Latest Intelligence From the Field

Setup An Awesome Adwords Campaign In 6 Steps

Many of our blog posts can get very advanced, but today we wanted to take it back to the basics of best Google Adwords practices.

When push comes to shove, the best functioning Adwords campaigns are not built on great ad copy, stellar keyword choices or obnoxious keyword bidding, they’re built on a great setup. Before you setup your first Adwords campaigns, take serious note of where, who and when your customers want your business. Additionally, consider when, who and how you want your customers to come to your website.

Even if you’ve been doing Adwords for years, it’s always good to double check your settings against best practices to make sure you’re utilizing everything Google offers on their ad platform. If you don’t check often enough, you might be surprised to see a lot of updates on a month-to-month basis in the settings tab.

After creating a new Adwords account, the next step is setting up your campaign. The most prominent features in campaign creation are deciding what kind of campaign should be made, location targeting, budget and bidding strategy.

1. Search vs. Display vs. PLA vs. Combination

The easiest decision might be the decision between search, display, product listing ads (PLA) or a combination of search and display. Generally speaking, you’ll want to choose one or the other. In fact, we rarely recommend creating a “Search & Display” campaign. Choosing one specific function for each campaign will allow you to better focus and, thus, better optimize your campaigns based on searcher trend.

Search Network:

For text ads on Google’s search network. You can chose between “standard” search or “all features.” Standard only offers a few targeting methods while “all features” will expand your settings to specifications like ad rotation, ad scheduling and more. When in doubt, chose “all features.”

You can also choose PLAs from the search campaign menu. For a PLA campaign, you will need to have a csv or xml of your products ready to upload. For more information on how to prepare for a PLA campaign, view Google’s information:

Lastly, you can choose to add dynamic search ads to your campaigns. If you choose this option, you can also create regular search ads to split test with dynamic search ads. Dynamic search ads are recommended if you have a lot of products on your site with various landing pages. Essentially, you write a generic text ad for your products and Google uses its super powers to match the searcher’s keyword with the most appropriate landing page or product page on your domain. We haven’t found significant results from these ads, so we would recommend PLAs if you want to advertise by a product’s view page.

Display Network:

Primarily for image ads on Google’s display network. Here you can make distinctions between “all features,” mobile-only, retargeting or engagement displays. If this is your first display campaign, we recommend choosing all features. If you are creating a retargeting campaign, though, be sure to distinguish this here.

2. Location

Location targeting isn’t just for choosing where you want your product advertised. You can also use it to geographically target different places with different ads, increasing the relevance of your ads and potentially increasing click-through rates (CTR).

3. Budget & Bidding Strategy

The last important feature here is deciding how big your budget is, and choosing a bidding strategy. Picking a budget should be the simple part, but deciding how much to bid will vary depending on what you’re trying to advertise and the amount of competition there is for your keywords.

Your bid can also be changed individually for each keyword, so you can bid more or less depending on each keyword’s performance, you aren’t locked to only setting the bids at the campaign level.

4. Devices

Devices – desktops/laptops, tablets, and mobile devices – used to be a part of the campaign setup until Google realized it was important enough to edit more closely. These settings have been moved to the ad group level settings since enhanced campaigns were introduced. Now, instead of having to create separate campaigns for mobile and desktop ads, they are all tracked separately for you in the same campaign. When you create new ads, you can choose if you want your ad to be focused on “mobile,” or “standard,” which will give it preference on desktops/laptops and tablets. You can also segment your budget by percentage between desktop, tablet and mobile in this view.

5. Ad Group

Once you have your campaign set up, ad groups are the next step. They allow you to keep keyword groups and ads separate within the same campaign settings. The key to having effective ad groups, and overall an effective account, is keeping the keywords in an ad group focused on a particular phrase or subject. Typically, it’s more effective to have multiple ad groups, each with their own focus and keywords, than it is to lump all of the keywords and ads into the same ad group.

Making sure your ads are focused and condensed into only the most relevant keywords will have a positive effect on quality score. Quality scores are measured by Google based on how relevant your ads, keywords and landing pages are to the terms that searchers use to find your ads. Quality score is one of the factors that determines how much you actually pay for clicks. A higher quality score will allow you to show in a higher position for lower costs. Since quality score is calculated based on not only your keywords, but also the contents of your ad and landing page, it’s important to make sure all three of these are closely related. Lastly, if you decide to use a phone number or sitelink ad extension, Google just announced these affect quality scores too now.

