This interview with David Rodnitzky of 3Q Digital is part of the Convirza Webinar Series. The following is the Q&A portion of the interview.
McKay Allen: Awesome. David, thank you very much, that was awesome. Very detailed, very good information. We’re getting a few questions just asking for clarification on what exactly the Lin-Rodnitzky ratio shows. Like, give us a sort of a synopsis, if you will, or maybe a more in-depth detailed answer but what exactly that indicates about your campaigns and about your ads.
David: Yeah, so it’s a benchmarking tool. And it indicates in a nutshell whether your campaign is too conservative, too aggressive or just right. So if you had a ratio that was under 1.5, that would indicate that you are probably being very conservative in your campaign. You’re probably only buying the keywords that are the absolute slam-dunk winners. There’s no doubt that these are going to thrive… pass the performance for you, but you’re not doing enough testing in keywords that may not seem obvious to perform, but could open up a lot of traffic for you. Now the flip side of the that is if you have a high Lin-Rodnitzky ratio, it means you’re buying a lot keywords or you’re being matched in a lot of queries that are junk, and you’re not cleaning up your account enough. And in that case, your goal is to bring your Lin-Rodnitzky ratio down by cleaning up those keywords or queries that are not performing for you.
McKay Allen: When you initially run- yeah, that makes perfect sense. When you initially run benchmarking tests, do you see… are people higher or are they low initially? I mean where do the majority of people of make mistakes and on which end of the ideal spectrum, there?
David: That’s a good question. I would say nine out of 10 people are higher, so it’s usually… I would describe it as a high-class problem. If you’ve created such a tight campaign that almost every click you get turns into a conversion, it’s usually the case that people don’t realize that they’re getting matched on a lot of bad queries. And we see a very high Lin-Rodnitzky ratio. I would say that, you know, in general most ratios I see are in the sort of three to four range, so not bad. But the opportunities. I’ve certainly seen ratios in the 15, 20 range, which means that the person who has set up the account doesn’t really understand what’s happening to their account, or doesn’t understand how to make the adjustments. And that’s when you can sort of have dramatic improvement which is save you a lot of money and spend a lot more money on the winning terms.
McKay Allen: That’s what I was going to say. It sounds like from what you’re saying then the majority of people are wasting a bit of money. In some cases probably a fairly large sum of money on… because their ratio is high. Because they’re not cleaning up the account as you say. There seems to be a fair bit of waste, then, with this.
David: There is. There is, and I mean, I think that’s the blessing and the curse of keyword buying. Keywords are great in the sense that you have complete control over your destiny, there’s no other advertising medium where you can buy a word or phrase that exactly represents what you’re selling. But they’re a curse in the sense that Google, if you give them enough leeway, they will find a lot of queries that really aren’t what you were intending when you bought that keyword, and that’s when people run into trouble and end up seeing their costs skyrocket. And I often sort of get this quizzical look, where they say, “I don’t know how my competitors are able to buy this keyword, I try to buy it and it’s nowhere near profitable.” And it’s often because the competitor is just buying that query exactly, and you’re buying that query, plus 50 other queries that you don’t realize you’re buying.
McKay Allen: Very interesting. Very good. Couple questions here that I think are pretty good from our audience. How do you factor, or do you factor in assisted conversions in the ratio? Is there a way to factor that into your algorithm or your metric there?
David: There really isn’t. The way that we look at assists would be, not from a Lin-Rodnitzky ratio perspective, but from just an overall query performance perspective. So we will typically work with an attribution partner, and when we get the performance data on queries, we will use the attribution partner to show us whether that query gets a 100% of the credit, or whether queries upstream get more credit, or even whether there’s other marketing sources that deserve the credit. And what that does is it allows us to adjust it, rather than the credit that we’re giving to that query and make a better bid adjustment. But it doesn’t really inform us as to whether or not the Lin-Rodnitzky ratio has changed.
