Transcription
[Music] Hello and welcome back to another episode of the Collective Clicks Podcast. This is your host, Brandon Baitman, and today I'm joined by Garrett Craigan, our lead of paid media here at Baitman Collective. He's one of the best PPC masters I know. We're going to talk all about keywords and negative keywords—just the first part of eight essential elements for a successful PPC campaign. We'll cover details you don't normally hear, and some common investor mistakes that, if fixed, would dramatically improve PPC results.
How you doing today, Garrett?
Doing good, how are you?
Hey, doing excellent, thank you. I know we planned some banter, but we're completely awkward as people, and I think that made this impossible.
Yes, we suck at being people.
Yeah, that's right. So we're good digital marketing robots, and I guess we'll just jump right into the stuff today. Sorry for everybody listening, we don't have anything fun to say.
The topic for today is the first in a series. We're going over eight essentials for a successful PPC campaign—or conversely, eight things we see investors doing wrong. It's in-depth, definitely one for nerdy investors wanting to dive deeper into PPC. We'll discuss technical terms, but if you're familiar with the platform, you'll find this interesting. It's eight things, so we'll start with numbers one and two today: keywords and negative keywords, and ads.
To kick things off, we do a lot of audits. An investor comes to us saying, "I don't know what's going wrong. My PPC isn't working," whether self-managed or through an agency. We look at the account, see what's happening, and why they aren't getting results. A common issue is the interaction between negative keywords and keywords. Garrett, any thoughts on common mistakes you're seeing?
Yeah, I think there's a tendency to focus on the "sexy" parts of PPC—dynamic localization or different ad types. But at its core, PPC is bidding on search terms. Good keywords and search terms usually lead to good results. If that's not there, everything else underperforms. It's crucial your keywords target the right intent, and you aren't bidding on out-of-market or low-quality searches.
You basically just said what I've said in so many audits recently, but with different words. It's like, "I'd love to talk about your problems, but you're not even eligible for most of them because you don't have the right search terms!" That's step one. People think fixing their campaigns will be great; it's like going from an F to a C. Good searches are the bare minimum; the rest won't fix a bad foundation.
I think the root of this for many companies is the difference between keywords and search terms. Many don't understand this. Could you explain?
Yep. The search term is what the customer actually searches—e.g., "How can I sell my house fast for cash?" A keyword is what we tell Google is the rough intent we want to target. We can target a keyword and have the search term not match, but most of the time it does. Most of the time we're targeting a kind of search, and Google matches it as best it can. That's why reviewing search terms to see what you're paying for is important. You can have good keywords but still not have good search terms. That's where we get stuck if we aren't diligent in keeping those search terms pruned and cleaned up.
That makes sense. An example: a keyword is "We Buy Houses," but the search term is "buy houses." You're telling Google you want "we buy houses," but you're getting clicks and leads when people search "buy houses." "We buy houses" is seller traffic; "buy houses" is buyer traffic—very different.
I also think many don't realize how many search terms there are. It's insane! I did an analysis (a while ago, so I'd be curious to repeat it) across our clients. We found that over 80% of their leads came from search terms with only one click in their account's history. You think it'll be five keywords getting all your leads, but there are thousands, and most leads come from long-tail search terms, not frequently searched ones.
Interesting. I saw a stat that a thousand brand new search terms appear on Google daily. What was historically searched isn't static; they're always changing and evolving. It's important to update your keywords based on what's currently working. Those search terms will be different every day, and if they aren't pruned, you're wasting budget.
The stat I saw is that about 15% of daily Google searches are completely unique. That rounds to about 500 million searches daily. That's a better stat than mine!
I wonder if you were looking at a specific account or niche, because 500 million and a thousand are very different numbers. But it's insane! I think this is marketing stereotype 101. Marketers like to put people in boxes, saying, "This is my customer." I saw a funny LinkedIn post about 2023 marketing predictions, sarcastically saying, "New CEOs and boardrooms will create oddly specific customer avatars, yielding no benefit." We do that all the time—expecting people to fit into these boxes. But people are unique, and how they search is always different.
One interesting thing: how many keywords do you really need? I've talked to investors who think the name of the game is guessing all these search terms and making a keyword for each. They think, "I'm working with a professional PPC company; they have tens of thousands of keywords and know everything people search." What would you say about that?
I don't think there's an ideal keyword count to maximize results. It depends on volume, intent, and budget. You can have a thousand keywords, but with a $1,000 monthly budget, you probably won't show up for most. You won't get enough data on a given keyword to have meaning. It has to be based on your budget, audience, keyword volume, and monthly searches. All these must be considered when building a keyword structure.