6. Ad Copy

As far as ad text goes, it is usually a good bet to have the main keywords once in the headline, once in the body, and once in the Display URL. The nice thing about the Display URLs is, as long as the main domain name is the same as your destination URL, you can add the keywords in after the URL for extra keyword relevance (see below screenshot for a search ad example of this). Display URLs work the exact same way but have a character limit of 35, so including the main keywords after our domain isn’t always possible.

Another small tip for improving your ad text: Capitalize the first letter of each word in the ad, this is more likely to catch the eye of someone skimming the page. Below is an example of ad copy Vuurr might write for keywords like, “digital marketing,” “data driven marketing,” and “site development marketing.”


These 6 steps will get you started on a well constructed and functioning Adwords campaign. Have questions we didn’t cover? Comment below!

Vuurr Turns 2, Opens Office in San Francisco

On this day two years ago, a team of four united (much like Voltron) and amassed to what is now known as Vuurr Digital Marketing consultancy. Today, June 13, 2013, celebrates two years of building and expanding metric-driven digital marketing campaigns for our clients.

With this occasion, we’re also taking the opportunity to announce the opening of our new satellite office located in the SoMa neighborhood of San Francisco, CA. We have a number of clients and partners in San Francisco already, and the opportunities for expansion and networking are obvious here.

In two years we have grown from just four partners, to a team of over ten with an annual revenue of over $1 million for 2013. Our current client portfolio includes companies such as Cox Communications, Cold Stone, MIT, Infusionsoft and PandoDaily. We’re looking forward to building up our portfolio and opening up more product partnerships in the bay area.

Two years isn’t a long time in the grand scheme of things but we’re proud of our team’s accomplishments. We’d like to thank our clients – without you we are nothing. And a thank you to our team, friends, and families who continue to support us along the way.

There will be changes; but we’re not going anywhere. Tally ho.

Copywriting for Conversions

You’ve probably noticed the articles, blogs or retail copy you prefer reads more like talking to your best friend and less like talking to a robot. Gone are the days of strict AP Style or Chicago Style grammar rules. In are the wily thoughts of creative writing, boastful copy and enticing new words that aren’t really words but for some reason still make sense in a trendy way.

And although I’m loving the freedom to take on a project with my real voice, there are still some key ideas to keep in mind when writing for conversion. Here, many copywriters might still argue with me and say there’s no way to put a value sign on a piece of copy – they’re wrong. In fact, if your copywriters don’t know how to track the effectiveness in their copy, it’s time they get to reading up on A/B testing and SEO. This post won’t make you a genius, but here are three things you want from your copy now.

1. Write Why

I’m reading a great book right now by Simon Sinek called Start With Why. After feeling inspired by his TED talk, I immediately bought his book. According to Sinek, just like there are only seven story plots, there are really only 8 ways to convert customers: Manipulation, Price, Promotion, Fear, Aspiration, Peer Pressure, Novelty and Why.

The first 7 copy strategies do work; however, these approaches are less likely to create loyal customers or inspire your customers to use your product. You want customers to be loyal because you always want them to come back, you always want them to seek you first with a problem or a need, and you always want them to recommend you first to their friends. You want customers to be inspired by you because you want them to believe your product makes their business/family/friends/sex/life better. When these two things happen, your customers belong to your brand so wholeheartedly that price, promotions and novelty don’t matter, nor will they feel fear before or after buying because they trust you.

When you’re brainstorming copy, be very sure you know why it needs to be written. Remember there are hundreds of thousands of words on the internet that will match yours, so make sure it’s ordered to answer the question “Why?”

2. Mindful Keyword Insertion

After you’ve done your keyword research, mindfully read through your drafts and insert keywords before finalizing copy. A couple keywords or iterations of each are probably already written throughout, so edit and reframe sentences to better highlight and explain the use of each keyword.

I call this “mindful” keyword insertion and not the more familiar term “keyword stuffing” because if you are actually stuffing keywords, your sentences are going to convert to total crap. I know you know those sentences… they’re high and mighty, and if you don’t have time to read them three times slowly, your readers certainly don’t have the patience to read them either.

