Why Top Advertisers are Going Back to Manual Bidding

Top advertisers are quietly shifting away from automation—and for good reason. 

In this episode of Google Ads Unleashed, host Jeremy shares why manual bidding is making a comeback, especially for brand campaigns, low-volume accounts, and high-CPC segments. Learn where smart bidding fails, how to run strategic A/B tests, and why taking back control of your bids might be your next best move. 

Get your free 30 minute strategy session with Jeremy here: https://www.younganddigital.marketing/

Scale your store with 1:1 coaching: https://www.younganddigital.marketing/1-2-1-coaching

Transcript

Hello everyone, and welcome back to Google Ads unleashed. I hope everyone is doing fabulously this fine Monday. So today I want to be talking about something that I've observed over many months, but really haven't only, only recently, I've really sort of rediscovered that not a lot of people are doing this, or only the best are doing this, and that is to not abandon manual bidding. What do I mean by that? You probably have sort of heard the old mantra and the old sort of stuff that Google is spewing, that, you know, you just got to trust, the algo, that you have to lean into, big data, that you have to do all this sort of stuff. But the reality is, do you actually have to right? A lot of people I know actually switching back to manual bidding strategies, especially in a lot of sort of cases. And in this episode, I'm going to be talking about what Google has promised, what is, you know, the sort of best practices that Google says. But why automation doesn't always actually outperform manual bidding. Some examples which even we have run, where manual bidding has actually improved performance. And also I'm gonna give you sort of a few tips for the future. Let's dive right in. So like I said, smart bidding would outperform manual bidding probably in 99% of cases. And I would argue that is actually true, because I remember very well when we moved away from manual bidding strategies to TCPA and with a lot of features of Google, very often when it's first introduced, it's actually really bad. And when it was, I think 2018 or something like that, Target, CPA was a new thing. Oh, my God, you could automate the bidding. And it was absolute garbage, right? It was terrible. It just didn't work at all. It was just really poor. So as a result of that, I personally, to be honest, had, like quite a bit of a sort of aversion to manual bid strategies for very long time. In fact, maybe to my dismay, I was one of the late adopters of T ROAs and TCPA. It didn't really hurt my career, I'd say, but I've had a healthy skepticism towards new tools that Google brings out for very long time. Over time, however, this has changed completely. Google is actually massively outperforming us most of the time, there's no way you can run, you know, sort of manual bid strategies on 100 plus keywords anymore. Most of the time you don't even have the time anymore, right? You used to be able to do this with very few accounts, but most people are managing six, eight accounts, or they may have two, three in in house accounts, or they have their own account, and you just don't have the time really to do that. So minor bid strategies are not really an option, and like I said, it mostly outperforms what what you do anyway. So we've run this AV test plenty of times, especially in higher volume accounts. It just always will work. Troas will always outperform any manual bid strategy. TCPA will as well. Then, of course, we have other big promises. P max was one. First of all, broad match was one of them introduced a few years ago. Now, since then, been upgraded, which is meant to combined with smart bidding, expand, you know, our advertising efforts even further. We of course, have p max as well, which was promising to, you know, more or less cover a complete full funnel strategy for you, and set the bids completely at the right time, at the right place. But in the past year or so at least, I have seen that cracks have been falling in those promises, and I think there are a couple of areas where still these kinds of strategies don't really work, or where manual CPC is, let's say where automation to that degree is actually fallen short of the manual version and the more sort of controlled version of running PPC. I want to first talk about brand. Search so, as everyone knows, if you actually run branded search campaigns in your ad account most of the time, you probably should, sometimes maybe shouldn't. Have got plenty of episodes on that, which you can probably go back and listen to. You will see if that if you use any automated bid strategy, such as TCPA, T ROAs, etc, in 99% of cases, you will overpay for your CPCs. The reason is pretty straightforward. The that Google doesn't actually understand that branded search traffic has a kind of different value to non branded traffic. Okay? All it sees is, hey, this kind of search term is generating a crazy return on ad spend. Let's bid more for that, right? Even when you set a very, very conservative Troas, it'll still try and push for volume, because it understands that this type of search term will always convert regardless of kind of what bits you set right. And of course, Google will always choose maximizing their profits. So which is why you'll see and you'll see this very often. You can even do this experiment now, if your brand search is on TCPA or on max conversions, or it's on Troas or something like that, which Google, by way, will always recommend, the Google rep will be on the phone and say to or this needs to be on TCPA or whatever. If you switch it to manual run a B test, you almost always will get either the same number or more conversions and lower CPCs. And that is where you just see that branded traffic just needs to be controlled and the same Google tried with p max, right? So for a long time, Google tried or, but they still have to kind of incorporate that kind of branded traffic into p max and hide it within the search terms. Now the search terms there, you can see that if your brand, if your p max is running for branded searches, you can, of course, do branded exclusions, and you can split brand or non brand traffic properly. But again, you will see that if you actually analyze sort of the median of those CPCs in your in your brand campaign, that in your p max, that you will always see that Google actually massively inflates branded clicks there as well with where manual bidding would work a ton better, which is why that's how we set most ad accounts. We will have specific p max setup usually catch all shopping on high priority. All of those will have brand negated, and then we'll have a low priority and medium priority shopping campaign, which will be almost always on manual CPC or max clicks with a bit cap one of the two, and it works just better. Simple as that, you can just much better control your bids for your own branded traffic, and you will