Practical Advice on How to Actually Make Smart Bidding Work for You

Episode Description

Fred Vallaeys and Shawn Walker dive into what it takes to make Google’s Smart Bidding work effectively, especially for smaller advertisers. They discuss why Google now prioritizes audience signals over keywords, how low-volume keywords are becoming harder to target, and why a simplified account structure leads to better results.

For those struggling with low conversion volume, they share how soft conversions can train the algorithm. They also cover the balance between automation and manual control, offering practical strategies to optimize performance while staying in control.

Episode Takeaways

Google’s Smart Bidding promises efficiency and performance with minimal manual intervention — but it doesn’t always deliver the results you expect.

Here are the key takeaways from Fred and Shawn’s conversation to help you figure out how you can use smart bidding strategically to maximize performance without overspending.

1. Role of conversion thresholds in smart bidding performance

Smart bidding analyzes past conversion data to identify which keywords, audiences, devices, time of day, and locations actually drive conversions. If you have too few conversions it may be difficult for the algorithm to find meaningful patterns.

While Google recommends at least 15 conversions over a 30-day period, Shawn recommends having at least 30 conversions over 30 days per campaign.

“‘I think the safe bet is to look for 30 conversions over 30 days per campaign. The more data you have, the more accurate smart bidding becomes. What we see is if you don’t hit those 30 conversions in one campaign per 30 days, your campaign starts to get a bit unstable” says Shawn.

Further, our latest study on the impact of bidding strategies found that advertisers who clear upwards of 50 conversions in 30 days see better performance than accounts with lower conversions.

If you do not have sufficient conversions, consider using micro-conversions (add to cart, sign up for a free trial, book a demo, etc) to train the algorithm. Ensure you’re assigning the right conversion values for each action so Google knows how to allocate your budgets.

2. The increasing importance of audience signals

With how search is progressing these days, keywords are not the only factor that determines ad relevance. This is because Google’s AI is becoming more sophisticated in trying to detect the intent behind a user’s search. It uses signals like user intent, browsing history, income, and many others to find users who are actually ready to purchase a product or service.

“We live in a day and age where signals are becoming even more important than keywords. Google is leaning more towards targeting a lot of broad keywords. You can’t guess intent on a keyword alone, you’d have to work off in-market signals.” Shawn shared.

What this means is that keywords are becoming less predictable since Google uses a combination of broad match and audience signals to decide whether or not your ad is relevant to a user.

3. Focus on well-structured campaigns over multiple, fragmented ones

Traditionally, advertisers have worked with complex campaign structures with multiple campaigns and ad groups that go after very specific long-tail keywords. The problem with this approach is that campaigns become increasingly difficult to optimize.

Shawn advocates for a simpler approach.

“If you had to choose one priority, I would say to have a structure that is in line with getting enough conversions. Every account that we have now starts simple and we make it more robust over time as opposed to our earlier days when we started with huge account structures and giant campaigns.”

Clearly, it’s wiser to opt for a smaller number of campaigns that are super focused on budget, geographic targeting, and very specific conversion goals. When you have clear campaign goals combined with accurately set-up conversion tracking, Google’s automation is able to optimize your campaigns for better results.

4. Use automation wisely

The key principle behind automation, especially within Google Ads is to automate things like device bid adjustments, ad scheduling, budget adjustments, etc. It reduces manual intervention and gives room for customization.

For instance, Shawn’s team uses custom scripts and tools like Optmyzr for more precise control over their campaigns.

But it is important to remember that not everything can be left to automation. You still need to closely monitor performance and make strategic adjustments for optimal performance. Without careful monitoring, automation may result in unintended consequences like cost spikes, poor targeting, or missed conversions.

Get the most out of Smart Bidding with Optmyzr

Although smart bidding promises to maximize conversions and conversion value, it is far from a set-it-and-forget-it solution. It works best when you combine it with sufficient conversion data, proper audience signals, and accurate conversion tracking.

A tool like Optmyzr helps advertisers make the most out of smart bidding by helping them with custom reports, audience insights, smart spend allocation, and competitor insights. This gives them more control over their campaigns while still saving time.

Not an Optmyzr customer yet? Now’s the best time to sign up for a full functionality 14-day free trial.

Thousands of advertisers — from small agencies to big brands — worldwide use Optmyzr to manage over $5 billion in ad spend every year.

You will also get the resources you need to get started and more. Our team will also be on hand to answer questions and provide any support we can.

Episode Transcript

Frederick Vallaeys: Hello, and welcome to another episode of PPC Town Hall. My name is Fred Vallaeys. I’m your host. I’m also CEO and Co-Founder at Optmyzr, a PPC management software. So for today’s episode, we have as guest Shawn Walker from Symphonic Digital, he’s been doing PPC for a long time, and he figured out that the strategies that used to work in the past no longer work.

We all know that. But Shawn went deep on that, and he figured out how he gets smart bidding and automated bidding to work in today’s framework where there’s so much automation from ad platforms like Google. How many conversions does it take

What bid strategies to run? What to do if you don’t have enough conversions?

So I’m looking forward to hearing all of this very practical advice that I think will benefit all of us. So with that, let’s get rolling with another episode of PPC town hall.