Only use keywords for clarity or your readers will sniff your bs and forget you existed.

3. Split Test

Here’s a likely scenario: You like two titles or two sentences or two wireframes. You could choose, but you shouldn’t and you don’t have to anymore. A/B and multivariate testing software have made choosing the best of two or three or hundreds easier than jumping rope. Submit your best ideas and watch your very own customers choose for you. We like Optimizely because it’s intuitive and your customization choices are almost endless, but Adobe and Kissmetrics also have great products. When your test has achieved statistically significant user data, make permanent changes to reflect the copy your users think is best.

jQuery Diagonal Grid Fill Algorithm

This week, the team mulled over a compelling code challenge: “How do you fill a grid in diagonal columns starting at the top left corner?”

As a person who legitimately takes pleasure in solving trivial puzzles, I approached the problem head-on. And although easy to solve on paper, the evolution of a grid quickly became tricky to invent for web use. But alas, after a couple Redbulls I came to a solution.

Visualization of jQuery Fade and Fill Diagonal Algorithm

To achieve the diagonal grid fill algorithm, only the height and width of the grid are necessary. Each diagonal column must be filled from top to bottom, left to right. It’s achieved by a loop and decrements to the selected cell by (width-1) and (height-1). The loop continues until the selected cell no longer exists.

Below is a PHP representation of the algorithm.

	$grid = array(array());
	$h = 4;
	$w = 5;
	$count = 1;
	for ($i = 1; $i <= $w+$h; $i++) {
		$index = min($i,$w);
		$row = max(1,($i-($w+1)+2));
		while ($index != 0 && $row <= $h) {
			$grid[$row][$index] = $count;
			echo $row.','.$index.' = '.$count.'<br />';

From here, implementing the code into JavaScript was fairly easy. With only a few simple syntax changes and a new grid from divs, the desired product was created.

<script type="text/javascript">
	function vuurrPixel(arg) {
		var size = 40;
		var h = $('.screen').height() / size;
		var w = $('.screen').width()  / size;
		var count = 1;

		for(i = 1; i <= (w + h); i++) {
			var index = Math.min( i , w );
			var row = Math.max(1,(i - ( w + 1 ) + 2 ) );

			while (index != 0 && row <= h) {
				if (arg == 'show') {
					$('.screen').append('<div style="display: none;" id="block-'+index+'-'+row+'" class="pixel"></div>');
					$('#block-'+index+'-'+row).css('top', (row-1)*size).css('left', (index-1)*size).delay(i*(size*.75)).fadeIn();
					$('.block').animate({opacity:"show"}, 1500);
				}else if (arg == 'hide') {
					$('.block').animate({opacity:"hide"}, 1500);

My problem conquered, I see many uses for it in the future. The diagonal grid fill algorithm can be used in many applications (slides, menus, etc…). Personally, my next feat is to use HTML5 canvases to implement the same algorithm for mobile compatibility.

GitHub Gist

Track Links With Google URL Builder

So… you’re wrapping up all your holiday advertisements this week. You’ve got traffic coming in from Adwords, Facebook, email blasts, Twitter and more. You’re getting great results, your ROI is through the roof and you’re almost on to celebrating. Sadly, when it’s time to analyze trends and individual campaign performance next week you notice a serious problem: You have no idea which ads or sources to attribute fiscal success because you didn’t build links correctly, or at all.

Bummer. That sucks.

But you can redeem your future self. Here’s how:

In order to properly build your links, several tags should be appended to the end of all URLs you wish to track. We’ll use as an example URL representing a holiday campaign on Facebook.

Before you start, add a ? (question mark) to the end of your base URL and remember to place an & (ampersand) between each utm_tag. Also, all URL building should be written in lowercase:

