SEM How to Improve Account Performance and Buyer Personas

SEM How to improve account performance


Smoothly running and decently performing SEM campaigns require getting three things right: tracking, targeting and a good user experience. If you can master these three areas, your campaigns will “deliver the goods” and get you conversions at a decent cost per acquisition (CPA), or sales at a solid return on ad spend (ROAS).


Tracking is everything. First, how can you spend thousands of dollars on your company’s behalf if you can’t track results to prove you were successful? Second, if you aren’t tracking results, then how can you optimize your campaigns by improving poor ads, favoring successful ads and using targeting methods such as keywords or demographic segments?

Tracking is often overlooked or not properly maintained; without it, in the long run, most organizations will not allow you to have a budget. And if you can’t prove your results, you will get a reputation as a “bad steward.”

If you’re using AdWords in combination with Google Analytics (GA), you have the luxury of using “autotagging,” where Google automatically appends a parameter with a string to all of your ads’ URLs. This unique string is a magical index number that allows GA and AdWords to look up dozens of pieces of information about the searcher’s click — city, browser, campaign, ad group, search term and so forth. It literally is a simple checkbox you enable (assuming you go through the multiple steps required to get GA to talk to AdWords as well, that’s a bit more of a challenge, as you need to set a few things on each end).

If instead you’re using Facebook (or LinkedIn Ads, or any Display advertising platform), since those are not Google products, autotagging is not an option; instead you must append “utm” codes for Google Analytics to interpret, on the end of each URL. This is a bit of a pain but is crucial. If you do this, you can slice and dice by campaign, source and medium in Google Analytics, and you will be able to perform attribution analyses — deciding how much credit to give early-funnel clicks versus later-funnel clicks.


There are numerous items that can cause tracking to fail and clicks to be credited instead to (not Set) in Google Analytics; we can’t cover all of them here, but the most common cause (other than neglecting to put tracking in place) is the presence of redirects.

Let’s say, for instance, that your website is configured to have all URLs end in a slash. If you accidentally put into your creative’s destination URL a version of the destination page’s URL that is missing the required slash, often your web server will automatically redirect the searcher to the correct URL. Great, what’s the problem? The problem is often that many web servers will not preserve parameters on the end of a URL in the presence of redirects. So if, for instance, you have an AdWords autotagging parameter on the end of the URL — after it’s redirected — poof, your autotagging parameter is gone, and the keyword will show as (not set) in Google Analytics. Similarly, for Facebook, you can lose your utm parameters. So redirects are the bane of SEM account managers. Check for them, and fix your destination URLs.

Targeting… and toilet seats

I’m partial to AdWords, so I like to talk about keywords as targets, but all of this applies to Facebook and other channels, say, with demographic or remarketing target audiences. Regardless of the channel, if your campaign targeting is poor, results will be poor.

For new SEM account managers, one of the most important things to get right is Google’s keyword “match types,” which define how searches will match against keywords. Defining a keyword with brackets, which represents exact match [nike shoes], will match on the query “nike shoes” but will not match on “blue nike shoes.” Phrase match with quotes “nike shoes” would match on both of those but would not match on “shoes nike.” Expanded broad match with pluses +nike +shoes would match on all of those but would not match on the related phrase “tennis shoes.” Using the keyword with no notation at all signifies broad match, e.g., nike shoes, which might match on all of those. Broad match is best to avoid where possible, although it can be useful as a method of keyword research, or if it’s used in combination with numerous negatives.

Many new SEM account managers ignore match types and throw keywords into an account, not realizing the default is broad match. True story: I audited an AdWords account a few years back for an advertising agency (a classic case of the cobbler’s stepchild; I’ve worked with several ad agencies to help them get their own leads). This agency was bidding, broad match, on perfectly reasonable terms like “advertising agency,” “marketing companies” and so forth. I pulled a “search query report” from AdWords to see what actual searches had matched on their keywords, and — I’m not making this up — they had paid recently for a click on the search “toilet seat.” How that matched on one of these terms, I can’t say.

So, if you don’t know what you’re doing, if you’re starting out, I recommend you use exact match and also the approach with the pluses. It allows the words to be in different orders and other words to be present, but it guarantees the words you have specified will be in the query. That should ensure you’re not bringing in anything too crazy.

Running search query reports on a regular basis, and then adding negatives to the account, is crucial in order to maintain account performance. Google’s matching algorithms seem to drift and change over time, as do searchers’ interests, so you will always see new queries appearing. But a well-managed account, in my experience, should have no more than 0.5 percent to 2 percent “Wasted Spend.”

Good user experience

If you’re driving people somewhere, that somewhere had better be pretty compelling. Google understands this and has built features into AdWords to favor good searcher experiences, focusing in particular on landing page relevance, performance and quality.