Understandable. I'm curious, what's the largest number of keywords you've managed or seen?
I managed ads for a 500-company, and they had 8,000 keywords in their North America account. They had an account for each continent/market due to different currencies.
Understandable. That makes sense. Interesting. I asked because I have a ridiculous story. One client had six million keywords!
Oh my God!
Yeah. How many had clicks/impressions? A smaller number, but they had the biggest search term universe of any company I'd ever worked with. Six million keywords! They spent about five million a month, but that's a dollar per keyword. Insane! Anyway, that was tangential… cut those down, team!
I think there's been a shift in PPC strategy regarding keywords. The account I mentioned has been going since 2008; the world has changed. Most PPC marketers agree things have changed significantly in the past decade, even the past three or four years. What are some big changes?
A big change is the platform itself—match types. Google has become more liberal in how it matches your keyword with search terms. They say it's to match intent, not just the keyword. But I think they want to sell inventory on searches not bid on by advertisers. What was an exact match keyword is now much looser than a couple of years ago.
To clarify, if your keyword is "We Buy Houses," there are different match types: exact match (historically, only "we buy houses" would show); phrase match ("we buy houses" has to be in there somewhere); and broad match (if someone types "We Buy Ugly Houses," you still show up). If you want to learn how to completely evade an issue, read Google's help articles on match types these days—it's hilarious! They basically say, "If you do this match type, it means all of these things, and also whatever we want."
Yes, exactly! So wild! It used to be, you put in "apple," maybe you get "apple," but you want "apples." They changed close variants, so now you put in "apples" and get "orange"—Google says they're both fruit, so close enough. It's different, especially with nuances like "We Buy Houses" vs. "buy houses." How has keyword strategy adapted?
There was a popular keyword structure called SKAG (single keyword ad group). People built one ad group with an exact match type keyword, forcing each search term into one ad group. It matched search with keyword, ad with landing page. But since match types have changed, this is obsolete. It's harder to match one search with one keyword because they're so loose. It's become increasingly important to have a fine-tuned negative keyword list. Let's delve deeper, because they function differently than keywords.
There are positive and negative keywords. Positive tells Google what you want; negative tells Google what you don't want. Negative trumps positive. I could have the keyword "we buy houses" (phrase match), but a negative keyword of "scam." If someone searches "We Buy Houses companies that are scams," I theoretically wouldn't show up. Without the negative keyword, we would.
Have you ever played Whac-A-Mole?
Actually, I don't think I have, but I know the game.
Okay. Adding negative keywords is like Whac-A-Mole—new bad search terms constantly pop up. Just as there are those 15% of daily Google searches that are new and beneficial, there are negative ones. I've seen people play this game for years and never get anywhere because they have negative keywords they need. How do you protect against this?
I've audited accounts years into this, and half their search terms are junk. They're whacking moles, but they pop up as fast as they can whack them.
There are a couple of things you can do. First, go through those search terms frequently—daily, in the beginning. If you have great search term quality, you can relax a bit later. The other thing is the match type you add those negative keywords to. Google is liberal with positive keyword match types, but restrictive with negative keyword match types. It requires specific negative keyword content to exclude it. A common reason for the Whac-A-Mole game is not having a wide enough range of negative keywords. Google finds loopholes around what you told it was negative.
I saw an example: hundreds of search terms related to home buying grants. Each time one popped up, they added it as an exact match negative keyword. The next day, a misspelling pops up, then something without one of those words—a constant game! You just need to add "grant" as a negative keyword. Anything including "grant" won't be there. It's simple, but I've seen accounts destroyed by negative keywords. You have to be careful it doesn't exclude a good search term.
Absolutely! For example, "buy houses." You might see "buy houses," "buy houses in Orlando," etc. You exclude "buy houses," but then exclude "we buy houses." Google never tells you what you accidentally excluded. You restrict your volume forever without knowing it.
One way to navigate this is leveraging a shared list of negative keywords across all our clients. Each time a client reaches a unique negative keyword, all clients benefit. It's shared learning built over years. We've already done the Whac-A-Mole for you.
So, a little plug—you mean to tell me that if a bad search term has shown up in any account over millions in spend, across hundreds of accounts over five years, it would be excluded?
I do mean that.
Wow, what a brilliant strategy! It makes a huge difference to have an extensive negative keyword list. Every bad search term click costs money. If that click becomes a lead, Google gets positive feedback. But if it's a bad search term, it's a bad lead, and you've taught Google that bad search term is good. That gets worse. That's why it's important to nip those in the bud quickly.