    utm_source tells Analytics which source the link is coming from. In this example, it’s “facebook,” but it could also be adwords, bing, nytimes, /blog or wherever you embed your link. EX:
    utm_medium tells Analytics what kind of source (a.k.a. medium) the link represents. Adwords, Facebook, Bing and the like are “cpc.” If you’re putting a link in your email, you’d append “email,” in Pinterest you might append “photo,” etc. The example URL builds as follows:
    utm_campaign is an easy one – it describes your campaign. Because our example is a holiday discount campaign, you’d attribute “discount” (or “holiday_discount” but I prefer to be succinct when possible). If you were sending out an email, you may attribute this to the date it was sent or the discount in the email. For our example:
    utm_term is generally only appended to PPC campaigns because it represents the keywords used in the ad (or the adgroup name). However, sometimes I use it just as another modifier for the ad. If you need to use two words, combine them with a + (plus sign), like “holiday+shopping.” In our example, the campaign term is “holiday,” since it’s a holiday ad, making the link now look like this:
    utm_content is best used when you’re split testing ad copy. You can either append a description of what you’re testing in the ad or you can number them. If your content tag has more than one word, make it one word in the URL, like “=bluebutton” instead of “=blue button.” I usually just number ads, so if this is the second ad for this campaign or adgroup, the example link would now finish:

You don’t need to use all the utm_tags; your data will attribute to the correct dimensions in Google Analytics to the extent you describe it. When traffic starts to come in and site metrics are recorded, go to Traffic Sources > Campaigns and/or Traffic Sources > Search > Paid to view your incoming campaign links. Switch between “Source/Medium,” “Source,” etc. to see your link building hard at work and compare source performance and future opportunity.

Building correct links allows you to properly and accurately analyze which sources in your marketing strategy are working best and which are ineffective; thus, saving you time, money and resources while you fuel the fires that make you money.

If you have the patience for automatic URL building forms, Google Support has a good one! Keep in mind this URL builder from Google only works for Google. If you’re using a link tracker other than Google, they may use a different _tag.

Minimum Sample Size: How Many Users is Enough?

You’re running a test on one of your landing pages with the primary goal of getting the user to click on a button. You are split testing (A/B testing) two different button colors (green and orange) to determine the impact of the button color on click-through rate. You collect the following initial results:

Button Color Visitors Clicks CTR (y)
Green 38 2 5.2%
Orange 39 3 7.7%

The orange button is better, right?

Not necessarily. Sure, the orange button has a 7.7% click-through rate (CTR) compared to only 5.2% for the green button. However, the orange button has really only earned one more click. If the next visitor on the page clicks on the green button, both variants will have a 7.7% CTR, indicating that button color is irrelevant in this application.

Here is what happens if we run this experiment for several months with 16,000+ visitors:

Button Color Visitors Clicks CTR (y)
Green 8,238 486 5.9%
Orange 7,893 734 9.3%

Is the orange button better now? Hell yeah, it is.

While the danger of making decisions based on too little data is an incorrect conclusion, the danger in collecting too much data is a waste of time, effort and money. Even though there was little data to determine orange was best in the first variant, had you decided to go with the orange button anyway, you could have sent all the above 16,000+ visitors to the page that performs at a 9.3% CTR. Thus, giving you a whole lot more sales/leads/etc.

So how many users are required to make this decision? We can use the following equation:

minimum sample size equation


  • n is the minimum sample size required to prove that the two variants are statistically different.
  • Z is the z-value corresponding to the chosen confidence interval in the Table of the Standard Normal Distribution.
  • E is typically known as the “error.” In this application, E is the difference between the mean values of two samples.
  • σ is the standard deviation.

The Z-score that corresponds to 95% confidence is 1.64 (from the Table of the Standard Normal Distribution).

For example, let’s say you are using the data in the second table above, and you want to be 95% confident in your decision. E is the difference between the two sample means, so

E = 0.093 (orange CTR) – 0.059 (green CTR)

E = 0.034

To figure out the standard deviation, we can treat binary data like continuous data because of the Central Limit Theorem, which states that as a binary sample gets larger, its distribution approximates a continuous distribution. So, determine the overall CTR as follows:

total clicks = 486 + 734 = 1,220

Total visitors = 8,238 + 7,893 = 16,135

That means the overall conversion rate was

1,220 / 16,135 = 0.0756

In Excel, you can get the σ-value by using the function =NORM.S.INV(1-0.0756) which returns 1.435.

Putting all of that together, the equation to determine what minimum sample size is required to show an accurate, statistically significant improvement of CTR for the tested page with the orange button is:

n = ( (1.64*1.435) / 0.034 ) 2

n = 4791

Therefore. In this example you would need to make sure you have 4791 samples in order to prove that you have enough to make a statistically significant decision over which variant is better.

