Susan Waldes – Search Engine Land News On Search Engines, Search Engine Optimization (SEO) & Search Engine Marketing (SEM) Tue, 05 Apr 2016 21:01:10 +0000 en-US hourly 1 The 4 Cs: A recipe for precise remarketing campaigns /the-4-cs-a-recipe-for-precise-remarketing-campaigns-244386 Mon, 04 Apr 2016 13:43:06 +0000 http:/?p=244386 A lot has been said about how to best structure search and shopping campaigns, but what about remarketing campaigns? Columnist Susan Waldes discusses her formula for successful remarketing campaign structure.

The post The 4 Cs: A recipe for precise remarketing campaigns appeared first on Search Engine Land.


Last fall, I had a great time participating in “The Great PPC Account Structure Debate” at SMX East. And I loved reading how the debate shook down recently with a different panel at SMX West.

You don’t have to have attended one of those sessions, though, to know that paid search people are passionate and vocal about their preferred account structure approaches. From speaker panels to articles to white papers, you could drown in the amount of information on structuring search campaigns (and shopping campaigns).

If the well is overflowing on how to best structure “query-based” campaign types, it’s comparatively dry for another important campaign type: remarketing.

In the last couple of months, I’ve built two different remarketing campaigns where the initial couple of days of performance data was not looking good. In both cases, I was so confident in the setup of those campaigns that I knew that something was wrong with either the website or the conversion tracking.

In both cases, I discovered that I was correct: the remarketing campaigns were working, but there were cart issues in one case and conversion tracking issues in the other.

The reason I had that level of confidence is that I always structure my remarketing campaigns with the same approach: the 4 Cs.

About the 4 Cs

The 4 Cs stand for content, cookie duration, creative and context. They each provide an indication of the “value” of a return visit from a prospect. Therefore, each C or combination of Cs lives in its own ad group, with a bid that addresses the hierarchical value of that visit.

Before we get any further into the structure stuff, let’s dig into the Cs and why each matters.

Content: This is the area of your site that somebody has visited. You don’t need to segment every page of your site here with its own audience and ad group. You want to create audiences that are grouped by the likelihood to convert.

For an e-commerce site, you might target four main content-based audiences: people who have been to home, category pages, product pages and the cart. Similar segmentation can be done for Lead Gen, depending on your product and site architecture; people who have started a form or been to your pricing page likely have a much higher chance of converting than those who have only been to your home page.

Cookie Duration: For most sites, recent visitors have a higher likelihood of converting than less recent visitors. The right cookie durations to use for your remarketing audiences will depend on your product and funnel. Just as with the content, you don’t need to break out every single day since visit into its own audience; you want to create groupings that have similar expected conversion rates. As a basic rule, I most often use two-day, seven-day and 30-day cookie durations on my audiences.

Creative: Refers to the type of ad that you are serving to your audience. You might want to target your seven-day cart abandoner audiences with dynamic display ads, static banner ads and text ads. If that’s the case, you should have three unique ad groups: one with only dynamic display ads, one with only static banner ads and one with only text ads.

The reason we segment by creative is, again, varying performance expectations. Each ad type has its own set of expected behavior in terms of click-through and conversion rates. Your static banners might have higher CTRs than your text ads because they are visually attractive, but lower conversion rates because they don’t contain as much information. Your text ads might have low CTRs and great conversion rates because all that text prequalifies the click more effectively. To bid appropriately in each scenario, each creative type should be broken into its own ad group.

Context: Context refers to the placements on which your ad is appearing. Context is a strong signal of intent and readiness to move down the funnel; a person is much more likely to convert to a B2B software lead when they are browsing software-related content than they are when they are checking the weather.

From a practical level, context layers can be added via keywords, topics or placements. It’s important to segment your ad groups by high contextual relevance and low contextual relevance so you can bid appropriately to each target.

Building 4 C remarketing campaigns

Content and Cookie Duration are applied when you create your remarketing audiences. In my e-commerce example, I’d want to create four different content segments (home, category, product, cart) and three different cookie durations (30-day, seven-day and two-day) for 12 total audiences.

Content is defined in the “people who visited a page with the following” section. Cookie duration is defined in the “membership duration” section. Make sure to name your audiences in a way that makes them easy to define and work with later.


Now that I have my 12 audience segments, I create 12 ad groups and apply one audience to each ad group.

Next, we want to duplicate these groups for each type of creative I’ll be using. For this e-commerce client, we’ll be using text ads, image ads (banners) and dynamic display ads. I clone my initial 12 ad groups for the three types of creative, naming each appropriately. I now have 36 ad groups: four content segments X three cookie duration segments X three creative types.


After I apply the creative units to each respective ad group, I clone everything one more time to add a contextual layer. That contextual layer might consist of a topic category or categories, contextual keywords or a list of placements.

When choosing contextual targets, remember that this is a remarketing campaign. You already know this person is somewhat interested in what you offer; you are just applying a contextual layer that indicates they are more likely to convert.

So for this example, an athletic shoe retailer, if I were creating a regular Display network campaign, I might target a very narrow topic of “Shopping>Apparel>Footwear>Athletic Shoes.” However, within my remarketing campaign, my contextual layer could be much more broad.

I would choose the whole topic category of “Shopping” and “Fitness.” If this person has been to my site and is in “shopping mode” or thinking about “fitness,” they are more likely to convert versus when they are reading business news or checking the weather.


Bidding to the 4 Cs

We created 72 ad groups with 72 different combinations of targeting and creative so that we can bid precisely on each one. The value of a click from a two-day cart abandoner on topical content can easily be 20x or more the value of a 30-day home page visitor.

Start with a base bid on everything. Then filter your ad groups and increase it for each more targeted audience. Your seven-day ad groups might get +50%, your two-day +100%, your product pages +80%, your cart abandons +150%, your topical content +50%.

You’ll end up with a hierarchical tiered bidding system. For my example client, my bids start at a $.60 CPC on a 30-day home page audience and go up to $12.22 CPC on my two-day cart abandons on topical content.

If you have data within prior remarketing campaigns or Google analytics, this bid hierarchy can be very data-driven, and the bids can directly correlate to the conversion rates that are presented in your data.

That said, don’t be afraid of just guesstimating. You can adjust your bids as the data comes in, and ultimately, our goal is actually to utilize the 5th C, Conversion Optimizer.

The 5th C: Conversion Optimizer

Once you have reached the 15-conversion threshold on your remarketing campaign, switch your bidding to a CPA (or ROAS) target.

When Google takes over the bidding, you have created a foundation of hierarchical conversion rates that each can be automatically bid to the right level. In addition, you allow Google to layer in additional factors like device, geography and browser to determine precise CPCs based on the value of each impression.

It’s a precise and well-oiled remarketing machine that will bid appropriately to each user segment and their value to your company.

The 4 Cs are about control

The 4 Cs are not the only factors that determine a successful remarketing campaign. Your device strategy, impression caps, negative audiences, creative strategy, even your product itself, will determine how well remarketing works for you and how large you can scale a remarketing program.

However, this campaign structure will provide a foundation that you can count on, and that gives you the control to optimize effectively.

Think of it like the precision campaign approaches that you use for your keyword search campaigns. Advertisers often isolate their highest-value head terms as exact match keywords in very precisely targeted ad groups with high precise bids and supplement that with lower bid, less precision broader match keywords in other campaigns or ad groups.

Structuring your remarketing campaigns with the 4 Cs gives you that same level of control, predictability and precision on your remarketing campaigns.