see that, yeah, it just outperforms another area where automation really falls short is in low volume accounts. So this is maybe something bit of a learning, even for ourselves, in the past, sort of half a year, is that very often when you have, let's say, Poor attribution, right? So not as much as many data points, let's say, even along a more complicated funnel, such as some sort of offline conversion tracking, or if you have a lot of low volume, like in terms of search volume, or if you have a longer sales cycle, that means from first touch point to further down the line, People buy in, like much longer. Think of SaaS, for instance. Then often automation is a bit erratic and it doesn't actually perform as well. I found this, of course, nuances to all of this, right? However, generally speaking, with volatile conversion data, or even when you start a new ad account, this is a very good example, right? You can't start a new ad account with automated bid strategies because of exactly that issue. You don't have enough or very volatile data. That is where manual CPC does really well, real life. Example of this, we have a client that basically sells a specific kind of software is a legacy client. We don't really
take on this kind of business anymore, but it's a legacy client. And basically they have maybe a few 100 searches for their specific kind of software per month. And the reality is, is that a manual CPC outperforms there, right? And the reason is pretty specific, because they have extremely low conversion volume. They don't have offline conversion tracking because they are so GDPR compliant that they're afraid of transmitting certain data to Google, which, in my opinion, is idiotic, but they can't be helped, and that's why their legacy client for that exact reason. But no, they won't listen. They don't want to, they don't want to do that. But also, automated bid strategies simply fail because they don't have enough data points, right? So that is where manual CPC simply outperformed, and we just have much tighter control over where we want to be in the auction. Something where automation falls short as well is that sometimes CPCs just creep up with no explanation or any levers that you can pull. For instance, you've probably seen this before in certain accounts that you switch, let's say to even a higher T ROAs, let's say a campaign has done really well. It's on T Ross three, and it's consistently been performing at four and five, even when you increase spends. And then you think, you know what? Now I can actually adjust the Ross target, maybe to four or five. That because it's doing well, right? I have actually seen campaigns not only start and spending more, which is completely counter intuitive, but also then driving higher CPCs, because it learns of this new paradigm, which makes absolutely no fucking sense, but that is where automation just falls short. You've probably seen the same. If you've ever gone into search terms? Right? You will sometimes see, let's say you run TCPA or or max conversions, and you don't use a portfolio bid strategy where you can limit the bid cap that you'll simply see search terms with like something like 10x the account, CP, CPC, right? I went into an ad account yesterday with, with, with a coaching client, and we went into his branded traffic, which was a max conversions at the time, and his brand, plus the keyword discount code, had a CPC of 37 euros. That is fucked up, right? Like, and that's just, and that's where automation simply falls short. Like I said, we had this example quite a few times. For instance, the brand thing, we've seen that over and over again, right? Branded search just needs manual bid caps. It's also another example that we had is we have one client at the moment where a lot is on broad match because they really wanted to test it, and it just drained the budget instantly, right with with smart bidding, and it just didn't make any sense manual CPC doesn't make any sense with broad match anyway, but so we didn't even test that. But yeah, so just automation plus broad match, and took full automation just simply was a wasted budget, right? So, yeah, I think manual CPC is making a comeback, and I think there are few cases where this is still ideal. So brand campaigns have already mentioned both in shopping and in search. I think low volume B to B is manual CPC is absolutely fine, or local, PPC, manual CPC, absolutely fine. There's nothing wrong with that whatsoever. It's almost like a lost craft, if you like. I suppose if you have some form of seasonality, that could work pretty well as well. Let's say you want to dominate something for a very short amount of time. Boom. Manual CPC is great. Or sometimes you simply know more than the algorithm, right? You've probably observed this yourself, where you think you know it is. It's hike in the CPCs. Nonsensically with any anything that I'm trying just run everything on manual CPC. Boom. You're You're sorted, right? I think with remarketing, it can work pretty well as well. We use this for RSAs, for instance, very often, right? So if we run RLSA both on shopping and on search, I think manual CPC is pretty good. I think if you maybe want to, when you start an account. I think it's a very, very good idea to use manual CPC as well, or simply for tests, all right. So it's it's just a really handy thing to have. So I think what this means for you, the dream of full automation is still. Ongoing, but blind trust is something that I would highly doubt if I were you, so I would probably always double check and question myself. Is full automation really always the way to go and and the better solution, right? So look at your accounts, look at certain sections. Could this actually be performing better if I took a little bit more control back? I think this is particularly helpful for anyone where CPCs are going out of control, or ROAs is shaky, or if you just need something to shake things up, right? So to try something different, especially in sort of low volume accounts and low conversion accounts. And if that has been interesting to you, then please subscribe to the pod. Hopefully some of the tips have been helpful for you. You can always ask me about it as well on LinkedIn Jeremy and Google ads. You'll also find me on young and digital dot marketing. Our website, we don't have any spots anymore. For August, our doors are closed, but for September, we're opening our doors already ready for the q4 time, if you want to take your q4 performance to the next level. Until then, we also have one to one coaching still available. There's been one spot so going in q3 but of course, that's sort of the end. Is an eye of q3 already. I know it's only just about August, but yeah, only one or two months ago, we'd have to speed through it. If this is something of interest to you, do let me know. Other than that, I wish you a happy and productive week ahead, and this has been Jeremy young with Google ads and niche Thank you very much.