All right, Shawn, welcome to the show. Good to have you on.

Shawn Walker: Yeah. Good to be here. Hello, Fred.

Frederick Vallaeys: Hey, a fellow Californian, right? And this doesn’t happen very often. California, but I live in California. Hey, tell us where you live and what’s your background.

Shawn Walker: Yeah, of course, I’m in San Diego right now.

So we’re in a good weather place. Everyone gets jealous yeah, I moved from Australia. So I ran a performance team out there for about three years. And then before that was in LA, but my career started at Yahoo when Yahoo was cool. Now, when you have a Yahoo email, people look at you funny.

But that’s where I got my start. So I was actually on the publisher side and then ended up working all through PPC. So whether it’s search or social display even worked at SEO as well. And my big background, I think I’ve done this probably for about 13, 14 years going on, but I’m a closet nerd. I will say we live in a world where there’s so much free education.

And I took advantage of that. So got certified in Google Analytics and lots of coding platforms. So JavaScript, PHP, Google scripts, all that, and any certification you can get your hands on. I got my hands on it. So lots of education there. But really, I guess my bread and butter at Symphonic is I’m the lead for strategy and product.

So when it comes down to putting strategy together, I orchestrate the team. I put together strategy with the team as well as provide technology and products for whether it’s reporting or lead generation. We’ve built even CRMs at one small point in my life, actually built an onboarding platform because I do a little bit of coding on the side to help bring new employees on.

So I’m really into the tech part. And we live in a really interesting world with A. I. and things like that now. But that’s a little bit of my background.

Frederick Vallaeys: Yeah, no, that’s a fantastic background. And I feel like in California either everybody has a script ready to go and sell it to Hollywood or everybody’s building a piece of technology trying to be the next Unicorn SaaS company.

So sounds like you’re just one of them now between Australia and San Diego, do you like surfing or do you like the beach? That’s what’s with those two places.

Shawn Walker: Fishing and music. That’s my go-to. So what’s great is the ocean is 10 minutes away from my house and then you have lakes that are 10 minutes away and mountains that are 10 minutes away.

So everything’s close. Big bass fishermen. So I like to go out, get away from the computer, and stare far away rather than up close which helps. And then I also play some music as well, playing a yacht rock band. So that’s pretty fun.

Frederick Vallaeys: Very cool. Good. It’s great to have you on the show.

There are a lot of topics I want to talk a little bit more about, like education what kind of software you build, and how you’ve been using GPT. But let’s get into the key topic here today, which was about bid management, right? And how do you outsmart Google’s conversion?

Based on bidding algorithms, I think one of the first questions that most advertisers always have is how many conversions I need before I can trust Google’s algorithms to start doing a good job. Can you talk a bit about those numbers?

Shawn Walker: Yeah. So Google is just following a suite of statistics, right? So statistics says if you remember in school, you need a sample size of 40 to get some kind of statistical significance and a reasonable level of accuracy. Google brings this to 30, but I’ve seen this jump around in some cases, even 15. But I think the safe bet is 30 conversions per 30 days per campaign.

That’s the important part. Even if you have 30 conversions in your whole account, it doesn’t matter. It’s all about each campaign, which I know we’ll probably get into this, but your structure is gonna change things, right? If you have 10 campaigns versus one. Then Google is going to have a harder time with the 10 campaigns versus one to try to reach that statistical significance.

Frederick Vallaeys: So it sounds like that need for a certain number of conversions plays a heavy factor in how you structure your account ultimately. And then now in terms of folks who have maybe much bigger accounts and have hundreds of conversions or those that have much smaller accounts. We’ll talk about that as well, but like how much sensitivity have you seen to this number of 30?

If you go from 30 to 60, do you get like twice as good performance because Google has just that much more data, or is that like a diminishing return? What have you seen there?

Shawn Walker: Yeah. This is such a big difference from when I started. So now we see a lot of sensitivity, right? So the more data you have, the more accurate you’re going to be.

But what we’ll see is if you don’t hit those 30 conversions in one campaign per 30 days, your campaign acts a bit wonky. Google tries to spend as much as it can in the beginning when you’re on, let’s say, max conversions or some kind of conversion-based bidding strategy. To learn and that learning phase can kill some small businesses versus big.

You can handle it. You got the budgets. What we’ve seen with bigger businesses is let’s say you have a thousand conversions, let’s just go crazy. The system does a pretty good job, right? So when you leave it on max conversions or you do some kind of ROAS targeting. If you’re in eCommerce or cost per lead, it does well if you can get there.

But the problem is we work with a lot of small to medium-sized businesses, and it seems like the bigger advertisers, we don’t have a huge problem there. We work on other things like ad copy and improving landing pages and things like that, where Google does all the heavy lifting, but the smaller advertisers.

It’s a ton of work and it seems like it’s just a struggle just to get to that point of statistical significance and trying to get purchases or leads is all over the place. One keyword that worked for you last week, isn’t the same keyword that works next week. So we live in a day and age where I think signals are more important, even than keywords.

And Google might even back that up there. They’re leaning more towards targeting a lot of broad keywords. If you’ve seen out in the ether where people get lazy, I’ll explain this. So I went to a conference in Australia for YouTube. And one thing that kind of stuck with me is people go to YouTube by typing you.