C is for Cookie (and Content, and Creative, and Context, and Conversion Optimizer, and Campaign Control)! Now get cooking with Cs!

The post The 4 Cs: A recipe for precise remarketing campaigns appeared first on Search Engine Land.

The Never-Before-Seen AdWords Data Points That Have Been Quietly Rolled Out To Your Account /never-seen-adwords-data-thats-quietly-rolled-account-237998 Mon, 14 Dec 2015 14:31:32 +0000 http:/?p=237998 Columnist Susan Waldes dives into the new Report Editor in AdWords and explores the new data points available to AdWords advertisers.

The post The Never-Before-Seen AdWords Data Points That Have Been Quietly Rolled Out To Your Account appeared first on Search Engine Land.


When I started writing this article a couple of weeks ago, I was going to write a guide to the data points that are available in the new AdWords Report Editor which aren’t available in the regular “Campaigns” section of AdWords.

However, over the last couple of weeks, AdWords has launched additional updates that also allow us to view almost all these new data sets throughout the “Campaigns” section of the platform. As I’m writing this, there is only one view of the data that you can get in the Report Editor and not elsewhere. What are these new data points? What is this Report Editor? Why is this important?

Because more data! Duh! Let’s cover the Report Editor first.

Report Editor

The new Report Editor section of AdWords has been rolling out to accounts since the spring, and just over the last couple months, it is finally available in all accounts. The key benefit of the new Report Editor is time savings: the filtering and report-building tools allow you to construct views of your data that previously required downloading a report (or many reports) and working with the data in Excel (or other external tools) to create the right pivot or extract the needed data points.

Before the new Report Editor,  if I wanted to see performance by day of the week, only on my ad groups containing a certain phrase and only on mobile devices, it would take heavy lifting (“heavy” used in a search-nerd-behind-a-screen context). I would have had to filter my ad groups, download the report with a day of week and device segment, extract only the mobile data in Excel and create a pivot table on day of the week.

Now, I just slide the needed columns over, add a couple of filters and voila!

report editor day of week filter

I get these data in 20 seconds instead of 20 minutes. [Click to enlarge.]

Manipulating data in Excel when you’re digging for performance insights and optimization opportunities is no big deal. When a set of data like this is something you need to look at on an ongoing basis, the 20 minutes you spend putting together this data weekly adds up quickly.  These new reports can also be saved and scheduled for email delivery at your chosen interval, which is where the time savings start to multiply.

If you haven’t bothered to play around with these new tools yet, take the time to do so. You will find something here that will give you back at least a few minutes a week that you spend on reporting or analysis. While there, if you look closely, you will also run into some new views of data that you’ve never been able to see before.

Ad Performance By Keyword, “Placement,” Query & Match Type

In my first version of this article, I was going to tell you that when you explore the new Report Editor, you’ll find the ability to see ad text data segmented by keyword, search term and match type. Then, just a couple of days ago, a new segment popped up on the “Ads” tab of the “Campaigns” section. You can now segment by  “Keyword/Placement” in the main UI section of AdWords, as well.

new keyword placement segment

We’ve never before been able to see which keywords are generating activity on a given ad text within AdWords. In fact, much of the reasoning behind a SKAG strategy (single-keyword ad group) is to facilitate one-to-one reporting on a single keyword and ad text performance.

If you are using multiple keywords and multiple ads in each ad group, this ad + keyword-level reporting, available in both the regular UI and the Report Editor, is sure to help find inefficiencies and/or opportunities.

You’ll see instances where certain ad groups should be segmented into smaller keyword sets that perform best with one ad and others with another. You may also see combinations of ads and keywords that are just weighing down an ad group and should be cut altogether.

If you do use a SKAG strategy, this reporting will only be valuable to you on your broader match type campaigns at a search term (query) level, which currently is only available in the Report Editor and on the Search Term dimension report by segmenting by “Ad” when downloading. Also only available in the Report Editor is the ability to segment ads by match type of the triggering keywords; I’m not sure of the use case for this report, but I’m sure there’s at least one.

On the other hand, only available as a segment on the ads tab, is the ability to execute similar segmentation on your GDN (Google Display Network) campaigns by “placement.” Note that “placement” data refers to the specific “Target and Bid” targets within a GDN campaign, whether they’re literal placements, remarketing audiences, topics or somehing else.

Automatic GDN matches, including automatic placements, all fall under the “AutomaticContent” row. In addition, the “placements” that are recorded here use the esoteric language of the API, such as “boomuserlist” (translation: a remarketing audience). The data are exciting, but a usability improvement will be required before a lot of advertisers are able to apply the insights.


Conversion Action Names, Everywhere!

Conversion action name data down to the query level was another exciting development in the Reports Editor that I intended to write about. Until recently, we could only break out data for each of our different conversions actions via the “custom columns” functionality at the campaign, ad and ad group level.

“Conversion Action” refers to the multiple conversion points/pixels that we create in the tools section and the goals we import to AdWords from Google Analytics. This may include phone calls, leads, sales, app downloads, newsletter signups and more; sometimes all of these are legitimate conversions on the same site/AdWords account. Formerly, we did not have full visibility into the granular keywords and queries that were generating phone calls versus those that generated web leads versus sales.

As of a few days ago, not only is the keyword and query data by conversion action name available in the Reports Editor, but almost every view in every UI tab has “conversion” segments for conversion action name and conversion category. We can get this data segmentation by demographics, locations and even on ad extension reporting.

Conversion action name data is available virtually everywhere, even on structured snippets extensions

Conversion action name data are available virtually everywhere now, even on structured snippets extensions. [Click to enlarge.]

A couple of notes on conversion action data:

  • You won’t find every metric in these segments, only conversion-related data (total conversions, cost per conversion, conversion rate, total conversion value and so on). We can’t see impression, spend and click data, as you can’t tie these “pre-click” metrics to this data split, which happens at or after the click and where one click can generate multiple conversion actions. That said, beware of the cost/conversion data and the conversion rate data in these segments. These data points are calculated using each segment against the total clicks and costs, so they aren’t necessarily indicative of true performance. For instance, we’ve spent $14,133.08 on the keyword below which generated 207 leads and one phone call from a call extension. The conversions aren’t “either/or,” though, so saying my CPA on the phone call is over 14k is a bit misleading.
    cpa conversion action segmentation
  • These data only apply to “conversions” and “all conversions” columns (not clicked conversions), which makes sense, as we need to be click-agnostic to see the total site conversion actions that have been performed. In the Report Editor, you are only able to choose the more inclusive columns. If you are using the segment in the main UI, make sure your columns are not set to “Clicked Conversions” metrics, or you will not see any data under your segments.

But Wait, There’s More!

A couple of other never-before-seen data points are now available, as well. There is a segment on the keywords tab that splits data by the “Search Terms Match Type,” which allows you to see if a phrase or broad match keyword is generating exact, phrase or broad matched queries. Also, we can now segment location reporting by “Ad Type” (text ad, image ad, PLA ad and so on) when we download a geographic report from the dimensions tab. Surely other new data views that I haven’t found yet are lurking about, as well.

Big picture, we are getting a new layer of visibility into our AdWords performance and better tools for aggregating and reading those data in flexible ways. Some of these new data views change the need for certain account structures, such as the use of single keyword ad groups. Others eliminate “hacks,” such as pushing fake data to your “Total Conversion Value” column to distinguish leads from signups (a less than ideal setup that I’ve been guilty of more than once). The Reports Editor gives us a new, quick and flexible way to gather, present and optimize towards the new data and old data.