Understood. Let's break this down structurally regarding keywords and negative keywords. How do you do this wrong, good enough, and insanely well? Many generalize, saying, "This is the one thing that matters in PPC," but it's all eight things—and arguably more—that matter.
If you do it wrong, your search terms aren't good—people searching aren't looking for you. If you're doing it right, you at least have seller intent. Before we get to the best-case scenario, let's talk about something else people do wrong, even in the "good" category. Many are too restrictive with keywords. They got one deal from a keyword one day, so they try to put all their spend there. They make judgments about keywords before having data. Where do you draw the line on keyword quality?
We had this discussion in a different episode. The answer in PPC is often "it depends." It's gray; you need data. It's less about finding the absolute best keyword and putting all your budget there, and more about maximizing the peak of the curve between volume and efficiency. Different keywords bring different points on that curve. The best accounts balance budget and volume allocation.
I absolutely agree. There's a spectrum of search terms. Some are extremely high-quality; others are people who don't even have a house to sell. We generally think: get rid of anyone not selling their house. If the keyword means someone wants to sell their house, you should probably have it in your campaign. There are ways to adjust bidding to make it not just coincidence if you get results. We'll talk about bidding in a future episode. Too much emphasis is put on keywords and not enough on bidding. You could have every keyword, but if you bid right, you'd be fine.
Let's talk about doing this at the absolute best level. What are some things we do with keywords to take it from B+ to A+?
It comes down to organization and tiering. We identify keywords with seller intent. Then we organize them into themes/categories (fast cash sellers, companies, etc.). Next, we organize by intent and quality, ensuring we're finding the right audience and speaking to their needs. Then we organize by search profitability.
You're touching on the dynamic motivation algorithm (DMA)—the technology behind our strategy.
Exactly. It's an oversimplification; there are many algorithms. One is a keyword structuring algorithm. How familiar are you with machine learning cluster analysis?
I have some knowledge, but I imagine our audience would like an explanation.
Let's say you have a graph with three dots in the upper right corner and three in the bottom left. You can create clusters—those three dots belong together. With machine learning, the plot could have infinite dimensions and values. Machine learning algorithms can group things together in thousands of variables, just like our eyes can group things. We use a form of cluster analysis to build our keywords. We start with search terms; keywords are a means to an end. We care about search terms. Thousands of keywords could trigger a single search term. Why choose one keyword over another? We analyze on two levels: intent (what does the person want?) and expected conversion rate (expected profitability). We build it into a tiered structure. That's the gold standard of keyword structuring.
In practice, how can the average investor apply this? What can they do to go from poor to average, or average to good?
It's about search term quality. Look through your keywords/search terms and ask: does this person want to sell their house? Aim for above 95%. You could make it 100%, but you'll restrict yourself. You need bad search terms, but fewer of them. Getting this in check is number one. Number two is getting your keyword structure right—organizing keywords in ad groups by topic. Don't have too many or too few ad groups. We usually do 13; I wouldn't have a problem with 20 or 7, but I'd have a problem with 2 or 1000.
If you have major issues with search term quality and can't solve it with negative keywords, consider going broader on negative keywords or narrower on keywords. A good example is broad match keywords. We've had success testing them, but it's advanced, not beginner. Google makes it hard because the default is broad match, which will destroy an account. For beginners, avoid broad match until you're ready.
Anything else?
Build an account maintenance schedule—daily, weekly, monthly. For search terms, probably weekly. Add bad search terms as negative keywords. Review keywords monthly, looking at cost per lead/cost per qualified lead, and pause poor keywords or bid lower.
If you have someone managing your PPC, remember this requires industry knowledge. If you're working with a non-real-estate-specific agency, ask for search term reports and help add negative keywords; you likely know better than they do. If you work with a niche-specific agency, they'll probably be good at this, but if you work with a generic agency, be careful. I consistently see rough stuff from some agencies.
A clear sign of whether an agency is doing their job is if their search terms are strong. Agencies often coast on keeping search terms clean. If you're worried about performance, look at your search terms. If they're poor, the agency probably isn't doing their job.
How do you look at search terms? Go into your Google Ads account, under campaigns, pick a campaign, go under keywords, then search terms. You'll see all the search terms; you can change the date range. Google has removed search term visibility as a privacy measure, so you won't always see every click, but you'll have an idea of overall quality. In my experience, it's two-thirds of the data.
The shared negative keyword list is valuable because there could be so many search terms in one account that you don't see, but the chances they've popped up in another account are decently high. The worst thing about Google is wasting money in places you can't see—that's a topic for another day.
Any closing words?
I think that was a good intro to keyword structures, the benefits of having good search terms…
Absolutely. That's it for the podcast today. Thank you for joining us. I'll see you next time. [Music]