Landing page relevance

Relevance and a great user experience are critical in order to get people to take action. User experience is linked closely to targeting; if you’re targeting people with a particular interest or intent, then they probably deserve a dedicated landing page that speaks to their issues and concerns. If you’re targeting 3,000 keywords, how can one landing page possibly satisfy a wide range of concerns?

One rule of thumb, based on my field experience with several dozen clients over the years, is: have one landing page for every $1,000 of monthly spend. If you’re spending $500/month, then one landing page, or simply your site’s home page, or contact page, is probably adequate. If you’re spending $100,000/month, this rule of thumb suggests 100 landing pages. This can easily be achieved if your business is an e-commerce one where you’re sending people to multiple product pages; if you are doing lead generation with calls to action like “Download our free guide,” you should at least consider generating numerous landing page clones with different titles/copy on them.

Landing page performance

Performance is, in my opinion, about two things: the mobile experience and speed. Most landing pages look great on desktop, as that is where they were developed. But it’s critical to make sure that they are mobile-friendly. Many marketers are developing their landing pages as mobile-first, with the desktop version as the afterthought. Speed is also critical. Users have little patience and time, and Google, via its “Quality Score” mechanism, actually rewards you directly by discounting what you pay. Landing page load time is one component of Quality Score.

Landing page quality

Landing page quality means many things to many people, and there are numerous perspectives on this. Some things that can rapidly communicate quality to the viewer include endorsements/badges, the company’s name/address at the bottom of the screen and the presence or absence of a Privacy Policy and Terms of Service on the page. Note that of these, several of them are possible for Google to measure.

I have had several clients who failed to have a Privacy Policy on their landing pages. Their quality scores (QS) all capped out at 8 out of 10; several weeks after they added a Privacy Policy, quality scores improved in their accounts across the board, and suddenly their top terms (i.e., brand, due to high CTR, which is rewarded with a high QS)  moved from 8 to 10. So, although it’s not an aspect explicitly documented anywhere by Google, I’m convinced having a Privacy Policy is a key Quality signal Google relies on.

So, for landing pages, at a minimum, make sure:

  • they load.
  • there are no redirects.
  • they are fast.
  • they have a privacy policy.

If you ensure these four things, you have a good shot at having your landing pages perform at least adequately, and you can then focus on usability and messaging to increase conversions.



 You’ve done your research and created buyer personas — now what?


If you have a love/hate relationship with buyer personas, you’re not alone. According to a 2016 survey by Customer Think, 72 percent of marketers are familiar with these personas, and 60 percent have created their first persona within the last two years.

Unfortunately, lots of search marketers are confused about how to use their personas or are frustrated that their personas are not producing the hoped-for results.

For this reason, buyer personas have been criticized as an exercise which, although well-intentioned, can distract marketers from actually producing meaningful deliverables.

So, let’s get practical.

A typical buyer persona contains a lot of “touchy-feely” stuff: the interests, values, behaviors and pain points that motivate a particular market niche, all wrapped up in a cutesy name and a stock photo picture to match. That’s nice, but it often isn’t all that useful when it comes to marketing metrics like cost per click (CPC) or return on investment (ROI).

However, if you add a few cold, hard financial facts to a buyer persona, you can instantly start to use it to refine your search marketing budget. Now, I know, this isn’t a typical way to use personas, but it’s a great way to start using them to actually change your business.

With that in mind, let’s take a look at three financial aspects of buyer personas and how to use them to build a better budget:

1. Lifetime value

Each buyer persona you target has different goals, behaviors, background… and monetary value for your business.

For example, if you sell bicycle parts, one target audience might be mechanically-minded kids who don’t want to pay for repairs at a shop. You should expect small purchases from this group.

If you’re also targeting people who restore vintage bicycles, you can expect that they will be willing to pay more to get the quality they want. If you target bike shops, however, you might sell at a lower per-product cost but more than make up the difference in volume.

To make things even more complicated, the value of a customer often includes more than a single purchase. This is where lifetime value (LTV) comes into play. Lifetime value includes all of the revenue produced by a single customer over the lifetime of that customer.

For example, a bike store might buy a box of chains for $50, but if they keep buying a box every month for five years, their LTV is actually $3,000!

When you add customer worth to your buyer personas, you’ll be selling them short if you neglect LTV. After all, a customer who buys a Ferrari on a $0 down payment promotion is worth a lot more than $0!

2. Cost per acquisition

Knowing how much your buyer personas are worth is important because the monetary value of a persona directly influences how much you’re willing to spend to get them.

Think about it this way: In order to recruit the nation’s best high school quarterback, colleges are willing to pay for airplane flights, campus tours, fancy dinners and scholarships. For a no-name linebacker… not so much.

In the business realm, cost-per-acquisition (CPA) is one of the easiest ways to measure the success of your paid search campaigns. The question, then, is: “What is a good CPA for my persona?”