I’m pretty sure that 2015 has seen more updates to AdWords then any year prior (and I’ve been around for them all). Yet these new features suggest 2016 might have even more updates in store.

So put the final touches on your New Year’s Eve promotion site links and automate their scheduling. Then, take the last couple of weeks of this year to familiarize yourself with these new reporting capabilities, and spend some time considering how they will play into your 2016 strategy.

You may just end up saving yourself a bit of time every week, or you may finally unlock the data that allow for your next big growth push. Either way, hold on tight — These updates indicate much more will be coming next year.

The post The Never-Before-Seen AdWords Data Points That Have Been Quietly Rolled Out To Your Account appeared first on Search Engine Land.

Think Differently About RLSA & Boost Your Conversion Volume /think-differently-rlsa-boost-conversion-volume-233508 Mon, 19 Oct 2015 14:24:18 +0000 http:/?p=233508 How can you unlock the full potential of Remarketing Lists for Search Ads? Columnist Susan Waldes says the key is to consider audience intent.

The post Think Differently About RLSA & Boost Your Conversion Volume appeared first on Search Engine Land.


Remarketing is an important part of most AdWords account strategies these days. Aside from search (including Shopping campaigns), remarketing campaigns are probably the most utilized and successful campaign type for SEM advertisers.

RLSA (Remarketing Lists For Search Ads) has been around for a couple years, and in a way, it combines the best of both search and remarketing. You’d think that it would be extremely popular given the success of search and remarketing campaigns. Yet it’s very under-utilized in the accounts I audit.

Many accounts don’t use RLSA audiences at all, and most RLSA implementations that I do see simply layer RLSA audiences onto existing ad groups with a boosted bid modifier.

It’s a valid strategy, but the impact it has at boosting conversion volume is very tiny — you get a fraction higher click-through rates (due to higher positions) on a fraction of your traffic that may have clicked even without the bid boost.

To unlock the power of RLSAs in boosting your conversion volume, you need to start thinking differently about what these audiences mean.

The Value Of Remarketing Audiences

The Google Display Network (GDN)-based remarketing campaigns that we’ve all been running for years have gotten us stuck in a rut in terms of how we think about the value of a remarketing audience. We think of our GDN-targeted remarketing audiences like this:

“These people have been to my site, so they are considering my offerings. If I serve them my ad at the right time with the right offer, they are pretty likely to go ahead and convert.” 

Granted, that explanation is simplified, but even in reference to highly segmented, sophisticated remarketing campaigns, it still holds:

“These people have abandoned a very popular product in my shopping cart in the last two days. If I serve them my ad at the right time with the right offer, they are really, really, really likely to go ahead and convert.” 

Because we are used to thinking about GDN remarketing audiences in this way, we extend this thinking to RLSA audiences, valuing the audiences simply for their higher likelihood of converting. Higher conversion rates mean we can bid higher, so we toss on a bid modifier.

However, because RLSA audiences exist in the realm of search queries, they carry important information about intent that can be much more powerful than simply carrying a higher likelihood of a conversion.

RLSA & Intent

Set aside that remarketing audiences have higher conversion rates for a moment (though it’s true and makes the economics of RLSA campaigns work well for most advertisers).

Remarketing audiences are valuable in the search universe (i.e., RLSA) because you “know” things about that bucket of people and their intentions that you do not know about generic audiences. A couple of examples:

  • You have a travel site that exclusively sells cruise packages to the Caribbean. Sally visited your site last week. It’s fair to assume that Sally is considering cruising to the Caribbean.
  • Your site is B2B and sells restaurant supplies. Jack orders from you a few times a year. Jack must be a restaurant owner or some other kind of restaurant supply purchase decision-maker.

Why do these inferences about intent matter? Because keywords are also indicators of intent, and these audience targets can replace portions of your keywords.

In your normal keyword search campaigns, you have carefully chosen keywords that align specifically with what you offer. Via time and optimization, you have developed stringent match typing and negative keywords to home in on the highly-specific queries that carry all the right intent signals.

For the cruise example, your normal keywords will be “Jamaica cruises” and “cruise to USVI.” Now, think about what we know about Sally. We already know she is interested in cruising to the Caribbean.

When Sally searches for just “cruise,” we can target her, knowing that she is considering cruising to the Caribbean. When Sally searches for “USVI,” we can target her again, knowing that her interest in the islands is to visit them via cruise.

Our restaurant supply company wouldn’t target “dinner plates” in our general search campaigns, as only a tiny portion of those searchers have bulk commercial intent and can afford our 200 plate minimum order size.

But when Jack searches for “dinner plates,” we can afford to show up in position #1. Because of what we already know about Jack, might as well be, and that “dinner plates” keyword becomes high-intent, viable and valuable.

What’s especially exciting is that by advertising on “cruises” or “dinner plates” (albeit to a limited bucket of people), you are showing up on totally new, high-volume auctions, expanding your SEM reach and driving (often big) incremental conversion opportunities that aren’t covered by your “normal” campaigns.


When setting up these layered campaigns, make sure your ad group “flexible reach” setting is set to “Target and Bid” for interests and remarketing.

This is not the default setting, so you need to proactively change this setting on the ad group level to make sure you are only appearing on queries made by previous site visitors.


[Click to enlarge.]

There are several ways to execute on this kind of layered search + RLSA intent strategy that broaden your reach into new, valuable auctions. The simplest ones forgo the keyword all together.

Dynamic Search Ads (DSA) + RLSA

Dynamic Search campaigns allow Google to match queries to relevant content within your site. Without the RLSA layering, advertisers see mixed results from these campaigns; usually, complaints are that the query matching is too loose.

When you are only reaching people who have been to your site before via RLSA, Dynamic Search Ads are a very powerful tool to ensure you draw them back in, at scale, when they are searching for anything relevant to your offerings.

When we add an RLSA audience target to a DSA campaign in this account, we achieve CTRs 2.5x better and conversion rates 4x higher than with our generic DSA campaigns.

[Click to enlarge.]

You can start with just one ad group targeting all site content and layer an RLSA of all site visitors for the easiest and broadest reach. This will not only drive incremental conversion volume from a broad swath of highly targeted searchers but will also serve as a “discovery engine” for queries that returning users make in their path to purchase.

PLA (Shopping) + RLSA

RLSA functionality for PLA/Shopping campaigns was just broadly released about a month ago. Just like Dynamic Search Ads, Shopping campaigns match queries to relevant on-site content (products driven by a data feed).

When you layer an RLSA audience onto a Shopping campaign product group, you can skip the very segmented product grouping and negative keywords that you might use in your regular shopping campaigns. The RLSA audience itself serves as an intent signal, so even tangential queries or less popular products can see great performance.

Broad Keywords + RLSA

If you take a keyword-based approach to this campaign type, a simple way to begin is by just cloning your exact match campaign(s), changing the match types to broad and layering on an RLSA audience of all site visitors. You may be used to rolling your eyes at the very loose or tangential queries you see in your search term reporting when you use broad match.

However, with the intent already carried by your RLSA audience layer, those tangent queries are positive new opportunities to gain conversions.

Another way to discover good broader keywords is within your Google Analytics site search data. In the restaurant example above, I mentioned that might as well be

For that very reason, when we are talking about your RLSA audiences, Google can be used as an extension of your own site search to draw people back in to your products and services that they are searching for.