Just you right and you’ll notice any site you go to you might type the first three letters and just hope Google gets it I almost don’t want to say we’re lazy, but we’re more efficient these days so what ends up happening when you have something like this is things get a little bit more difficult to guess.

Okay, so people may not have the same signals they had before where they say hey, I want to purchase this specific thing. And their keyword is, 20 years long. It’s one keyword and you have to guess the intent of that. And you can’t guess intent on the keyword by itself. If it’s just one word, you have to guess by signals like in market signals.

Could be income, the sites you browse, contextual places that there are probably a million signals that Google uses to find if someone’s actually in the market, you get the right person at the right time rather than just solely relying on the keyword these days. That’s one of the challenges we’ve found.

Frederick Vallaeys: Yeah, and that’s interesting to restate what you’re saying there is that historically, the search engines just had less of a memory around who a person was. And so it was much more important to have a fully qualified keyword that exactly explained what you were looking for. Because that was the predominant signal to fetch the correct and the most relevant ads.

But nowadays you’re saying Google knows so much about your generative AI, you’re having conversations with these assistants. That’s serving like your memory. So now you can type something as simple as three words, the three letters of the keyword that you were intending to type.

And the system already knows what you want. And so it’s able to pull in the right ads. So you can no longer rely just on that exact keyword. It’s about all of these auxiliary signals which also maybe speaks a little bit to like upper funnel activity, right? If you’re not present in those upper funnel stages where the memory and the intent are building, then it’s going to be much more difficult for Google to say oh, you are the advertiser.

At that bottom of the funnel, who is the right match for that thing that the user has been thinking about for two or three days and has gone on a complex consumer journey?

Shawn Walker: Yeah. And we find that like back in the day, you would just get these long tail keywords and they would work. But now those don’t seem to work as well as other keywords.

We’re talking one to two keywords versus, three to five, let’s say it is backward. We’re living in a backward world where all these signals are more important than the keyword itself. As to why that’s the case, I could only guess. I think it comes down to we’re trying to be efficient.

We expect things right away. Even when you’re looking for a movie, let’s say on Netflix, you type in two letters and you hope the movie gets listed to the right. You’re not going to type the whole thing out. We’re just moving that way. I think there was a big conversation before about voice search where I just don’t hear that much rumble about that anymore.

It used to be all about long tail keywords and now Google has switched to broad. So there’s something just changing out there.

Frederick Vallaeys: And I think when it comes to voice search, the problem that I always thought was, that consumers want variety and options. The problem with a voice search is that if you ask it, whatever it is, it’s really difficult to give it three to five options like you would see on the search results page, where you can visually very quickly take that in and go to the one you want.

For voice search, you have to wait for the first one to be read. To even be exposed to what that second one in the third one is. And for people, it’s all about shortcuts, right? People don’t want to spend more time doing things. They want to spend less time doing things. But I think generative AI is changing that game.

Now for voice search, we no longer have to think about keywords for results. It’s a conversation I can explain to the assistant what it is that I want. And it doesn’t have to give me three options. It can say, okay based on what you’ve explained, here’s a summarization of the options that you have.

And as you start to understand what it is that you want, it leads you down to that one answer. And so I think voice search and voice-based advertising or chat-based advertising, obviously is going to come back in a, in full force, just because we have different and better technology that we just didn’t have in the past.

But like you’re saying, that means it’s no longer about one keyword to a bunch of ads. It’s about, many conversations that consisted of thousands of words of actually talking to your bot. And that’s the thing that creates the signal that Google or ChatGPT, or whoever can use, to say, here’s the ad.

And then the other thing that’s fascinating about that is there’s some notion that your assistant or Google knows before you’ve even started typing anything, what it is that you’re likely to do. And you think about your phone, right? You open the app to find the search screen on your phone and it already has the 10 most likely things you’re about to do right now because it looks at the context of your day, your location, and what you typically do.

Google is very similar. In that regard. And so maybe rather than them just showing us some stuff that they already knew we were going to search for, they were like, yeah, let’s wait let’s have them five and three characters. So to at least feel some sense of control. But it is pretty crazy what the technology can do these days.

Shawn Walker: Yeah. And I’ll say one thing, I feel like I’m trying to be a little bit of a hero for the small guy here. So let’s say five years ago, even this is all great. But now, Google has taken away from us, unfortunately, targeting these long-tail keywords. So we write in these long prompts, right? Or you’ll go to Reddit and type in this long thing.

So the searches do exist, but the problem is if you try to buy these search terms, Google will say too low volume, right? And then you can’t use it. You can’t buy it. You just won’t show you can buy the keyword, but you’re just not going to show for it. And then also, unfortunately, in your search query report, you don’t get to see all the terms.

I forget the percentage, but it’s a very low percentage of what you get to see. So if that came back, and if Google’s listening, bring it back. But if that came back with all this AI prompting, that would be a great marriage, but unfortunately, the little guy is stuck, right?

You can’t buy those terms that you know are going to work or are very pointy. You have to go broad and you have to spend those dollars because you’re just not going to, you’re not going to show up. Google will flag you and say that’s been going on for years and years, and that’s where we struggle with some clients, I will say, because a lot of people live in the past and they have terms in their head that they know people search and they say, let’s just buy these and then, we don’t even show for them.

