So It Begins
In November 2015 the CEO of Duo was contacted by a representative from Google who was wondering if we were interested in using their AdWords platform. At the time, we were working on moving the Duo site to HubSpot and in doing so, I was reassessing keyword strategy and on page optimization.
With the redo in progress and the holidays coming up, we decided that late 2015 was not really the best time to start an AdWords advertising campaign. I told the Google representative (here on referred to as “The Google Man” or the “GM” for short) that we would talk in 2016, and I went on my way.
We finished the new Duo site, and I used some of my time to learn more about Google’s many platforms. I was already very familiar with Google Analytics and Google AdWords, having worked with both platforms extensively at my previous job, but I decided to take that to the next level. I got certified in both Analytics and AdWords. After solidifying my knowledge of these platforms, I felt ready to take on an AdWords campaign.
The Right Time to Start Using AdWords
This was all well timed because soon after, I got another email from our CEO. This time, he meant business. Since Drupal 8 had recently launched, he knew people would have questions about it and we knew they'd search for the answers. I wasted no time in calling the GM to get the ball rolling. It took one full month of research, planning, organizing, writing, designing, building, and communicating to get this all put together. Now that it is live, I want to talk a little bit about the strategy behind this AdWords campaign.
What Exactly Is AdWords Anyway?
To keep it short, AdWords is an advertising platform in Google. There are a few different ways to advertise, the two most popular being Search (pictured below) and Display. These ads show up all over the Google Network - a collection of websites that show AdWords Ads - when you search for a particular phrase or visit a specific website. When someone sees your ad, they have the opportunity to click and learn more about what you have to offer. Each time someone clicks on your ad, you are charged. This is called Pay Per Click, or PPC.
Here's an example of search ads:
The price and position of these ads is calculated by Google through something called a Quality Score. Google ranks your ads based on copy, keywords, landing pages, URLs, and price. This is actually more of an artform that you would imagine. A lot goes into these tiny little ads. It requires tons of keyword research, and the production of various pages and content on your site.
Ads are grouped by campaigns. Each campaign has Ad Groups (sets of related keywords) that run Ads on the various networks. So for example: the campaign for “Web Design” has Ad Groups for Web Design, Responsive Design, and [Industry] Web Design. All of those different groups are composed of their respective Ads and Keywords. You can also add Negative Keywords, which are words that, when included in a search query, cause the ad to NOT show up.
Creating Keyword and Ad Group Strategy
To start, our CEO asked me to focus on just Drupal words. I did some initial research (you can learn more about this in my last blog post about SEO), and found that Drupal related words vary pretty widely in search volume, competition, and price per click. Another thing of note is that Drupal is a fairly technical term. I did a little research on my own, and picked a few brains around the office for some more insight into this.
According to the sales and marketing folks, a lot of people actually find us organically, as we are a fairly well known and regarded Drupal shop. So if people are looking for Drupal, more often than not they will find us. I also used some of the takeaways from redoing the our site, specifically that Drupal is good for people who know Drupal, but it might be not help to get people looking for just a website redesign. With that in mind, I took the liberty to expand the search to a few different categories.
I added categories for Web Design, Responsive Design, [Industry] Design (focusing on the main industries we have historically attracted), Web Development, Collaboration Platforms, HubSpot, Marketing, Content Creation/Strategy and a few more. These groups and their subsequent keyword listings were selected because they really encompass what Duo does as a business and the opportunities we were looking to catch. The more generic terms, like focusing on web design, were also selected based on the idea that non-tech types were more likely to search for these terms, rather than the more tech heavy Drupal terms. At this point, I had my campaigns and Ad Groups pretty well organized. It was time to call up the Google Man.
Working With Our Google Contact
Throughout the entire process, I kept in contact with the GM. I knew what I needed to do to get this together, but I also valued his opinion. At this point though, we had to have a serious talk about budget. Without going into specifics, we needed to spend a certain amount of money to work with them, as well as to increase the odds of success for our campaigns. With this in mind, I felt that we could get more out of this experiment by expanding away from just Drupal.
So I went back to our CEO with the numbers, my suggestions, and a little back up in the form of various opinions from the GM (who thought web design was a good area to target) and members of the Duo team. From there, we decided to target Drupal, Web Design, Web Development, Responsive Design, and Collaboration Platforms terms, and we set a budget.