Final Thoughts

This strategy can drive big boosts in conversion volume within CPA or ROAS goals across many different verticals.

For e-commerce advertisers in particular, now is the perfect time to implement this strategy, as we are just about to enter the holiday season. The season itself is an additional boost to conversion intent: November and December drive 30 percent more e-commerce revenue than non-holiday months. This is also the first holiday season that RLSA audiences can be used on Shopping campaigns.

Effective use of this incredibly powerful combination of multiple intent signals based on prior site activity, current broadly targeted queries and a higher probability of converting due to the seasonality can boost your 2015 holiday SEM to new, record-high success.

The post Think Differently About RLSA & Boost Your Conversion Volume appeared first on Search Engine Land.

Gmail Ads: An Old Ad Type, But A New Addition To AdWords /gmail-ads-old-ad-type-new-addition-adwords-231228 Mon, 21 Sep 2015 15:09:37 +0000 http:/?p=231228 Not familiar with Gmail Ads? Columnist Susan Waldes has the scoop on the ad type formerly known as "Gmail Sponsored Promotions."

The post Gmail Ads: An Old Ad Type, But A New Addition To AdWords appeared first on Search Engine Land.


Gmail Ads have been around for nearly two years now in one form or another, though they used to be known as Gmail Sponsored Promotions (GSPs) and existed within a unique advertising platform. As of earlier this month, however, Gmail Ads have been integrated into AdWords.

Though the experience for consumers remains unchanged, the switch from the original GSP platform to the AdWords interface has brought about some notable changes for advertisers. SEMs who are already familiar with AdWords may find it an easy and intuitive transition, while others may be scrambling to figure out how to manage their Gmail Ads within a new system.

In my column on Marketing Land today, I discuss what’s changed (and what hasn’t) with Gmail Ads as a result of the AdWords integration. You’ll also find a guide detailing how to set up this new ad type within the AdWords interface. Take a look:

The post Gmail Ads: An Old Ad Type, But A New Addition To AdWords appeared first on Search Engine Land.

Because My Crystal Ball Is Broken And Other Reasons To Bring Broad Match Back In 2015 /crystal-ball-broken-reasons-bring-broad-match-back-2015-228219 /crystal-ball-broken-reasons-bring-broad-match-back-2015-228219#respond Mon, 24 Aug 2015 14:24:54 +0000 http:/?p=228219 Broad match has gotten a bit of a bad rap over the years, but columnist Susan Waldes explains why it could -- or at least, should -- be making a comeback.

The post Because My Crystal Ball Is Broken And Other Reasons To Bring Broad Match Back In 2015 appeared first on Search Engine Land.


If you’ve walked past a group of experienced search engine marketers discussing the industry in the last 15 years, you’ve surely heard something like this:

“When I took over the account, everything was broad match. A million dollars a year and all broad matched! It was so awful!”

For many years, broad match — or at least too much of it — has been considered cringe-worthy. It’s been the first mark of an un-optimized SEM account since match types were first introduced.

That’s why you can find thousands of articles and best practice guides that tout using exact, phrase and modified broad match to drive major performance improvements.

Too Many Accounts With Too Many “Best Practices”

Years and years of this focus on tight match types has led to a new problem: I now audit many more accounts that are struggling because of no broad match than ones that have too much of it.

The move to more exact matching, often without a holistic strategic compass (because its the supposed optimized way to do things), has put more advertisers head-to-head on each user query.

Ten years ago, there may have been 10 advertisers on broad match “online mba program.” The one or two advanced advertisers that used exact match [online mba program] would get an edge on the queries that were exactly matched to that keyword. Now, it’s more like eight out of 10 accounts competing on the exact match keyword, and that exact-match edge isn’t worth anything any more.

To simplify our discussion of match types, let’s throw out the concept of Quality Score and Ad Rank (though it is certainly a factor in real auctions). Without considering Quality Score, and with so much exact match, the only way to get top positions is to bid more. More exact matching is driving up CPCs on many keywords.

Changes In The Marketplace

In addition, Google’s move just one year ago to include close variants in exact matching has added to the premium costs associated with exact match. We talked about the eight advertisers that are all competing for that [online mba program] search. That is compounded now that the advertisers using the [online mba programs] keyword are part of that auction, too.

Prior to close variants, even in a very competitive auction with lots of exact match advertisers, you were still competing with a smaller subset of competitors. Factor in that each competitor would have a different approach to device, geography and time of day, and it’s easy to see how you could find pockets of traffic (certain devices, geos and times) when your bid got you high positions, even if it wasn’t  particularly competitive all of the time.

Now, the chance of winning within those “pockets” is smaller; close variants dramatically increased the number of advertisers eligible for each exact match auction.

Why Broad Match Could Be Key To The Performance Improvements You Need

I’ve spent many of my 16 years in this industry wrinkling my nose at broad match. Lately though, I love it! Match type choice should be a strategy, not a dogma.

If you aren’t using broad match, I encourage you to consider it in 2015. If you are using broad match in the right accounts, in the right way, you don’t need any convincing — the performance speaks for itself.

However, your boss, colleagues, clients and friends may be asking you why-oh-why you could support such a terrible match type. Here are five reasons you can fire back with:

1. My Crystal Ball Is Broken

Somewhere between 10 percent and 20 percent of  Google queries on a given day have never been seen before. Some of these are just momentary cultural memes due to news items and don’t align with advertiser intent (There are no paid ads on “target troll” at the time I am writing this). More common are just the individual and weird ways that the wacky world of individual people search for things.

To validate that the point above is applicable for paid search, I pulled search term (query) reporting from our MCC for the last year on a sub-set of accounts that I knew had very balanced match types. I looked at queries that only had one conversion and only had one impression — one-of-a-kind searches. The result?

30 percent of conversions in the last year came from queries that had only one single impression.

If you don’t have any broad match, you can only capture the portion of that 30 percent of conversions that you already predicted with your crystal ball.

2. I’m Following Google’s Lead

Over the last couple of years, Google has released new campaign types such as DSA (dynamic search ads) and Shopping campaigns that inherently skirt match types. Other changes, like close variants, pull back on the level of control an advertiser can have with match types.

The industry is always speculating that these moves mean the end of keywords altogether. When you are Google, and you have a huge batch of good direct-response-intent inventory (weird but good one-off searches) and nobody to buy (because everything is exact-matched), changing the rules is the best way to serve company profits, consumer desires and advertiser interests.

Google is like the parent who told us to be careful crossing the street. We are the child who responded by never leaving the house. Adding more broad match to your account in 2015 is… finally leaving the house and being careful crossing the street.

Shopping campaigns, DSA and close variants have improved performance and provided opportunities to grow profit for many advertisers. Would these campaign types even have been necessary if the industry hadn’t been overusing exact match?

3. I’m Focused On Mobile

Wait, you are focused on mobile, right? In terms of broad match, it’s particularly voice search to focus on.

The number of words per keyword has been declining for a decade or so; in 2015, voice search is driving a shift to longer queries again. If the long tail was dead, it’s coming back to life. Long live zombie long tail!

It’s not just mobile, either. Windows 10, just released, brings voice search to the forefront on desktop, too. We don’t know if this will change the way everybody searches, but there are undoubtedly missed opportunities to be had if you don’t have strong broad match keyword representation to capture those long tail, free-form voice search queries that will never make it to an auto-suggest or keyword tool.