There may be instances where there’s a query that we have decided is not performing, and we’re going to exclude it in the alpha beta structure, and then the attribution technology suggests that actually it’s a good assist keyword, so we might resurrect it. But I think the thing to figure about any benchmarking tool is, you know, benchmarks by their nature are, you know, averages and estimations. They’re not exact science. Looking into attribution, I totally recommend it. I think it’s becoming more and more important as we live in a multi-channel, multi-device marketing world. But for the purposes of benchmarking we haven’t, sort of, integrated it in that function.
McKay Allen: Awesome. And then a question here that is, sort of interesting. Question from Ed, he says, you know, of course, Google announced enhanced ad words campaigns a couple months ago. How does that impact all of this? Some of their new enhanced features that they’re rolling out. I believe they’re rolling them out in July.
David: Yeah, July is – sorry. It’s July, it was initially June and they moved it back to July.
McKay Allen: Right. So how does that work into all of this?
David: Enhanced campaigns doesn’t really impact the alpha beta structure very much. You know, what we’ve done in the past is we have… when we had more granularity around device segmentation, we could sometimes take an alpha campaign and break it into an alpha mobile, and an alpha desktop campaign and bid differently. You know, enhance campaigns limits your ability to be granular, so at the end of the day in some respects, it actually, you know, I would say in a way it makes life simpler. It requires less. It’s going to require less campaigns. But, you know, as a best practice, before and after enhanced campaigns, it’s always important to look at an account level, a campaign, and a keyword level whether to keep the keyword that’s performing better on one device versus another.
And, you know, if there’s significant difference in how a keyword’s performing on, let’s say a mobile versus a desktop, then you’d probably want to split that keyword into separate campaigns and have a different mobile bid boost, which is the new terminology for enhanced campaigns for that keyword, in a sort of mobile-focused campaign versus the desktop campaign. But again, I mean, from an alpha beta perspective, we’re primarily just looking at the efficacy of the query and not initially concerning ourselves with the device that it’s being targeted against.
McKay Allen: Great. You get a-
David: I mean, enhanced is a pain in the butt.
McKay Allen: Yeah, it seems — I wasn’t going to say that, but you’re… I think it seems that way. And we actually got a couple questions, if you could speak generally about bidding strategies. And I know that’s obviously with how you, you know, it goes down to targeting and it goes down to some of the stuff you talked about. But any broad strokes advice on bidding strategies, or is it all just a matter of testing and benchmarking, that sort of thing?
David: Well, I think, I’ll say I guess two broad comments. One is that I prefer to think about my bidding for my head terms, for the most profitable, highest-volume queries in a more manual fashion. So I try to treat those like my little babies, and I’m making the bid adjustments myself. And I usually rely on bid management tools, like a Marin software, for the tail terms. So if you happen to have an account that has, or let’s say you have a website that has 50,000 SKUs, you’re an e-commerce company. And there are 500, or 200 SKUs that are making up 95% of your revenue, you might handle those by hand, or by some very custom bid. For the 49,000 others that collectively get, the remaining 5% of your traffic, you can look to a bid algorithm to make those bid changes.
The other thing I’ll say is that the best practice that I have in bidding is to get to the most granular… is to integrate the most granular tracking possible in making your bid decisions. So a lot of people for example, settle for cost per acquisition as their metric, when they really should be looking at return on ad spend. So looking at the actual revenue that that keyword is providing and then dividing that by cost, instead of just saying, “Well, I’m going to bid all my keywords to a $20 CPA,” when in fact one keyword should really only be worth $5 and another keyword should be worth, you know, $300 based on the revenue that it’s bringing in. And then there are people who only bid to return on ad spend, who should actually be looking at profit.
