Originally posted on BuyerGenomics. What follows is a deep dive by the Head of Strategy for Endai, Damian Bergamaschi, on how to use paid search in an omni channel marketing strategy.
- Don’t look at the aggregate CPA. It is not reality.
- Look at your search term data to create a search acquisition matrix to better understand how your metrics really look between brand and non-brand and net new acquisition and existing customers- the mix of those four items.
- Build a strategy that shifts the dollars to the best performing quadrants for your business.
- Create an omni-channel strategy that nurtures individuals from brand aware and net new customer acquisition into the brand loyal quadrant using quadrant shepherding as a strategy.
- Have the goal within your organization to get to the place where you can do what the most sophisticated brands do which is actually to bid on search based on lifetime value- also known as ROI-based bidding.
So today we’re doing a little bit of a different format. I will be doing my first solo episode. What we’re going to try to do today is understand the effectiveness of your paid search investment, and we’re going to go through it using my experience, having worked deeply in segmenting paid search accounts to basically uncover some opportunities for you guys to think about. If you have any questions about stuff that we cover because it might be a little in-depth, feel free to email me at firstname.lastname@example.org or at email@example.com
So the very first part of understanding the effectiveness of your paid search investment is to first determine well, what was the objective of that investment to begin with? So for most paid search strategies, it’s about driving sales, and sales can come from a few different places. Primarily, it’s from repeat transactions and also from new customers. So when it comes to hitting the quarter in that particular period, they’re both very valuable. However, I tend to think that net new customer acquisition is a lot more strategic. When you’re driving new customers to your business, that is the lifeblood of a successful organization, and it really speaks about hitting your quarter next quarter too. So the very first thing you need to do to determine how you’re doing at your objective, specifically I’d say new customer acquisition, is to segment your results.
So often I see agencies and even internal marketing organizations look at the aggregate and blended result of, let’s say, something like Google AdWords or your paid search accounts, and what you’ll typically see is, Oh the Blended CPA or the CPA –they won’t even say that– is $6 per sale. And it may feel very good because you may intuitively know that your average order value is maybe $50 or $100, and even after you take the cost of goods sold out, you’re highly profitable on that sale or that acquisition.
But what is typically not known is: would I have gotten those sales anyway because a lot of them are branded search? And how many of them are new customers to begin with? So, the very first thing you need to do to answer those questions… –and by the way, we work with people and organizations that are very very large and also very experienced marketers. Not a lot of people do this. So if you don’t do this today, don’t beat yourself up. It’s very rare that I find an account that is set up and being thought about in this way.– So, the first thing you need to do is: stop looking at keywords. Go into the search Term Report, and look at what people actually type in the search engine. So, a keyword is what you bid on. That’s the thing that you buy. The search term is what people actually typed when you bid on that keyword.
So very often, brands will think that they’re bidding on something that is non-brand, but they’re actually getting all their sales from the brand. So I’ll give you a little example. I’ve done a lot of work with one particular client in this. They’re a steakhouse. So I’ll just refer to that example because it’s pretty helpful to think about it. So they are a- let’s call them the XYZ steakhouse. So when they’re bidding on the term “steakhouse,” that analysts may be thinking alright, that’s someone who’s looking for a steakhouse, but what could also show up if you don’t set up your account the right way is the XYZ steakhouse as the search term. And what you’ll see is -because I see it every time- that a campaign that looks like it’s doing really good on net new customer acquisition actually isn’t because all of the sales are actually from a branded term. So that is one of the very first things you do when you’re trying to deconstruct how your brand versus non-brand search is doing.
Okay so that’s the very first step in segmentation: brand versus non-brand search. It is the quickest proxy for you to identify if you’re generating net new customers or if you are acquiring–you’re retaining– existing customers or getting incremental repeat transactions. Both are valuable, but I tend to think this is a personal opinion that net new acquisition is a little bit more strategic -because a lot of the sales from brand you would have likely gotten anyway, but we’ll deconstruct that further.