You can’t even predict the things that do make it to an auto-suggest!

4. I Don’t Like Wasting Money

Broad match keywords are less expensive. We talked about the factors that are driving exact match CPCs up, and that is precisely why broad match CPCs are so affordable. With broad match, you open yourself up to more auctions and the pockets of available traffic at your target costs open up, as well.

5. I Am Into Growth (Maybe I’m Even A Growth Hacker!) 

When you have good broad match coverage, you discover all kinds of new keywords that convert, and you can use those findings to drive your account expansions and consistently grow.

I once took over an e-commerce SEM account of a women’s dress retailer. All this client sells is dresses. All keywords were exact or modified broad match and almost all keywords contained the word “dress” or “dresses.”

We opened up true broad match and discovered that there was a whole world of non-dress-containing keywords that converted. For instance, when an event implicitly was one you’d wear a dress to, people didn’t explicitly say “dress” — e.g., “outfit for a ball” or “what to wear to a wedding.”

Most accounts have similar keyword “seeds” or types that only emerge when you use broad matching. You can take the surprises you find and add them back into the account and consistently expand your converting keyword set.

Have The Naysayers Convinced?

If you have convinced the naysayers that broad match is needed to account for surprises, regional differences, efficiency, device changes, cultural changes, growth and the economics of the marketplace… congratulations!

One quick caveat: If you expand your use of broad match, make sure you are prepared to check your search term reporting (queries) quickly and frequently.

Some query reporting now comes through within hours, and most is there the next day (unlike the old 48-hour delay), so you no longer have to pay a “delay cost” before you can see what you may need to cut. The ability to get those negative keywords needs in “much closer” to real time is another reason why 2015 is the year you should be bringing broad match back.

The post Because My Crystal Ball Is Broken And Other Reasons To Bring Broad Match Back In 2015 appeared first on Search Engine Land.

/crystal-ball-broken-reasons-bring-broad-match-back-2015-228219/feed 0
Defeating The Dog Days: 5 Tips To Boost SEM Conversion Volume During The Summer Slump /defeating-dog-days-5-tips-boost-sem-conversion-volume-summer-slump-225860 /defeating-dog-days-5-tips-boost-sem-conversion-volume-summer-slump-225860#respond Mon, 27 Jul 2015 13:44:44 +0000 http:/?p=225860 Summer is a slow season in many industries -- which makes it the perfect time to review your PPC accounts and conduct some experiments, according to columnist Susan Waldes.

The post Defeating The Dog Days: 5 Tips To Boost SEM Conversion Volume During The Summer Slump appeared first on Search Engine Land.


If you happen to be in the sunglasses, summer camp or air conditioning repair business, congratulations on a great last few weeks for your SEM program. If you are the rest of us and feel like you have to swim upstream to meet your conversion volume goals lately, take comfort that we all are suffering from the “summer slump” right along with you. As temperatures rise, search volumes fall.

The summer means people are out and about. Kids are out of school, workers are on vacation, and everybody spends more time on leisure activities and less in front of screens. This means there is just less searching across almost all verticals.

Search demand will remain slow for most industries until the second week of August, when most verticals will start seeing an uptick. Though we can’t change this seasonal phenomenon, there are some proactive optimizations that will give your low conversion numbers a boost and help you make the most out of the summer slump.

Here are five tips to juice a bit more out of your SEM program while everybody else melts in the heat.

1. Try Enhanced CPC

In your campaign settings, test turning on “Enhanced CPC.” This will allow Google to increase your bid up to 30% more than your actual bid when the algorithm determines a user has a strong probability of converting. It might not be a setting that you use year-round, but when impression volume is at its lowest, enhanced CPC can mean snatching important incremental conversions from your competition.

2. Review Your Negative Keywords

Some negative keywords are “forever.” If you don’t sell blue widgets and never will, then “blue” should remain a negative in your widget shop account.

However, in a mature account, other negative keywords are added for performance reasons due to seasonal inventory, or sometimes by mistake. A period of undesirable performance on a certain keyword phrase may have been years ago, for a couple months, and with a different offer or landing page from the one you currently use.

Periodically, negatives should be vetted, and there is no better time than when you need a volume boost. Aside from removing any negatives that no longer apply, negatives that have been added purely for performance during a limited period of time should be pulled off and retested.

(Tip: Use shared campaign negative lists where possible to group negative keywords, with a “name” for the list that indicates why they are there and if they should be removed or reviewed at some point. )


3. Try More Broad Matching

Tight match types are often the hallmark of a mature and well-managed PPC account, but all too often there is so much focus on exact, phrase and modified broad match that regular broad match gets cut all together.

Broad match is not something to be afraid of. Granted, the queries are more diverse and less “sure” in terms of consistent conversion rates, so bid lower. However, if you don’t have any broad match (or have very little), you are missing out on significant conversion volumes.

Adding more broad match keywords can give an instant boost to your conversion volume and, secondarily, help new keyword discovery and account growth via matching to queries that you hadn’t thought of before.

4. Filter For Below First Page

On your keyword tab, choose the “Status” filter, and uncheck everything except “Below First Page Bid” under “Eligible: limited.”


You’ll see a list of keywords that are below the first page bid estimate. To get even better data, select a long date range and add another filter for keywords that have had significant conversion volume during that long time period.

Formulaic manual or automated bidding systems will lower a bid sometimes after only a short period of poor performance. If a bid gets lowered below the first page, it may never get enough impressions to garner the satisfactory performance data which pushes the bid back up again. In effect, you can “kill” good keywords over time via bidding automation when they fall off the first page.

Sometimes, bumping the bid by a couple of cents will get these keywords showing again without having any particular impact on cost per action (CPA). Download your filtered list and create a column that calculates the bid increase needed to get to the first page, and test bumping bids up in tiers.


5. Experiment With Your Bidding

Want to know what would happen if you doubled your bid on your top keywords without destroying your overall CPA metrics? You can.

Set an experiment at the campaign settings level. Depending on how conservative you want to be, you can only allow as little as 10% of your impressions on the target keywords to participate.


Choose a set of test keywords (or even just one very important keyword) — perhaps those that had higher conversion volume in the past, or keywords that never got much activity but you think should work well. Set a high CPC multiplier on the “experimental bid” sub-tab in AdWords Editor on your chosen keywords.


As data accrues, you can segment for “experiment” in the Adwords UI and see how the higher bid impacts position, CTR, conversion rate and CPA on the small portion of traffic you are testing. With only 10% of traffic in a certain campaign participating, even if CPA gets uncomfortably high on this bit of traffic, you won’t destroy your overall metrics.

On the other hand, you might find that the higher positions lead to higher conversion rates and significantly more profit. If the experiment wins, you can  bid higher on 100% of the traffic moving forward and increase your total conversions significantly on an ongoing basis.

Final Thoughts

Don’t sit around passively waiting for the summer slump to end to start growing your conversion volumes. Take a proactive stance and start testing with these tips. You’ll likely find some techniques to boost your conversion volumes when search is seasonally slow. At the very least, you’ll have some great data that shows higher-ups that you’ve tried everything, and the summer slump is unequivocally “a thing.”

The post Defeating The Dog Days: 5 Tips To Boost SEM Conversion Volume During The Summer Slump appeared first on Search Engine Land.