That’s been a big struggle for us with small businesses. And especially with singles, just going back to bidding the problem is small businesses have a higher cost to operate per unit, right? If you’re a big business and you bought a big building and you fill it up with employees, you’re at economies of scale where to produce one unit, it’s going to be a lot cheaper.

But if you’re a startup, that unit that you’re selling is going to be so expensive. So your targets are harder to hit. You might be able to operate at a two-to-one as a huge business, right? Getting rid of operational costs but as a small business, we see often targets where people say, Oh, we got to hit a five to one or seven to one, which is that much harder because you’re competing with these big people, right?

We all have the same car these days with AIs and algorithms, but the challenge is how we drive it differently. If you’re a small business, do I drive it differently than the big business? That’s a struggle and I’m sure if any small business ends up watching this, probably the best piece of advice would be things don’t work how they used to.

Frederick Vallaeys: So you’re supposed to simplify. One of the key things that I keep hearing there is instead of having thousands of campaigns and tens of thousands of ad groups like these complex structures that go after the long tail, it’s much more about simplifying it and campaign structure.

Just have a handful of campaigns and the campaign is focused on what is your budget or what’s the geo-target. Or what’s the specific type of conversion you’re trying to get for that campaign? Then you match that up with some of the automation from Google, like the broad match keyword, responsive search ads, and automated bidding.

You put those three together. Presumably, something good can happen, but that good thing can only happen if you give it the key signal. And this is what you were talking about, right? The key signal. To make all this automation work for you, whether you’re a big guy or a small advertiser is that conversion data.

And so ideally you have at least 30 conversions per 30 days per campaign. But tell us what do you do if you’re one of these smaller advertisers and you can’t get through those 30? Are there some other benchmarks, some other signal that you can push into Google to get things maybe rolling in the right direction?

Shawn Walker: Yeah. So this is probably the most important part of the conversation, I would say. Most small businesses are not going to hit that. So let’s think of a scenario if you say you have the $ 2,000 budget, right? And your CPCs are 5, which isn’t crazy. That’s normal. And you have an e-commerce product where you have a conversion rate of 10%.

I think you get about 40 purchases. Okay. Would you say, great, that’s 40 signals to my conversions. But if that’s spread across 10 campaigns even if that’s spread across two campaigns, you have not hit the threshold that Google needs. And for some small businesses, that’s an impossible feat to get to.

So the question is, what the heck can I do to beat this? There’s a couple of things. What’s the next thing above a purchase or let’s talk B2B and eCommerce at the same time? So above a purchase would be added to the cart. Or for B2B lead generation, let’s say, if you can’t get somebody to fill out a form what’s something else that you can get that’s right before that?

I see the order of operation going like this. You see an ad, that’s your first signal. Your second signal is you click the ad. Then you go to the website. If you make it there, because some people drop off. And then when you’re on the website, you do something. You engage, right? You either scroll down the page, you read some content.

You go to another page past that, it’s more engaging content. It might be a gated asset or even an ungated asset where you open up a PDF or something like that. And you see where I’m going with this. So there are all these conversions that you could build out from the point of someone clicking your ad, visiting your website all the way to purchase and lead.

And I’ll just give some examples of things that we’ve built out that have worked for us with small businesses. So one thing that we use is this conversion called engage visitor. And it’s pretty interesting. It does two things. So you go to the website and if you scroll down about 50 percent of the page, that would be a normal conversion that you could launch.

However, if you come back a hundred times, let’s say it’s going to fire a hundred convergence and you might not want that, but we built a cookie-based conversion that only fires per unique user. So if you go to the site, you scroll down 50 percent of the way, a cookie will fire and I’ll say, all right, Fred’s been to this site.

He’s engaged with the site. He’s not a bot. He’s a higher-quality user, right? And if you come back a hundred more times, it’s still just going to count you as one. So we’ve been using this signal for a lot of B2B. We’ve been using it for a lot of e-commerce- just to have something that’s going to get rid of the bots.

And to get people who aren’t going to go to the site and be idle, so that’s been pretty successful. And I think we have it across most of our clients now. It takes a little bit more scripting. I will say you can get it done in Google Tag Manager where you have to have some kind of script that can set a cookie once you scroll down the page and then also read the cookie.

And say, okay, Fred’s been here before. He’s already been tagged. Don’t fire this conversion again. If someone wants to know more about that, reach out to me on LinkedIn. I could probably give my link at the end of this, but that’s one good thing. The other thing I will say is product views for e-commerce.

That’s a big one. If someone is looking at a product. And I don’t mean you landed on the product page. You went to the website, you ended up going and looking at a product. Product remarketing is a good signal, right? For B2B, I will say, any kind of content clicks or outbound emails, where you’re not necessarily giving a lead, but you’re showing intent to contact anything a little bit softer.

So I guess the summation of what I’m saying is to find a soft conversion where you’re going to get that volume that can be fed into the system and give you a good indication that this person is much more likely to convert to something else like a lead or a purchase versus someone who just went to the website and bounced right away, maybe even spent two seconds on the site, or it’s a bot because we have a bot problem these days.