So one of the things that you’ll see when you segment brand versus non-brand is that the brand conversions, you might be able to pick a sale up or an acquisition for pennies to a few dollars. And what you’ll also find is that after you do the search term analysis the right way that I was just talking about, you’ll typically find that the non-brand conversions, there are almost none of them. I kid you not. I’ve looked at campaigns where they spent grids of thousands of dollars over a period of weeks or months, and there’ll be like one or two true non-brand conversions. And I challenge you to look in your account because if you’re not really thinking about it to this level, that’s probably what your account looks like.
So the next thing is: figure out what is the true cost per acquisition for your non-brand search. And when you do a really good job at it, it’s still going to be higher than you probably expect. But it has to be a profitable acquisition over some time-frame. But people that do it really really really well, it depends on the industry, but really well is break-even on the first sale because if you’re doing that –if you have a credit card set up on the account, you’re actually getting the money before you even pay the credit card. You get the money that day when you make the sale, and then before the credit card bill comes 30 days later, you’ve already broken even on that. It’s pretty challenging to do much better than that at scale, but I have seen it. And it is possible. Easier to do in industries where the AOV (or the Average Order Value) is higher.
OK. So the first thing is: don’t lose hope. While it’s not an easy fix overnight, it’s actually pretty simple to turn that around. So it’s easy: No, but simple: Yes. But first what we’re going to do today is continue to deconstruct the segmentation of that account because there’s another very important aspect which is the New versus Existing within brand and non-brand. And if you have technology in place where you can track the customer that you acquired to the campaign or the keyword that drove them, you can build a pretty interesting search acquisition matrix that would have four quadrants on it.
And that’s where you’re basically saying on one side, what is the search type? Was it a brand search or a non-brand search? And then even further, what was the purchase history? So, you can go to the Website for the show notes here to see: we made a little infographics that has the four quadrants for this. But basically, if you have somebody that searched your brand, they sought you out directly, and they were an existing customer in the database when you got that sale, they are what we would call “Brand loyal.” Okay. If somebody sought you out, they typed your brand in, and they were never in the database before, that is somebody we would call “Brand aware,” but they’re a new customer.
Okay. So it’s a little bit different. So, not all brand search is searches you would have gotten anyway, and particularly –this is another little hint or tip rather– if you’re being conquested by one of your competitors, and by that I mean one of your competitors is actively buying your search brand, then it’s actually pretty common for you not to get all of your brand search. So, it also depends. Let’s say you’re doing branded advertising, and you have lots of other touches, when somebody comes in on the brand, they can also be net new. So, we shouldn’t assume that all brand search is search you would’ve gotten anyway.
In fact, I have seen on several clients where when you pull back the brand search, yet a lot of it you see go back up in organic, there’s absolutely some cannibalization happening there, but I’ve almost universally in every single time I’ve done this seen that the aggregate, absolute level of sales is not as high as when you’re hitting 100% search impression share for the brand. And the theory that we have- the running theory for this- is that the “brand aware” component, that is the one that’s not a shoe in. And that “brand aware” component at a higher rate when you have a 100 % penetration on the brand aware, that’s what makes the difference. Typically, it can be a 10-15% lift in absolute sales for this relatively low cost per click segment of your search.
Okay. So, the next segment here is the non-brand. So in this quadrant. So on non-brand search, you can have somebody that doesn’t seek your brand directly. They’re searching more of like a category, and an example like that would be “steakhouse near me.” And if they were an existing guest in that case or an existing customer, you may call that a promiscuous customer because they’re in your database. They’ve already purchased from you, yet they do not seek you out directly. They’ll go to whomever. Okay, so sometimes we’ll call them their promiscuous customers. They will literally buy from whomever is, in this case, closest, cheapest or serves their need at that time that they’re looking for.