/defeating-dog-days-5-tips-boost-sem-conversion-volume-summer-slump-225860/feed 0
Use Google Analytics To Stop Wasting Remarketing Dollars On People That “Just Aren’t That Into You” /use-google-analytics-stop-wasting-remarketing-dollars-people-just-arent-223759 /use-google-analytics-stop-wasting-remarketing-dollars-people-just-arent-223759#respond Mon, 29 Jun 2015 13:43:47 +0000 http:/?p=223759 Let's face it -- some of your website visitors just aren't interested in what you have to offer. Columnist Susan Waldes has some tips on how to stop spending money on people who will never buy from you.

The post Use Google Analytics To Stop Wasting Remarketing Dollars On People That “Just Aren’t That Into You” appeared first on Search Engine Land.


Remarketing is a powerful way to gain incremental conversions. After all, there is no better audience for buying your product than the one that is already familiar with it. Your visitors are warm prospects, and your remarketing pushes them right down the funnel.

But wait, is that always true? Is every person who visited your site actually in-market for your product? Or, are some of those site visitors just people with “fat fingers” that accidentally clicked a link on their tablet? Was that surge of traffic from Tumblr actually people interested in buying insurance, or was it curiosity clicks from a humorous take-down of your stock photo choices?

Face it — a chunk of your website visitors “Just Aren’t That Into You.” Coming to terms with this can be tough on the ego. Even tougher, though, is realizing that your remarketing campaigns are wasting dollars and impressions trying to reengage these people who don’t really care in the first place. Ouch.

I’ll share three strategies, using Google Analytics, that insure your remarketing efforts are actually targeting people who are potential customers and not wasting impressions on the visits that “Just Aren’t That Into You.” Your click-through rate (CTR) will increase when you aren’t endlessly pestering disinterested users with wasted impressions, and Google’s algorithm will serve your ad more frequently, on more premium placements, to the users that are interested in your offer.

Check Your Content

I frequently see remarketing campaigns that are targeting an audience of all site visitors over the last 30 days. This can be a decent starting point for a remarketing campaign, but even without looking at data, you can surely define some parts of your site that don’t warrant inclusion.

Careers pages are an obvious one. Contact pages which attract current customers are another one that should often be excluded. To find the less obvious pages, navigate in Google Analytics to the “Behavior” category, then drill down to “Site Content” > “Landing Pages.”

Filter for landing pages that have high session numbers but no conversions. You’ll discover content that is driving users for reasons other than purchase intent. Maybe one of your blog entries went viral. That’s great for your SEO, but all those people aren’t qualified targets for your remarketing.

For the non-converting content that you discover, create remarketing audiences (in AdWords or Analytics) of people who visited those pages and exclude those audiences from your remarketing campaigns.

Exclude “Barely There” Visits

Navigate to the “Audience” category  in Google Analytics, then select the subcategory of “Behavior.” You’ll find a report called “Engagement.” There, you’ll see a tiered report breaking down session duration (or time on site). On most sites, you’ll see that close to 50% (or more) of sessions are less than 10 seconds.



These <10 visitors are not good prospects for driving future conversions. You may have spent marketing dollars already on some of these initial visits, and the last thing you want to do is incur more costs trying to re-engage users that “Just Aren’t That Into You.”

To eliminate these people from your remarketing campaign target, create a Remarketing list in Google Analytics for sessions <10 seconds.


The list will appear in your AdWords account within a day and can then be added as a negative audience to your remarketing campaigns.

Leverage Affinity & In-Market Audiences

People visit websites for a myriad of reasons aside from simply being “in market” for a certain product or service. For instance, as a PPC professional you likely spend lots of time browsing client and competitor websites without any intention to purchase summer dresses or health insurance or whatever vertical you work in.

How can you separate the truly interested people from the looky-loos? Go to the “All Campaigns” report under Acquisition > Campaigns and filter for your remarketing campaigns. Then, set your Secondary Dimension to “Affinity Category (reach).”


These affinity categories map exactly to the affinity audience targets within Adwords. They will not always illuminate exactly why these people came to your site; after all, there are not affinity categories for “Digital Marketing Professionals Doing Competitor Research.” However, they do “bucket” people in ways that allow you to slice out most of the “Just Aren’t That Into You” users.

I’ve worked with a few clients that have products targeted to parents, or just a demographic that contains lots of parents. These advertisers have consistent problems with little kids clicking on their ads when mom or dad hands over the tablet. If this problem impacts you, you’ll likely see lots of clicks, and little to no conversions, coming from the “Comics & Animation Fans” and/or “Avid Gamers” affinity categories. Add them as negative audiences and voilà, no more wasting ad dollars on 3-foot tall Sesame Street fans.

You can also utilize the same strategy with “In-Market Segments.” The process is the same except when you choose your secondary dimension, choose “In-Market Segment.” Depending on your product and vertical, you’ll usually find one or the other illuminates the undesirable audience segments.

If you haven’t actually launched remarketing campaigns yet, you can look at these affinity categories and in-market segments as a secondary dimension on all your traffic within the “Acquisition” > “All Traffic” > “Channels” report to identify bad audiences before you launch. If you do this, you’ll still want to check your actual remarketing data periodically, once launched, as remarketing can “magnify” bad audiences by continuing to reengage and cookie users over and over that never really cared in the first place. 

Take Control & Start Negating

Hopefully these tips get you on the path to take control and stop pining after sales from users that “Just Aren’t That Into You.” Eliminating these people will have a positive impact on your remarketing performance results by eliminating wasted spend, increasing your CTR and gaining algorithmic favor with Google.

Once you start tactically thinking about your audiences as prospects and non-prospects, and see the performance gains, you’ll likely discover your own reporting dimensions and techniques for finding more unqualified audiences to cut from your campaigns. You can also find more tips for cutting out junk from your remarketing campaigns in this deck I presented at SMX Advanced in early June.

The post Use Google Analytics To Stop Wasting Remarketing Dollars On People That “Just Aren’t That Into You” appeared first on Search Engine Land.

/use-google-analytics-stop-wasting-remarketing-dollars-people-just-arent-223759/feed 0
Sneak Peek Into AdWords’ New Multidimensional Reports /sneak-peek-adwords-new-multidimensional-reports-219982 /sneak-peek-adwords-new-multidimensional-reports-219982#respond Mon, 04 May 2015 14:05:18 +0000 http:/?p=219982 Google has begun rolling out its new Report Editor to AdWords accounts, but very few advertisers have access to this feature yet. Luckily, columnist Susan Waldes is among them and has decided to share her findings here!

The post Sneak Peek Into AdWords’ New Multidimensional Reports appeared first on Search Engine Land.


Just over a year ago (April 22, 2014), during a livestream about upcoming AdWords features, Google announced it would be rolling out advanced reporting for AdWords. The accompanying blog post stated:

Advanced reporting: To help you analyze your data better (without the endless downloading and re-formatting of data) we’re providing you with new multi-dimensional data analysis and visualization tools so you can perform most, if not all of your data analysis, right here inside AdWords. We’re also making it easy for you to turn your data into tables, graphs and charts so that you can download them and share with your teams.

This announcement generated a good deal of excitement from the SEM community. With a whole year come and gone and the buzz quieting to a whisper, Google’s timing of that announcement, from the perspective of achieving optimal PR, can certainly be questioned.

Currently, the new functionality is starting to roll out to a very limited number of accounts. Of the approximately 100 accounts that I have access to, I’m only seeing the new “Reports Editor” in one. (The account also happens to be completely shut off at the moment.)

