/a>.<\/p>\n\n\n\n\n\n\n\n

<\/div>\n\n\n\n

Q1: Smart versus manual bidding – what is your position on which strategy to use and under what circumstances?<\/strong><\/h4>\n\n\n\n
<\/div>\n\n\n\n

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<\/a><\/p>\n\n\n\n

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<\/a><\/p>\n\n\n\n

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<\/a><\/p>\n\n\n\n

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<\/a><\/p>\n\n\n\n

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<\/a><\/p>\n\n\n\n

My thoughts are:<\/p>\n\n\n\n

  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.<\/li>
  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.<\/li>
  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<\/a><\/li><\/ol>\n\n\n\n

    B2B and Smart bidding are not yet friends. come on technology, find an answer! @JuliaVyse<\/a><\/p>\n\n\n\n

    Yeah, B2B is a whole different animal. @Ichasse<\/a><\/p>\n\n\n\n

    @JuliaVyse<\/a> \u00a0I 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<\/a><\/a><\/p>\n\n\n\n

    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 \u201cigure out how to try to make ecomm centric features work for lead generation. @NeptuneMoon<\/a><\/p>\n\n\n\n

    B2B you really have to make sure smart bidding is getting down funnel signals so it knows the leads are good. @revaminkoff<\/a><\/p>\n\n\n\n

    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<\/a><\/a><\/p>\n\n\n\n

    @navahhopkins<\/a><\/a> \u00a0I 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<\/a><\/p>\n\n\n\n

    @navahhopkins<\/a><\/a> \u00a0I 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<\/a><\/p>\n\n\n\n

    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<\/a><\/p>\n\n\n\n

    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\u2019t often see talked about. Think surge pricing but for ads. @NeptuneMoon<\/a><\/p>\n\n\n\n

    All bidding is “smart” – even CPC is eCPC. @ferkungamaboobo<\/a><\/p>\n\n\n\n

    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<\/a><\/p>\n\n\n\n

    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<\/a><\/p>\n\n\n\n

    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<\/a><\/p>\n\n\n\n

    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<\/a><\/p>\n\n\n\n

    <\/div>\n\n\n\n

    Q2: Performance Max campaigns, both generally and relative to brand terms – what is your position on using PMax and under what circumstances?<\/strong><\/h4>\n\n\n\n
    <\/div>\n\n\n\n

    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<\/a><\/p>\n\n\n\n

    My main struggle with PMax is that it mixes 2 (or more, depending on your definitions) channels. @ferkungamaboobo<\/a><\/p>\n\n\n\n

    I have complex thoughts on PMax\u2026It 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 \u201cforget all that!\u201d And that is not a minor undertaking. It also muddies waters that clients are used to being clear and separate. @NeptuneMoon<\/a><\/p>\n\n\n\n

    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<\/a><\/p>\n\n\n\n

    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<\/a><\/a><\/p>\n\n\n\n

    Did anyone submit to Search Honours? All the categories were AI-focused, which = PMax adoption.@JuliaVyse<\/a><\/p>\n\n\n\n

    \u00a0@ferkungamaboobo<\/a> 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<\/a><\/a><\/p>\n\n\n\n

    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<\/a><\/p>\n\n\n\n

     @Ichasse<\/a> 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<\/a><\/p>\n\n\n\n

    The platforms don\u2019t want us asking those kinds of questions any longer though @JuliaVyse<\/a> 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<\/a><\/p>\n\n\n\n

    @navahhopkins<\/a><\/a> 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<\/a><\/p>\n\n\n\n

    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<\/a><\/p>\n\n\n\n

    @NeptuneMoon<\/a> 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<\/a><\/p>\n\n\n\n

    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<\/a><\/a><\/p>\n\n\n\n

    Mixed feelings about pMAX. Yet to see it work really well for smaller accounts. It\u2019s more like a hit & miss, though we have dialled it in for a few Ecom accounts. Excluding – brand search. That\u2019s the approach I am taking for now. @alimehdimukadam<\/a><\/p>\n\n\n\n

    @navahhopkins<\/a><\/a> I do think you\u2019re onto something with PMax being at least in part a preemptive step against data and monopoly claims\/issues. @NeptuneMoon<\/a><\/p>\n\n\n\n

    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<\/a><\/p>\n\n\n\n

    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<\/a><\/p>\n\n\n\n

    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<\/a><\/a><\/p>\n\n\n\n

    Scripts work well to see some insights, better than nothing. @alimehdimukadam<\/a><\/p>\n\n\n\n

    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<\/a><\/a><\/p>\n\n\n\n

    @navahhopkins<\/a><\/a> I need to get someone from Microsoft Ads on a chat this fall\u2026@NeptuneMoon<\/a><\/p>\n\n\n\n

    Do you want me to ask? I can try to pull in our friends over at Optmyzr. @navahhopkins<\/a><\/a><\/p>\n\n\n\n

    That would be great! @NeptuneMoon<\/a><\/p>\n\n\n\n

    Oof, Microsoft’s inventory is uh, inoptimal. It’s a real challenge for them in display in general. @ferkungamaboobo<\/a><\/p>\n\n\n\n

    @ferkungamaboobo<\/a> If I could get Xbox placements in Microsoft PMax, I would SPEND. SO. MUCH. @JuliaVyse<\/a><\/p>\n\n\n\n

    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<\/a><\/p>\n\n\n\n

    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<\/a><\/a><\/p>\n\n\n\n

    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<\/a><\/p>\n\n\n\n

    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<\/a><\/p>\n\n\n\n

    <\/div>\n\n\n\n

    Q3: Keyword structures (SKAGs, STAGS, match types,  etc.) – what is your position on which strategy to use and under what circumstances?<\/strong><\/h4>\n\n\n\n
    <\/div>\n\n\n\n

    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<\/a><\/a><\/p>\n\n\n\n

    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<\/a><\/p>\n\n\n\n

    I do tight exact match and I keep it to 2 to 3 ad groups with 3 to 5 keywords. \u00a0I use match type as a lever. @runnerkik<\/a><\/p>\n\n\n\n

    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\u2019m sure there are some legacy SKAGS still doing well). @NeptuneMoon<\/a><\/p>\n\n\n\n

    STAGs \u00a0@alimehdimukadam<\/a><\/p>\n\n\n\n

    It is also clear that Google wants STAGs (search theme ad groups). They repeat it often. @NeptuneMoon<\/a><\/p>\n\n\n\n

    STAGs was new to me when someone asked about them a while ago – it just sounds like a normal campaign. @ferkungamaboobo<\/a><\/p>\n\n\n\n

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

    \n
    \"google's<\/a><\/figure><\/div>\n\n\n

    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<\/a><\/p>\n\n\n\n

    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<\/a><\/p>\n\n\n\n

    SKAGs just has always seemed like SO much squeeze for very little juice. @ferkungamaboobo<\/a><\/p>\n\n\n\n

    @ferkungamaboobo<\/a> Google is a bit contradictory in their advice these days regarding ad group themes\u2026 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 \u201cwhen in doubt, you can always let the AI figure it out for ya\u2026@NeptuneMoon<\/a><\/p>\n\n\n\n

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

    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<\/a><\/p>\n\n\n\n

    Having swapped geo modifiers match each other is also nice these days. @ferkungamaboobo<\/a><\/p>\n\n\n\n

    To be fair, this push to consolidate started with keyword data getting smushed together. Remember when you could easily see:<\/p>\n\n\n\n