The next quadrant is what we would call the net new customer quadrant. And that’s somebody that types in a non-brand search, and they’re not in your database. Okay. That’s a pure acquisition. That’s typically where your scale-able volume is. It’s hard to quickly scale the “brand aware” net new acquisitions because you have to spend a lot of money on branding/advertising. Sometimes it even has to be saturation branding advertising, and there’s a lag to that. But when you figure out how to cost effectively acquire the net new customer in that quadrant, that is typically something you can scale pretty quickly. It’s going to be tied to your media budget.
And if you can do it in break-even on the first sale, if you are a small organization that’s more closely held, you can go very quick because you literally are– you might not have as tight of a budget, and you can pay the credit card bill within 30 days of getting that capital, you can go very quick. I’ve seen organizations be able to leg up in a matter of a few months using this type of strategy. And if you’re in a large organization where you have set budgets, then that’s okay too, you’re just going to be a lot more efficient at net new acquisition than you ever have been before.
So there’s also another interesting phenomena that I found, and that’s the concept of taking your search terms because now you’re looking at what people actually type and tying them to the quadrant where they belong. So if you have the type of customer marketing database or customer database platform where you can tie the search click to the individual in your database, you’ll start to see what types of campaigns are driving the “brand aware” versus the “brand loyal” versus the “promiscuous customer” versus the “pure net new acquisition.” And what I’ve found is that, for example, brand aware is a great example. Typically, you want to have 100% of the impressions share on search queries that seem like they could be the first time guests, but they know who you are.
An example of this would be when somebody types the brand in and also attaches an informational query to it. So for example, it might be XYZ steakhouse menu, XYZ steakhouse prices, XYZ steakhouse hours, XYZ steakhouse dress code, reviews about XYZ steakhouse, and you can kind of synthesize that for your own business. The point is, if somebody was truly a brand loyalist they would probably know the answers to these questions. So, at a higher ratio, what you’ll find is those types of searches are “brand aware,” and there’s a very strategic share of brand search to purchase because it is driving net news at a very low cost.
I mean, especially if you’re not being conquested, it could cost you under 10 cents to buy the clicks for those types of searches, and the conversion rates on them may be extremely high. The effective CPA on those types of searches will probably be better than any you’ve ever seen in any other part of the business. And it’s good to know, if you look at your data, that Wow those are actually net news. So even if you’re going to turn your brand search off and focus on organic to scoop that up, it still would probably be strategic to say which part of my brand search should I pull back on, and then maintain 100% share of these low-cost, high-conversion brand aware net news.
Another thing that I’ve found is, when people type your brand in with an exact match, disproportionately those individuals are brand loyalists. They’re in the database at a higher rate than other brand searches. So, if you were going to be more aggressive about shifting away from brand spend, that would be an interesting place to start. Incidentally, I’ve also found that it makes more sense to maintain paid search on brand on mobile than it is on desktop just because it dominates the initial render when you do a search. So that’s like a side little hack, to make sure that you don’t really change the nature of the account too quickly all at once.
Another thing is, there could be certain modifiers to search terms that help cluster people into promiscuous versus just a pure net new acquisition that’s of high value and high potential value. So, one of the things that we found for this steakhouse client was that if you had a modifier like “top rated,” “best,” things like that, the value of those, or the potential value of those individuals is much higher. And in that one, you will see, especially when you tie it to the point of sale data, you’ll see trends emerge. Typically, you have to look at it a little bit. And knowing your business is really valuable in this particular exercise.
Another interesting way to do this, even if you don’t have the ability to tie your search term to your actual customer file, is think about constructing conquest campaigns where you target the brand search of a competitor that has the customer that you would like to have. If you know there’s another competitor out there which you can effectively steal share from, cost effectively of course, and they tend to have the customer profile -the attributes- that make them very similar to your “Most Valuable Buyer,” those are very strategic because you can know and feel very good about intuitively that you’re going after high potential-value customers on search without having the certainty of the data tied to your database.