So you might have a 300% return on ad spend objective, when in fact some keywords have 90% margin and other keywords have 10%, and if you bid those to the same row as one’s going to be massively profitable, you should be bidding more, and one’s going to be massively unprofitable and you should be bidding less. So at the end of the day, I mean, maybe this is a good plug for Convirza. But the tracking that you’ve set up, whether it’s tracking phone orders or off-line transactions or profit dollars, all needs to be set up properly to get the right bidding strategy in place.
Because the more sort of general you are about your bids, you may have a really great bid algorithm, but if you’re bidding everything to the same CPA, when you should, in fact be bidding to a combination of attributed revenue and profit, you’re going to be, as I think Abanash Koscheck said, “garbage in, garbage out.” You’re going to make bad decisions. So those would be my two tips.
McKay Allen: That’s great. And I’ll piggyback off that too, if I might. I mean, one of the data points we’re finding is that in many industries, even if the number of calls produced via PPC are lower than the number of form fill-outs, whatever the call to action is on your landing page, those calls, in many cases that we’ve documented, actually turn into more revenue dollars than the form fill-outs do. And so…
David: Yeah, exactly.
McKay Allen: Right. It does. So something to keep in mind.
David: I’ve heard anecdotally that a call is three times more valuable than a form.
McKay Allen: Yeah. Yeah. I mean, our data actually backs that up. We actually think the… more valuable, that’s very interesting actually. Our data shows that someone who fills out a form. And this is fairly documented, I think, Mech Labs and Marketing Sherpa and HubSpot have done data to this effect. But you know, anywhere from 5% to 6%, in that neighborhood of people who fill out a form, end up becoming customers in some industries, if they’re not filling out an actual… you know, an actual… becoming a customer as they’re filling out the form, e-commerce style. But on the phone, I mean, gosh, you can have close rates that are 20% higher of everybody who calls. So you need fewer calls, is my point.
David: Yeah. Well I mean, think about it this way. How serious is a customer who fills out a form, versus who calls you, versus who walks in your store? You know?
McKay Allen: Right.
David: You know? I mean, obviously the more they’re willing to interact with you, the more likely they are that they’re serious about making a purchase.
McKay Allen: Right. They’re lower in the funnel, if they’re calling you.
David: Exactly.
McKay Allen: Yeah. That’s very true. We’ve got another specific question about bidding, if I can throw this at you. It’s from Rick. He says, all right, I’ll read this to you and we’ll decipher. He says, “We like to bid on the term “free” with different bids than we do other terms. For example, we might bid on “free lead generation” but would want to pay less for that than we would for queries just for “lead generation.” How would I get Google to respect those two bids from a structural standpoint?” Does that question make sense?
David: Yeah, it does.
McKay Allen: Okay.
David: So the easiest way to do that would be to buy the term “free lead generation” on exact match, and buy the word “lead generation” in exact match and then bid separately. You buy those two on exact match, if someone types in “free lead generation” you cannot be matched on your exact match term “lead generation,” and vice-versa. So when you buy phrases on exact match, you are basically telling Google, “This is the only combination of words that you can match me on,” and then you can create separate bidding, separate ad text and separate landing pages against that specific phrase. So end of the day, the exact match is your friend if you have that sort of distinction that you’re trying to bid differently on.
McKay Allen: Okay. That makes a lot of sense. Good stuff. Good questions from people today and good information from you. What other last comments or thoughts do you have to leave with us today, David, before we all hit the road?
David: You know, I think the beauty of online marketing and ad words is the granularity of data and the mathematical information that you have available to make decisions. And, you know, at the end of the day, I look at all advertising online as zeroes and ones. So the word “Britney Spears” sells lead generation software, then I’m going to buy that keyword. So just keep digging into the data and make sure data is set up properly, and running analysis over and over again and you’ll get good results.
McKay Allen: That’s great. And I do appreciate you too, David, taking the time to do this. I know you’re a very busy guy. And I know your time is valuable. So we appreciate you spending 40-45 minutes with us today. We really do.
David: Thanks McKay, I appreciate it.