So every advertiser I would recommend if you’re small, you have to set up these soft conversions. And don’t assume that you’re just going to get a flood of purchases or a flood of leads because we’re just not seeing that these days, especially because you need so much budget to get that statistical significance.

Frederick Vallaeys: So think about it as a hierarchy of what is it that the typical consumer goes and does as they go to the ultimate conversion, which would be the sale, and both in eCommerce and legion, I think the ultimate conversion is the sale is closing the deal. But then the beauty is that if you think about that funnel, there are different stages, right?

So what is your approach for not necessarily saying, okay, I have to pick this third stage in the funnel of desirability? But do you create conversion actions for each of those steps and then toggle those on and off as you start seeing there’s enough volume in that more desirable action? So let’s get more specific.

Let’s feed more specific data to the ad platform. So it can bid for the thing. That’s a little bit closer to what I want, or how do you manage that whole like dealing with shifts in volume at the different conversion action levels?

Shawn Walker: So we have a specific approach. If it’s a big client and we have budgets, what we’ll do is the same thing as we do for a small client.

We’ll still build out those soft conversions and the hard conversions, I will say, which are purchases and leads. And we’ll start with the thing that’s more difficult to get, and we will at least try max conversions. We’ll try the cost per if we have a goal, right? And if we don’t get there, we have our life raft.

We have our backup plan, right? So we do build out actions for all these before we even start and some campaigns might get there quicker. If you’re just using a setting per campaign that just goes with the full account, I almost don’t want to recommend that because each campaign might be different.

One campaign might be doing better than the other, whereas the other one never gets that statistical significance. So what you can do is just pull in that softer action, right? And when it gets to the point where you can get a harder lead, you can switch it.

Frederick Vallaeys: Pause on that for a second.

So you said something along the lines of Account level and campaign level. So I think you’re talking about the different levels of which conversion action can exist. Can you go a little bit deeper on explaining that for folks who may not have necessarily used that capability?

Shawn Walker: Yeah. So by default, when you set up conversions, Google will put this as a primary target, right? You have to manually go in there and set it to secondary. The difference between the two is your primary is going to be included in your optimization settings, whereas if you just start a campaign and choose account level, it’s going to pull in all your primary assignments by default.

But you can also use the secondary ones if you want. So you can pick and choose what you want. But by default, you’re going to use the account level. However, if you use the campaign level, which you’re allowed to, you can choose whatever you want to optimize, because I think every campaign is going to be treated a bit differently.

Especially let’s say you’re a big brand. And you’re running non-brand and brand. Maybe your brand is doing much better than your nonbrand. You might have different strategies. You might have different conversion points per those. Using that account level may not be the right approach. It might be better to use the campaign level.

Frederick Vallaeys: Okay. So you have these different levels of conversion actions. You can make them primary or secondary and basically, you can make them primary as there’s enough volume to say this is the signal that you now need to use Google for my bid optimization. And maybe make some of the old primary secondary.

So they’re no longer considered because they might not be as desirable. Maybe they are upper funnel micro conversions. So that sounds like one approach, right? It uses micro conversions or softer conversions to guide the system. There’s another approach that I think you’re advocating, which is to do with account structures.

So talk a little about how you think about the number of campaigns you have and maybe as opposed to making more campaigns and building more ad groups, how should someone think about that?

Shawn Walker: So I came from the world of working in automotive for search. So I thought every account was going to be like this, right?

Oh, you get 10 million and you can build whatever the heck you want, not the world we live in. And especially not in the world we live in with small and medium advertisers. Using Adobe or Google, you can get some idea of what your conversion rate is for something, whether it’s direct or organic. You start somewhere and you get an idea of this is my conversion rate typically advertising is not going to be as good as someone who just goes directly to your website, but you can get some idea of where you can reverse engineer and say:

“Okay. I’m going to need this many clicks or visitors to my website to get a lead.”

Then you get back into your budget, right? And if you find out that you know what it’s going to take me a hundred thousand dollars to get statistical significance. In that scenario, it’s not a great idea to break out into a ton of campaigns, right?

Google is going to look at each campaign and say, are you reaching that statistical significance? Now, if you’re in a case where you can reach that and do so with 10 different campaigns it’s advantageous for you to split out themes and you can still reach that statistical significance.

That’s fine. But I would say if you had to choose one, the priority is to have a structure that is in line with getting enough conversions. Every account that we do now starts simple and we make it more robust over time versus back in the day. We used to build out these giant campaigns. This huge account structure and then dial it back where now you want to do the opposite.

Start small and then build as you see where there are areas you can build on.

Frederick Vallaeys: And that’s fascinating. So you start small and then you expand from there. There’s a different approach to achieving the same thing, which is more narrow targeting, right? So I think most advertisers, they come in and they say, hey, we’re ready for a lot of business, send it my way.

But is it sometimes better to take a more. Narrow approach and maybe do some geo-targeting and do some exclusions and build from there. Or do you tend to not put too many restrictions in place from the get-go?

Shawn Walker: Yeah. So we do want to put restrictions in. Say you’ve got to the point where you even have one campaign and you said, this is as much as I can do.

I only get a handful of conversions. I only spend 500 a month. Let’s just say in that case, you need to decide on feeding signals that you know beforehand. So Google advocates that let us do the work, right? Let us spend a bunch of money, we’ll find your people, and we’ll just keep going.