If you have the new Report Editor, you’ll see a “Reports” link on the top navigation between “Campaigns” and “Opportunities.” If not, the link to “Reports” is on the left navigation sidebar.

A couple weeks ago when I discovered the new reports, the update box on the Help Center article about the reporting feature said it would be rolling out to all accounts in April and May. However, as of today, that box says it will roll out “over the next few months.” It appears that whatever obstacles put a year between announcement and launch may also be delaying launch even further.

So while SEMs continue patiently waiting to get access to the Reports Editor, I thought I’d share some sneak peeks at the new functionality.

The Visual Charts

The Reports Editor can create on-the-fly line, bar and pie charts. Bar and pie charts are completely new to AdWords. Line charts have an updated look and much more comprehensive functionality.

I’m not typically a fan of pie charts for SEM metrics. The data reduction required doesn’t give a holistic view of the account reality. However, occasionally there is a use for pie charts, like this quick visual of how impressions are distributed across match types:


Line charts have always been available within the AdWords UI; the upcoming updates both make the data more presentable as well as introduce much richer ways to look at data.

Formerly, time was the only choice for the X-axis in a line chart. The visual chart was very thin and horizontal, making it less than ideal for using in client presentations. Additionally, a maximum of two trend lines on a limited set of pure metric data was available for charting.

The old line chart in Adwords

The old line chart in Adwords [CLICK TO ENLARGE]

With the new Reports Editor, line charts are much more visually presentable. They are still based on time as the primary X-axis, which is controlled by a drop-down at the top of the report.

You can still only have two metrics trended over time. However, there is the ability to add a data segment you are trending. For instance, you can set your segment to “country” and trend conversion volume monthly with each country as a line (though only top results will appear in the chart to a maximum of 16 lines). Or, you can trend the three match types (each with its own line) week over week:


The bar charts are another way to present data over time. They also allow many other ways to use your X-axis. For bar charts, your initial segment can either be time or something else like device, click type, or top vs. side (data points currently accessed through segment reporting).

You have the ability to chart bars for two metrics within your chosen segment. For instance, here I am looking at account-wide conversions and Cost/Conv by device:


These new, visual ways of looking at data are useful for client reporting and presentations. Surely, it will also be easier to hone in on new problem or focus areas within an account that emerge more clearly with these new visual tools.

The Data Tables

The new data tables will be much more frequently used than the charts. They represent improvements to the current reporting capabilities in a few ways:

  • They are created quickly and “on-the-fly” within the reporting system. Ever set up a complex report, downloaded it, opened it in Excel, then realized you forgot one column or segment and you had to do it all over again? That won’t happen anymore, as the data is built right in the UI.
  • “Segments” are presented as columns rather than indented rows in the UI, which makes comparison and sorting less cumbersome.
  • The data presented can be filtered regardless of whether or not the filter is also a column in your report.
  • There are some entirely new available data points.
  • Time periods on saved reports can be changed on the fly rather than having to “regenerate” data in each time period.

For instance, this report shows a totally new data point (conversions metrics on search queries). It shows a segment as a column rather than as an indented row (the “conversion action names” column) and I’ve filtered it to only show this data for mobile devices. Formerly this kind of filtering could only be done by including all devices as a segment, then downloading and extracting only the mobile data in Excel:

account-sensitive data obscured

[CLICK TO ENLARGE] – Account-sensitive data obscured.

The Impact

The most interesting reports to me are those that tease at new ways to look at data that will lead to new techniques, best practices and optimizations.

For instance, looking at mobile conversion volume by hour-of-day and day-of-week, I can quickly see that Wednesday mornings are key for driving mobile conversion volume for this particular advertiser.



It’s data that I could have extracted from the current reporting capabilities, but I’d have had to “know” that I wanted to look at data in this manner, then go through a complex process of report building, downloading, pivoting, and filtering in Excel to get to the point where I could see this.

In this new system, I just pulled in some columns and filtered in about five seconds; simply “poking around” drove this new insight. Surely, there are multitudes of similar insights that will be facilitated by this new reporting. Some of them will be very account-specific, and others are sure to lead to new industry practices.

I do not advise holding your breath while waiting for this (already very delayed) feature to roll out globally.

I do advise reading more about the capabilities and starting to think about how your reporting and data analysis can be improved in this new environment. In addition, as this rolls out to more accounts, stay on alert for new techniques that will surely emerge, facilitated by new ways of looking at data, which can meaningfully improve your results.

The post Sneak Peek Into AdWords’ New Multidimensional Reports appeared first on Search Engine Land.

/sneak-peek-adwords-new-multidimensional-reports-219982/feed 0
Everything Search Marketers Need To Know About Gmail Sponsored Promotions, Part 1 /gmail-sponsored-promotions-everything-need-know-succeed-direct-response-gsp-part-1-215816 Tue, 10 Mar 2015 17:49:47 +0000 http:/?p=215816 GSPs are now available to all advertisers. Columnist Susan Waldes has been there since the beta launch two years ago and shares her hard-won best practices.

The post Everything Search Marketers Need To Know About Gmail Sponsored Promotions, Part 1 appeared first on Search Engine Land.


They’re not quite AdWords, and not quite email marketing either. So, what are Gmail Sponsored Promotions? They’re very special beasts.

This Google offering, which allows marketers to place ads within the Gmail interface, has been around for about two years as a Google beta. I was fortunate to get access to this earliest beta test and have been using the platform ever since. It’s gone through many changes since those early days, and last month Google officially opened up GSP to any advertiser.

Given all the interest and availability, I decided it was time to share my experiences with GSP. I’ve had the opportunity to run campaigns for a dozen different advertisers, and that’s given me a good chance to develop some solid best practices.

Though it’s not really search marketing, I suspect AdWords aficionados may be interested in exploring the opportunities, so, though the two-part series is running on Marketing Land, I wanted to make sure paid search folks were aware so they could check it out over on Marketing Land.

The post Everything Search Marketers Need To Know About Gmail Sponsored Promotions, Part 1 appeared first on Search Engine Land.

Rip Off The Band-Aid: Use AdWords Conversion Tracking For (All) Your True Business Goals /rip-off-band-aid-use-adwords-conversion-tracking-true-business-goals-214038 /rip-off-band-aid-use-adwords-conversion-tracking-true-business-goals-214038#respond Mon, 09 Feb 2015 14:24:43 +0000 http:/?p=214038 AdWords has made many improvements to its conversion tracking capabilities in recent years, but columnist Susan Waldes notes that many businesses have been reluctant to catch up.

The post Rip Off The Band-Aid: Use AdWords Conversion Tracking For (All) Your True Business Goals appeared first on Search Engine Land.


Over the last year, Google AdWords has massively improved conversion tracking capabilities, reporting on conversion metrics and the automated bidding functionality that they drive.

Yet, over the last few months, one of the most common problems I see in faltering AdWords accounts I audit is companies that have a clear business goal  (or goals) for their paid search spend, yet are tracking other goals as conversions in their AdWords account.

Sometimes, this is because there is a lack of knowledge on the improvements that have been made; other times, it’s more about institutional complacency — people get used to the “issues” that existed years ago when initially setting up their conversion tracking and they fail to see that there are now tools to resolve them.

Changing your conversion tracking does have some costs associated with it. There is the development time, the loss of apples to apples year-over-year comparisons, re-development of reporting systems to work with new data points, and usually some time invested in “re-teaching” other parts of the organization what a “paid search conversion” means.