And also, you’ll find that certain search terms do cluster around promiscuous. For example, the “near me” was a little less loyal than if you typed in “best steakhouse near me.” So one of the things you can do -of course there’s many- is to shift dollars away from one quadrant and into others. So one idea that we talked about was , you have a certain dollar that you invest to cracking net new acquisition so that you can scale it. Then when you do that, you can strategically shift dollars from brand search -the brand loyal search- to net new acquisition, maintain full share of the “brand aware” net new and effectively just change your mix up a little bit because you’re not going to lose where you are.
But if you have a more strategic long term focus, this can be absolutely transformational to a business in a few quarters. I’ve seen organizations shift humongous allocations of dollars away from more traditional media to paid search because it was so easy to prove the result of it and also know how quickly you can scale it and how strategically you scale it at different times of the year. One of the other value creation strategies is the idea of Quadrant shepherding. So if you go back to the Matrix, there’s four quadrants and what you want to do is similar to a shepherd who leads his flock, you can basically find net new guests that are coming in because you’re able to flag them or know with high probability that they’re net new and create an omni-channel nurturing strategy to increase the value of those individuals so that you’ve shepherded them from the net new acquisition quadrant into the brand loyal quadrant.
And what that does over time is, it either brings your media costs down to zero for that individual if you focus on organic earned clicks, or even if you do have a very high impression share on your brand search, you will take that cost of repeat acquisitions down from, maybe that first acquisition cost you 50, 6,0 70, 80, 100 dollars maybe more depending on your vertical, down to something that costs pennies or dollars, and it’s a way to multiply the power of your budget over time when you take that strategy.
And it also sets you up for what I would call the ultimate goal of any paid search acquisition program which is ROI-based bidding because if you have that value creation strategy of Quadrant shepherding, what you’ll do is you’ll systematically increase the value of that new customer because they purchase more frequently or maybe they purchase more. And then, when you can acquire a new customer with a higher lifetime value, or even the higher value within the first year, you can now make ROI decisions on: how much money do we want to make? or at what point into the future do we need to break even on? And you can adjust your bidding based off of the lifetime value of that customer versus just the value of a sale.
And that’s kind of transformational because sometimes it’s hard to know if you’re looking at last touches to say, I want to keep investing in new customer acquisition when I know that I lose $10 a sale. If you look at it over different horizons like 3 months, 6 months, 12 months, 3 years, you will find very often that the price to acquire that customer the first time, if you’re acquiring the right customers, is very small to the lifetime value that is created when you execute a very good omni-channel strategy that nurtures the buyer throughout their lifecycle. So that was a very quick and high level, but hopefully with enough detail, framework for how you can judge the effectiveness of your paid search campaign the way that through the lens of somebody that is very critical of an account as opposed to just trying to make it look like it’s performing, this is how you get to the nature of: are you building the business or not with these paid dollars?
So if you have any questions, I’d love to answer them. I’m kind of a geek when it comes to this stuff, and I have a lot of passion about it. And I love hearing from both business owners as well as marketers working internally at organizations. So, in conclusion, #1: don’t look at the aggregate. It is not reality. #2: look at your search term data to create a search acquisition matrix to better understand how your metrics really look between brand and non-brand and net new acquisition and existing customers- the mix of those four items. Then build a strategy that shifts the dollars to the best performing quadrants for your business. After that, create an omni-channel strategy that nurtures individuals from brand aware and net new customer acquisition into the brand loyal quadrant using quadrant shepherding as a strategy.
And finally, have the goal within your organization to get to the place where you can do what the most sophisticated brands do which is actually to bid on search based on lifetime value- also known as ROI-based bidding. So I hope you guys got a lot out of that today, and we’re going to have follow up episodes where we talk about how to conquest and to go into more detail on some of these particular topics. But I’d love to hear from you. Shoot me an email. Leave a comment in the show notes. Catch you soon.