However, a lot of people don’t have the budget to do that to start. So what you have to do is use whatever data you have, your marketing data, consumer surveys talking to people, email. The power of communication is big and you find out where people are. So we’ve seen a couple of things, a different couple of signals.

Let’s talk about geo real quick. There’s city and rural, right? If you have a business that’s going to thrive in the city, do city targeting. If you’re going to have a business that’s going to thrive rural, then target that, right? Different states perform differently. I will say this- a lot of people put things on observation.

But if you know that, let’s say Texas versus New York, let’s say Texas as far as your clientele goes, has double the conversion rate as anywhere else. Start there. If your budget can handle it, just start there, pick the low-hanging fruit as people say, because you know that there are more converters there and you can put that on explicit targeting rather than observation where you’re still targeting everyone and you’re learning.

If you already know Texas does well, target Texas. I will say this as well. There are other signals, like if you’re selling a high-end brand, let’s say where it’s the thousand dollars to purchase something or more, you might want to just explicitly target the top 10 percent of income earners, which is a signal in Google rather than targeting everyone. Or what you can do, because what we find is the unknown when Google doesn’t know what income bracket you’re in that tends to be a big one that does well, but you can do exclusion.

So say, that the lower 50 percent income is never going to buy your product. Just start off excluding that don’t guess if you already know. Just start there and then other things like I’ll just give an example of one client. We have a client who works In sports, right? So baseball, volleyball things like that.

You can choose in market targeting explicitly where you can find people who are just interested in something and maybe what happens is you grow, you start by just targeting, let’s say, baseball, right? People are interested in baseball or baseball equipment. And you find that you start scaling, you could then put that on observation and go after other people.

But in the beginning, I think the best way to approach your budget is to do the best targeting you can in the beginning and see how far you can go with the low-hanging fruit and then expand from there. But at least you know how well you can do right from the get-go versus the other way around of guessing and spending a lot of money to guess.

Frederick Vallaeys: Yeah, and this is interesting because you’re, you have competing objectives, right? On the one hand, you’re saying you need to have around 30 conversions per 30 days. On the other hand, you want to target the places that perform better, such as in your example, Texas having twice the conversion rate, or maybe you want to be in the market for a certain audience.

But that reduces the number of conversions that you get. But so I think once you put those two together, what it sounds like you’re saying is target specifically where you will perform best and then fix the conversion issue as a second step by going maybe higher up in the funnel and not looking at the sale, but looking at a soft conversion event that proceeds the actual thing that you want. It sounds like that is the better strategy as opposed to going broad and getting lots of volume.

But maybe the wrong kind of volume, right?

Shawn Walker: And go broad where Google says you need to go broad. So you might have to do broad keyword targeting because we’ve seen that it doesn’t work. Otherwise, if you get too tight, go broad there and tighten up your targeting. Like you’re saying your geo tightens up your in market segments, and things like that.

And then with that same approach, start with your hard conversions. You’ll see in the first week, right? If that’s not working, then all of a sudden you have to bring in your soft conversions and feed it some kind of data to where it can go after more quality users. And you could eventually get to your sought-after goal, which is, sales.

And I’ll mention one thing with B2B. What I find interesting is. There are all these systems like Salesforce and HubSpot. We can have marketing qualified leads and opportunities, and we always connect these systems, but I’ll say even for large clients, there’s never enough volume on these to optimize to those there.

They seem to be nice to have in the system to say, Oh, I eventually got to this. But the problem is you just never get that statistical significance, right?

Frederick Vallaeys: Then what you’re describing is like you have the funnel and we’ve talked about like the lead is what you want, but all these things come before it.

When you go into the CRM, it’s even beyond the lead. Deeper down the funnel of higher levels of qualification. So the volume of data becomes even more sparse- a nice problem to have. If you spend a million dollars a month on ads, you probably do have enough CRM data to feed the system with those MQLs and SQLs.

But if you’re a smaller advertiser then CRM sounds like a total luxury and not something that’s immediately going to give you some gains.

Shawn Walker: Yeah. The problem is you’re trying to predict norms, right? If you’re a smaller client, let’s say you have five sales in Salesforce, which might sound like a small number, but it’s not in the B2B world for a smaller client.

And you might see that one person is in, we’ll just go back to my example, New York, and one person’s in Texas and they both like sports. You cannot use a sample size of five to say, all right, we’re just going to explicitly target sports and we’re just going to target people in New York and Texas. This is where I have a problem with some of the data that people lean on CRM data to make decisions.

I think you always have to go back to that statistics class and say, do I have enough samples to predict norms, right? Is this going to be the norm or is somebody going to squeak in there from Florida, or Washington and then throw your plan out the window?

Frederick Vallaeys: Yeah, and think about the agency perspective.

You have to teach your clients to have these fair expectations because if they’re going to be basing their judgment on five conversions in a month, guess what next month you only got four conversions, which was a crapshoot, right? That could have happened either way. That’s outside of the statistical range.