However, I’ve yet to see a case where ripping off the band-aid isn’t the right thing to do. Tracking your actual business goals as conversions within AdWords gives you immediate, quantifiable advantages that far (and quickly) exceed the costs and pain points. Your spend will become more efficient almost immediately, and a world of new, important data and features opens up.

Conversion Types

First, let’s quickly review the types of conversion tracking that are now available in the AdWords interface.

  • Web Page Conversions. These are the “traditional” conversions that the majority of accounts use. A pixel is fired when a certain page (usually a “thank you” page) is loaded. Tweaks to the code can also be made to fire this pixel when certain site areas are clicked or certain other events occur. (Tip: look into Google Tag Manager for a free solution to help with tracking atypical “conversion” scenarios.)
  • Mobile & Tablet App Downloads And In-App Actions. Your Google Play Store app downloads and in-app actions such as upgrades and in-app purchases are simple to track as AdWords conversions. iOS app tracking takes a bit more heavy lifting, but can be done.
  • Phone Calls. You can track raw phone call numbers from your call extensions, mobile landing pages and desktop landing pages. Each is set up as their own conversion type and can be given a duration (default 60 seconds) that is reported as a conversion in your reporting.
  • Google Analytics Goals. Any goal that you have in your Google Analytics can be brought into your AdWords conversion reporting if your accounts are properly linked. If things like session duration or multiple page views are important to your business, which are not associated with a specific page loading, you can still measure these business goals against your AdWords spend.
  • Imported Conversions. Any conversions that are internally tracked in your database can be imported and integrated into your AdWords reporting as long as you capture the gclid in your database rows. (Tip: It’s often hidden in a landing page URL row and can be extracted with a bit of Excel manipulation.)

Those cover the major goals of most businesses. These newer conversion types are important, but more notable over the last year is that there are a host of new conversion tracking features and choices.

Refining Your Conversions

Conversions for most businesses are typically more complex than just recording a “1.” Until more recently, though, Google did not allow marketers to customize many of those nuances.

AdWords has reached a level of maturity with the available conversion tracking features. There are tools that make tracking multiple goals possible and allow parity between how your conversions are recorded in AdWords and how your business as a whole measures the value and importance of each goal.

  • Assigning Revenue: Not new, but underutilized, is the ability to capture actual transaction values for any web-based or imported conversion types. If you are able to associate revenue to the conversion action that generated it, you should be bringing this metric into your AdWords account.

    For “conversions” that do not have literal revenue attached to them, such as a social media follow, you can assign a static value. If your business as a whole values a “Facebook Like” (for instance) at $1, you can track this goal (alongside others) within your account and show in more granularity how your AdWords spend is contributing to growing the business.

    More recent updates have rolled out a “value” assignment for all conversion types, including phone calls.


  • Conversion Cookie Length Control:  There was a (very long) time when AdWords conversions all operated on a 30-day cookie. Functionality to control the conversion pixel window started rolling out in late 2013. The earliest iteration had limitations on the types of conversions where you could control this tracking and only a few set length choices.

    With more recent updates, advertisers can choose the conversion window (in days) without limitations and for every conversion type. This allows advertisers to control the attribution of their AdWords spend to window that makes sense for their products or for each conversion type.

    For instance, if you track a “shallow” conversion such as a newsletter sign-up, you might only want to give AdWords credit for a day for that. However, if you also sell enterprise B2B software that costs 4 or 5 figures, typically a product that would have a long consideration period, a 60 or 90 day window might make sense for that transaction’s conversion.


  • All Or Unique Conversions: This rollout began in February 2014 with the release of the (somewhat awkward sounding) converted clicks vs. conversions reporting. After the roll-out of a couple updates, you can now choose whether each conversion type counts “all conversions” or “unique conversions.”

    For instance, for any conversion type that is more of a “lead” or “user acquisition,” you probably want to count only unique conversions; multiple conversions are usually duplicates of some sort. On the other hand, for transaction or purchase based conversions, you may want to count all conversions, the user who keeps coming back and purchasing several times, is highly valued and your reporting can account for that full value.

  • Bid Optimization: Open up the “advanced” section in your conversion creation process and you get the choice, for each conversion, to determine whether it is included in Google’s automated bidding calculations. Those bidding types currently include ROAS bidding, CPA bidding (target and max) and enhanced CPC bidding.

    For instance, say you are a SaaS freemium model software product, you can count your “free trials” as conversions, with no revenue associated, and not factored into your bidding. As another conversion, you can count your paid subscriptions, track that revenue and use ROAS bidding or CPA bidding (CPA meaning a cost per paid subscription, in this case) only on the conversion type that is truly driving revenue.

Conversion Reporting

The most recent and final piece of the puzzle that makes these new features truly usable is the associated reporting. Google has new columns that allow you to segment these conversion types, and all their associated metrics with true granularity.

  • Phone Call Reporting: The “call details “ set of columns within AdWords allows you to see how many users saw your call extension, how many raw phone calls your call conversion types have generated and how many on-page call conversions you’ve generated unique conversions and all conversions.
  • Unique Vs. Total Conversions: The “conversion” set of columns has been updated to include “conversions per click” vs. “total conversions” metrics that split out your unique conversions from all tracked conversions.
  • Conversions (Optimized): The “conversion” set of columns has also been updated to include the “conv (opt)” columns that report only on the conversions and associate metrics that you have set-up for inclusion in automated bidding strategies.
  • Custom Columns: The new custom columns set of columns, release in Dec 2014 (I fall into the this is a BIG deal camp for this very reason), allow you to show each different conversion name and all associated metrics in its own set of columns. You can break out your call extension calls, your calls from your website, your web leads etc, each in their own set of columns and truly see the full picture of the activity your spend is generating.


Making Your Spend Work For You

I frequently hear advertisers say things like, “Our AdWords account is measuring leads as conversions, but we would like to bid to a profitable cost per sale,” or, “We round up our conversions by 10% to account for the phone calls from the number on the website.”

These kind of proxies and assumptions can be useful when tools are not available to show the true numbers (the case for most of the last decade). Many advertisers also have the real numbers in other systems and periodically match them back to AdWords data as best they can.  However, a world of very granular reporting and optimization opportunities open up when the true measure of success are in-line with spend data and reporting segments that you cannot get outside of AdWords.

For the “we track leads but care about sales” advertiser, think about if you had the sales and revenue data matched up to a query report. You may discover that you can negate a query that drove half your leads on a certain query without a sale. Once that query is negated, you can double your bid on the keyword that is now twice as efficient.

For the “we round up for phone calls” advertiser, tracking real calls can mean discovering that 90% of the phone calls, previously spread to all campaigns at a 10% round-up, are actually being driven by your brand campaigns, on mobile devices during the times you are running TV commercials. Then, set up rules and structure to optimally capitalize on that finding.

Give It Some Deep Thought

It’s the ideal time to consider whether your conversion tracking is really working for you optimally. Invest the time and make the needed adjustments to maximize your ability to grow your account, quantify your spend and align with broader business goals.

The constraints of the past have been largely removed and the benefits are immediately quantifiable in terms of better account efficiencies. The danger of complacency is your competitors utilizing these new features first and gaining market share on the metrics that matter – and would you even notice it happening if you aren’t tracking the metrics that matter?

The post Rip Off The Band-Aid: Use AdWords Conversion Tracking For (All) Your True Business Goals appeared first on Search Engine Land.

/rip-off-band-aid-use-adwords-conversion-tracking-true-business-goals-214038/feed 0