Creating The Ads
With the keywords selected and the campaigns organized, it was time to start on the creative. To begin, I worked with Sarah, our copywriter/content strategist, to get some ideas. Ads are very restrictive, limiting you to a headline of twenty-five characters and two lines of ad copy at thirty-five characters each. With such little space, it's tough to get a message across, especially if you try to fit in a long keyword, like Collaboration Platforms.
We created a variety of ads to run some tests and see what ads actually got clicks. If you create multiple ads, Google will rotate through the ads, and use the historical data to give preference to those that perform best. So we came up with 10-15 different ad sets to try out. This gave us some options to play with. We let the Google team select which ones they thought would work best, using 5-8 ads per Ad Group in the final campaigns.
The message of these ads focused on the types of customers we wanted to obtain. Words like “enterprise” or “large scale” were specifically selected with the intent to weed out smaller businesses who are looking for smaller projects. We also used negative keywords to improve our odds, preventing people who searched for things like “cheap” “build it yourself” and “low cost” from seeing our ads.
Our ads were written to aim for the highest quality score we could, which would allow us to show up higher in the results and spend less money.
We also created Display Ads. These are smaller picture-based ads that will show up on relevant websites in one of two ways. One of these ways involves keywords. Sites that use similar keywords or contain similar subject matter are selected. If they are a part of the Google Network, then our ads might show.
Another thing we are trying is A/B testing. We are planning to run two identical ads, but with one difference - button color. We will be able to see which one converts more to help us select a secondary accent color for our own website. This is just a small added bonus, and clever way to use this system to get more insight into our business and customers.
Another strategy we used was retargeting. These ads would show up in a similar fashion, but to people who had already gone to our site. How does Google know this? A simple tracking code places a cookie to show the network who to remarket to.
Below is a perfect example of remarketing. I looked at this hat last week, and now I'm seeing ads for it on nearly every site I visit.
Relevant Landing Pages
With the ads created, I turned to Stephanie, our HubSpot expert, to make relevant landing pages and thank you pages for each group. This is necessary for two reasons. First, it will allow us to specifically track which Ad Groups result in specific conversions. Second, it will help improve our Quality Score. A relevant landing page, complete with appropriate keywords, should boost your score, while improving conversions.
We then created three types of landing pages. One was a simple Contact Us CTA to appeal to people who were already interested. We also tried a variation with an Ebook/Whitepaper download to incentivise people to fill out the form, and another called “Schedule a Consultation.” We created a different version of each landing page for each Ad Group so the visitor who clicks an ad lands on a page that very clearly matches what they were looking for. For instance, if someone clicks “chicago web design” but lands on a page for “expert web development” they might hit the back button thinking they’re in the wrong place. The message had to match and the page had to be clear if we wanted to convert visitors into leads. Much like the ad copy, we can use the data we will harvest to adjust these and improve performance.
Once built, we installed specific tracking codes on the thank you pages that would be linked to the campaigns to track specific conversions.
Landing pages were actually the last step in the process. With these done, it was time to go live.
To go live, I had to give all of this information to the GM. I did this in the form of a massive, but incredibly well organized, Google Sheet that I shared with the GM and his team. From there, they were able to take all of this data - Campaigns, Groups, Keywords, Negative Keywords, Ad Text, Display Ads, Landing Page URLs - and actually create the ads.
Once they were ready, I hopped on a final call to review everything. The Google Man actually told me (multiple times) that our campaign was one of the most thorough and well organized that he’s seen. That was pretty awesome to hear, not just because he was complimenting my work, but because it meant that we did everything we possibly could to follow best practices and really set up these campaigns for success.
With everything ready, I hit the button to go live. One month after our CEO had initially reached out to me, our campaigns were up and running. We’ve already been getting clicks too.
Working With Data For Future Improvements
One thing to note is that we have no idea how these will actually perform. What we will do is monitor our campaigns and make adjustments based on the data we take in. I will have weekly meetings with a Google rep to go over these tweaks. So as we go, we will create and remove specific keywords and ads. We will also get more insight into what landing pages convert and new keyword opportunities, and adjust our strategy as needed. We have this tied to our analytics for even more insights on user behavior, as well as site searches and conversions.
In the end though, this will come down a certain degree of luck. That huge project could find us, or maybe we will miss an opportunity when our daily budget runs out and our ads stop showing. However, only time will tell. Part Two to this post which will be focused on these adjustments and outcome of our campaign. Check back for an update!