But they’re going to look at that and be like, oh my God, Shawn, why did you drop 20 percent of conversions? It’s yeah next month we’ll have 20 percent more because we’ll randomly have 6 instead of 5. But that’s not an actual statistical pattern that we can do anything with. So you’re setting up, I think you’re setting yourself up for a lot of hard conversations if you allow your customer to look at those small numbers as the absolute truth.

Shawn Walker: Yeah. They’re nice to have, but I wouldn’t make any big decisions based on that.

Frederick Vallaeys: Yeah. At some point something happens, and automated smart bidding may not work, whether it’s due to data being broken, or maybe the volume just isn’t there, no matter what you do. What’s a backup?

Is there some level of automation you can still put in place that you control based on the type of signal you have?

Shawn Walker: Yeah. So a lot of the time if we’re starting a brand new campaign where there’s not much background and the client doesn’t know, exactly who the audience is, we will start with manual bidding, right?

Just to get some volume out there. But you can still make intelligent decisions. So outside of adjusting ad copying keywords and negative keywords, the obvious stuff, if someone, if you’re not handing out discounts and people are typing in your brand name and discounts, you probably want to negative out discounts, things like that.

There’s the obvious stuff, but then there’s how can I optimize when I have limited conversion volume, I think that is what we’re saying- something is better than nothing. So while you can’t make big decisions off of what you’re doing, you can do a couple of things. We built out some scripts that do some of the lifting that Google would do in their automation.

I will say this, we don’t have big decimators, right? Like they don’t go in and just start pausing everything. It’s very slight changes. So we have a couple of things like device automation. So even if you have five conversions. What you can do is just have a level setting where you can control the script, where you bid up, where you see performance and bid down, where you don’t based on conversion rate with limited signals, right?

But you do it in a way where you’re not changing bids by 50 or 60 percent and you’re not decimating the campaign. You’re doing something better than nothing. So even if it’s a 5 percent change. And you’re testing a theory you’re doing. So you’re doing what you would normally do as a human, but you’re letting a robot do it.

So this is pure automation. In market segments, we have a script that looks at segments and does the same thing it would for devices. Same thing for geos. I would say if clients don’t know exactly where their customer base is, I think it’s good to start by just uploading 50 states on observation.

So you can make changes to them, and apply bid multipliers. So we have a script that will make bid multipliers work. There’s also a campaign script that we have where if we see a campaign doing a human would go in there and up the budget. And decrease the budget for the other campaign.

We have robots that’ll do that. So you’re seeing a theme here. It’s anything we would do as a human, we’re letting a robot do it. And then there are also a couple of good ones we have here as a schedule creator. So you hit a button, it creates your Monday through Friday, sorry, Monday through Sunday, where.

Not only do you get to automatically create a schedule, but you get to add bid multipliers to that as well. And just the last couple of things, if you’re limited by budget from your keywords, there’s one quick fix there. So the quick fix is if you have the budget and a campaign is doing well, then put more budget into it.

But if you don’t, and you want to get rid of that message, what you can do is just decrease the bids of keywords within that campaign because essentially what you’re doing is you’re paying more than you should for your keyword and it’s stopping a partial way of the day.

So Google saying, you’re limited by your budget, we’re only going to show you for a portion of the day. You can get more clicks by just decreasing your bid. And we have a script where it will decrease bids for keywords that are not converting until that message goes away. I will say this is not perfect, but it’s better than nothing, right?

The fear is in some of these if you think a keyword is going to do well, for some reason it’s done well in the past. Could this bid it down because it hasn’t had a conversion yet? Yes. So there is a caveat there. If you turn all these on and 10 things are bidding down or up, could things get expensive?

Yes. So you have to keep an eye on it. It’s not robots taking over the world. It’s more you need to be there. You need to wield the hammer, make sure that it’s working correctly, and keep an eye on it.

Frederick Vallaeys: And instead of spending manual time doing these adjustments that you would have done anyway, get back your time by scripting it. Or you could use Optmyzr and the Rule Engine can do that for you without coding. But use these tools and technologies so that you can take back your time and have more of a strategic thought about what else should happen in that campaign.

That’s the human value at this point, I think. So it sounds like you’re leveraging a lot of cool scripts too. Give you back time now, as, as far as taking back time you alluded to some gen AI at the beginning of the show, but any cool tools that you’re using there beyond scripts, or are you even building scripts with GPT?

What are you using gen AI for that saves time in an agency setting?

Shawn Walker: Yeah. I will say that the coding has helped if anyone can stomach it, and sit there and stare at a screen for a while. Scripting does help just being able to read it, and being able to write it because then you could create these things, these tools, almost like widgets, right?

Leave the enterprise stuff to you, Fred, the little things being scrappy I think it has helped so much. And as kind of part of that, we use writer.me. Have you heard of that before?

Frederick Vallaeys: I don’t think I know that one.

Shawn Walker: So it’s like a version of Chat GPT, but I will say one thing when we were vetting a couple of things, this one is super affordable and it does a really good job with search ad copy.

We tried Jasper ChatGPT and a couple of other things. And we found that this one was just on the money with the ad copy. I was wowed by it. We ended up using it. This helps not only write ad copy as a start. So let’s say you have 20 ad groups you need to write for. This is a good start for it, where you can go in there and then tweak it to see if it makes sense.