Now, your knee-jerk answer to this question might be, “Anything less than the persona’s LTV.” However, while this makes sense on a superficial level, you have to remember that marketing is not your only expense. When you take the fixed and variable costs of your business into account, you can spend $2 on marketing, make a $4 sale, and still be losing money.

In my experience, to make a reasonable profit with a buyer persona, the LTV for a given persona needs to be at least four or five times larger than their CPA. If it costs more than 25 percent of your LTV to close a buyer persona, something in your marketing needs to change.

3. Buyer scarcity

The last important financial aspect of each buyer persona is how many members of that persona are actually up for grabs.

You may identify a buyer persona that is very profitable, but if there are only a few of those buyers in the nation, pouring more and more money into marketing won’t make more of them appear. As the Wall Street saying goes: “Trees don’t grow to the sky.”

If you are an established business, you can figure out how many customers you’re getting from each of your target groups by examining data from previous years. However, if you’re a startup, you may need to rely more on area demographics, industry averages and educated guesses, similar to how you calculate your market size and market share.

Knowing how many potential customers you have within each persona will help you decide how much you can spend marketing to them and still expect a return on your investment.

Putting it all together

Now, let’s put all of this together and see how you can use the financial data associated with each buyer persona to plan an effective search marketing budget for 2017.

Imagine a company called “Green Room Lights” that makes and sells energy-efficient light bulbs. Green Room Lights has recently researched their customer base and come up with three distinct buyer personas (complete with stock photo pictures!):


Jenny is the head of a single household. She cares about the environment and has decided to make her house more eco-friendly. Jenny is searching for an Energy Star qualifying bulb to replace her current incandescent bulbs.


Mark is a contractor who builds small-to-medium-sized apartment complexes. He advertises to his clients that he uses energy-efficient construction methods that will save them money in the long run. Mark is looking for the most cost-effective way to deliver on this promise.


Nicki is a buyer at a large chain hardware/home improvement store. As incandescent bulb sales have begun to taper, she has decided to explore more energy-efficient replacements to stock. She wants a product she can sell at a reasonable profit margin that will attract buyers like Jenny.

Green Room Lights has a goal of getting $4 million in gross sales through $1 million of marketing. It wants to know how much to budget for marketing to each of these three groups to accomplish this goal. So, Green Room’s clever marketing director starts looking at the financial aspects of the company’s buyer personas.

Lifetime Value

First, let’s take a look at each persona’s LTV:


Like the average American, Jenny’s household has 45 light bulbs. Whether all at once or gradually, Jenny intends to replace them all. She buys a more expensive bulb ($10 each) because an eco-friendly product is more important to her than price. Because the bulbs will last for years, she is not expected to make any more purchases in the future.

Mark is looking for a bulk order price of $5 per bulb. He installs 1,200 bulbs in each apartment complex and builds two apartment complexes per year.

Nicki wants an even lower wholesale price of $4 per bulb. She stocks 2,200 stores with 200 bulbs each over the course of a calendar year.

Cost per acquisition

Now, let’s take a look at what Green Room Lights can afford to spend on each type of customer:

Because Jenny buys her bulbs at full price, Green Room Lights will make a profit at $1 of sales for every 25 cents of marketing.

Mark gets his light bulbs at a price with a smaller profit margin for Green Room Lights, so their marketing needs to be more efficient for him (20 cents of marketing per dollar of sales).

Narrow profit margins mean that marketing to Nicki needs to be the most efficient of all, producing $1 of LTV for every 18 cents of marketing.

Buyer availability

Lastly, let’s take a look at Green Light Room’s customer base (CB):


Of the 124.6 million households in the US, 29 percent use no incandescent bulbs, making the “Jenny” customers an almost inexhaustible resource.

There are 477,950 contractors in the US, but let’s assume only 2 percent of these, like Mark, build energy-efficient apartment complexes. Based on last year’s numbers, Green Room Lights has also been getting 0.5 percent of this market and would like to up that to 1 percent.

In a given year, there are only one or two Nickis up for grabs. In fact, Green Room Lights has only made one “Nicki” sale in the past and would feel lucky to get even one more, but they will do whatever it takes to lock a Nicki down in 2017.

Planning the search marketing budget

Now that Green Light’s marketing director has this information, it should be simple to plan out the budget. “Mark” and “Nicki” customers are more valuable to the company, so they will try to get every last sale possible, and any leftover money will be spent on “Jenny” customers:


This distribution of funds provides direction and clear goals for each campaign and ensures that each type of customer is receiving an appropriate level of attention.

Make sense?

By the way, figuring out the optimum arrangement of buyer personas can be challenging, so to make things easier, I’ve actually put together a free marketing budget calculator you can use (click here to access the calculator). All you have to do is drop in your CPA and LTV for each buyer persona and identify how many sales you expect on a monthly basis from each persona, and the calculator will give you your expected revenue and marketing budget. Give it a try!

By Ted Ives & Jacob Baadsgaard