Most A/B testing software will take care of his for you. What you really need to understand about the equation is:

  • As standard deviation increases (more variation in your conversion rate), you will need more samples.
  • If you want more confidence (95% versus 90%), you will need more samples.
  • As the difference in performance between the two variants becomes smaller, you will need more samples (it takes more data to make sure the difference isn’t just statistical noise).

Easy right? Now get to work.

Infusionsoft Partnercon

We presented the slides below at Infusionsoft Partnercon. If you saw the presentation, thanks for coming out! If you didn’t, all of the slides, notes, code snippets and more are below. If you have questions or comments, please let us know by emailing [email protected] or filling out our contact form.

The Presentation

Form Abandonment Code

YouTube Play as Google Analytics Event

Track Twitter Button Clicks

Track Facebook Like/Unlike/Share Button Clicks

Call Tracking

Use Twimlbin and Twilio to create an easy call tracking solution.

Twimlbin October Updates

After meeting with some good friends from the developer evangelist team at Twilio, we’ve added a couple additional features to Twimlbin at their request.

1. XML Output

The first of these new features will give users the ability to force Twimlbin to render the actual XML rather than the HTML interface when viewing in a browser by appending “/raw” to the end of any Twimlbin URL.

For example, if you visit, you will see the entire Twimlbin rendered as an HTML friendly view. But with the new feature, when you visit, you’ll see the raw XML rendered to your browser.

Here’s another hint: This will work with any interface that accepts XML.

2. UTF-8 Encoded Character Handling

Another of the newly added features gives users the ability to properly validate UTF-8 encoded characters. Previously, if you put special characters (like emoji) in a Twimlbin, it would return back an invalid Twiml. This has been fixed and updated.

Keep an eye open for additional new features coming soon. We’ll also be adding support for all of the new Twiml verbs so things like Queues will properly validate as well!

Measure Right The First Time

We presented the slides below at BOLO 2012. If you saw the presentation, thanks for coming out! If you didn’t, all of the slides, notes, code snippets and more are below. If you have questions or comments, please let us know by emailing [email protected] or filling out our contact form.

The Presentation

Form Abandonment Code

YouTube Play as Google Analytics Event

Track Twitter Button Clicks

Track Facebook Like/Unlike/Share Button Clicks

Call Tracking

Use Twimlbin and Twilio to create an easy call tracking solution.

Why the RFP Process is Broken

Since the end of the KGB, few three letter combinations have evoked as many negative feelings as RFP. No agency I know enjoys going through an RFP process, and most customers dislike the outcome it produces. The process is broken and outdated for both parties. It needs to be replaced.

An RFP is ostensibly implemented in order to protect buyers who aren’t knowledgeable about which vendors are best suited to solve a specific need. Many times, they are required for government “fair bidding” requirements. The goal in these cases is to get an “apples to apples” comparison and remove biases from the procurer.

The reality, however, is a different story.

Here is where the RFP process fails for both parties:

Requesting Firm:

  • Has to predict needs for a project even if they’re not an expert in that field
  • Has to understand the responses to technical specifications and how the firm best fit their needs
  • Has to match up different response language to ensure comparing “apples to apples”
  • Has to go through a lengthy process designed to get a bid that is low priced as well as confident in the firm’s ability to perform up to specifications

Responding Firm:

  • Has to interpret a document of needs with minimal context – often without a chance to have a real conversation with the Requester – which may or may not actually outline the root problems to be solved
  • Has to provide a verbose response to prove their ability and reliability without actually building a working relationship
  • Has to bid low enough to be noticed and hope vendor can then manage the project to that budget or get change requests approved

Once all the above hoops are jumped through, there is still no guarantee of good work or good fit between the vendor and client – WHICH IS THE WHOLE FREAKIN’ POINT OF AN RFP!

So how do you find a trustworthy vendor for your next project? Ask around, make a few calls, and get a feel for who you trust and who you fit with. Ultimately, you have to hope it all works out in the end.

At Vuurr, we have found the only way to be successful in solving problems for a client is to
work from a place of mutual understanding, starting where the pain points are and using our expertise to find solutions. Without taking the time to properly diagnose our client’s needs, we may be wasting each other’s time, money and effort on superficial quick fixes.

So whatever your next project entails, make sure you’re using resources that actually help you. Friends don’t let friends use RFPs.