Starting keyword development, I know we all end up in the Google ads tool, but, I always have it open. I’ll just put in, the brand name and I’ll spit out 10 keywords that I can copy. And then put in Google ads. And then all of a sudden it’ll give me just more robust keyword examples.

Frederick Vallaeys: One thing I like about these generative systems, like the keyword generation and the attacks generation is that you can start to prompt by putting yourselves in the shoes of, and then you put in your target audience and the type of keywords and the types of messaging that it comes up with in the ad is it’s really specific to that scenario.

So I was recently testing this out for not translating, but localizing an ad. So taking an English ad and instead of straight up translating it to French, I said, could you rewrite this ad so that the French user would find it more appealing? And so it starts making word selections that keep the core message the same, but the nuance of the UVPs that it’s putting in are highly targeted to a French audience that cares about the elegance of something as opposed to the freshness of something, which the Dutch care much more about. And so the types of ads that come out fit those audiences. And that’s the brilliant thing that GPT can do. There are so many of the old-school tools that just weren’t able to.

Shawn Walker: Yeah, I like that.

And I think I think overall, I don’t want to speak for everyone, but I think the right way to use these things is just to speed up your workflow. I don’t think it’s supposed to replace us, eventually, you get there, right? You eventually get to write your ad copy, but instead of staring at the screen and doing a ton of research, let these tools do this for you, and fact check, of course. But I think it’s supposed to speed up workflow.

So that’s why I use these things. A couple of simple things too. I have it open just all day long, just to even answer simple questions. This is weird because I used to do that in Google, but I can’t even say why I think I have this open so much now that I’m finding myself prompting more though, more than when I go into Google just because I do a lot of things there scheduling, I gotta tell you a cool one that we did to one thing I was finding myself doing every single day.

People go, are you available? Send me your schedule or whatever. You can’t always go in and share your calendar with every single person. You might be talking with reps and clients and different things, and you have to go into your calendar. Look, go in there, type it out. And this was just becoming too cumbersome.

So with some scripts in Google Sheets, I wrote the thing where you connect your calendar. And now I just put in the start and end date of what date I’m interested in and I hit a button and it’ll just spit out the days that I’m not available and the times and I copy this format and it takes two seconds.

I copy this, I throw it into writer.ai, and it spits out an email that’s like I wrote it, and it says I’m not available at these times. And I got to tell you, this has saved me, I don’t know, maybe two hours a week because I was doing this all day long and I’m like, surely there’s an answer for this.

And I went on Reddit and all these things that nobody had anything like this. So I just hacked it. And I got to say, it’s one of those dumb things, like that’s such a little thing, but it added up so much time that I was like, you know what we got to do something here.

Frederick Vallaeys: I think a lot of people might pay you three, 5 a month just to have that little capability.

Maybe that’s a business right there.

Shawn Walker: I know I would pay myself. It’s helped.

Frederick Vallaeys: Very cool. Awesome. So yeah, tons of good advice there as far as how to make bid management work, especially if you’re a smaller advertiser. So thanks for all of those steps. Any final takeaways there that you think people should know?

Shawn Walker: I think there’s one thing I missed. I know you’re saying, how do you structure, campaigns and ad groups? I will say, that I think we live in the day where it’s better to separate ad groups than it is to separate campaigns because all the optimization is at the campaign level. So if you’re sitting there making decisions, I want to break out my themes.

And as long as you don’t need to control a specific budget, just break it out in the ad group rather than breaking it out at the campaign level. I think you should only really break out campaigns if it’s truly performing completely differently than other campaigns. If it’s brand versus nonbrand, for example, if you have to have a dedicated budget for it.

And if the targeting is different because you can’t target different locations and in a different ad group, let’s say so certain parameters you have to use campaigns for, but if you can avoid it, just do it at the ad group level and let Google use that statistical significance that will come with the reduction of campaigns.

Frederick Vallaeys: Yeah. Understand why these structures exist and what their capabilities are. And if you need to divide based on one of those specific capabilities, then it usually makes sense to build a new campaign. But if you’re going to have multiple campaigns with the same budget, the same geo-target, the same day parting, then, like you’re saying, you’re maybe just making more complications for yourself than is necessary.

You’re also going to confuse the machine and give it less data. Don’t use ad groups instead.

Shawn Walker: Yeah, and I’d say, just to recap, if I was going to say three things, be prepared for little conversion volume, right? And I know we don’t like to hear that but be prepared for it. Build those soft conversions from the start right away out of the gate and start with a smaller structure.

If there are three things to remember, I would say out of all this, those are the three things I would give away.

Frederick Vallaeys: Okay. That’s awesome. Thank you, Shawn. If people want to get in touch with you or Symphonic, where should they go?

Shawn Walker: So symphonicdigital.com. I know there should be an About Us section where you have a smiley face of me where you can get a hold of me on LinkedIn.

Or I don’t know if I can give you a link.

Frederick Vallaeys: So we’ll put all the links in the show notes, but reach out to Symphonic Digital. Shawn is also, like everyone, on LinkedIn—a good place to get a hold of him. If you’ve enjoyed this episode and want to see more of them, please subscribe and leave us a review as well.

And we’ll be back with the next episode of PPC town hall soon. Shawn, thank you so much for joining and everyone. Thanks for listening and watching.