A successful Adwords campaign is an ever evolving entity. All the preparation and planning is just a base, and will require constant adjustment. This adjustment is based off of the trends you see and the data you collect. Six weeks in, we have made a number of adjustments to our campaigns and are starting to see the payoff!
Initially there wasn’t much to do. We made a few changes to keywords, specifically the match types in the Drupal campaigns. With this change, we made sure that Drupal was required in all searches for these words. This was just a safeguard to make sure we were getting relevant traffic.
Another thing we noticed was with our Collaboration Platforms campaigns. Lots of people were searching “[Company] Intranet,” so we were showing up for a lot of search results for company portals. We noted this, and have consistently been adding these companies to the negative keywords list. Delta employees - you guys were the worst offenders. The thing that’s annoying about this is that a handful of people looking for their company intranet actually clicked our ads, and bounced. These are common problems with search advertising, and one of the main reasons why you have to be diligent and monitor your campaigns.
At this point, we had a bit more insight on what was working and what was not. So we went through the keywords, adding a few more specific ones and removing a few of the broader words.
One thing I also took notice of was the performance of our Display ads. The Google Rep suggested these, so I took his advice. However, after two weeks of data, it was clear this was not the right path for us. They were hogging up way too much of the budget, with little return. At this point, we gave these ads a separate, smaller budget, while increasing the budget in the other campaigns.
After another week of Display ads running on a smaller budget, we were still not having much luck. At this point, I was done with them. However, by this point we had enough traffic to turn our display campaign into a remarketing campaign. We set this up so that anyone who viewed 2+ pages of our site would see these ads across the Google network. The logic here was people that actually clicked around were the one who were truly interested in learning about what we have to offer. Since we figured there were not a ton of people seeing these (given the bounce rate), we decided on a really limited budget of $10/week.
With this in mind, we changed the rest of the campaigns to share a $90/week shared budget. I also went through and did an audit of where our words were in terms of showing on the first page and how much we would need to increase the bid to get there. On average, it turned out that increasing the minimum bid from $5 to $7.50 would give a huge boost to our ad placement, hopefully getting more of our ads on the first page. So we started with this jump, noting that we could possibly increase this more if we needed to.
Little did we know, but the changes we started to implement this week would be the start of seeing some real progress.
The next thing we did was more keyword refinement, adding more Drupal words and pausing a few other words that were either not getting clicks or not as relevant as we had initially thought. We also added a few more of those hotel and intranet negative keywords words.
At this point, one of the main things I noticed was that people were getting to our landing pages, but nobody was filling out the forms. People were also clicking the site links in our ads. We added heat mapping to the page, and noticed that all the clicks were on words that people thought would take them to more information. That was the “Ah Ha!” moment. I figured that the biggest problem here was the the landing pages were not working how they needed to. People would click the ads, but were not ready to fill out the form. They wanted more information.
So we made some major changes. First, we cut the form down to just Name, Email, Website, and Comments. We also made it more of a straightforward contact form, rather than suggesting a larger engagement or a giveaway. This was done to ease people into working with us. Finally, we added more links to help people find better information on our company. We added some links to our team and our work in the page copy, and added the navigation bar back to the top (note - removing the navigation bar is a suggested best practice for landing pages that does not always work for every single situation).
This week focused on a more refinements, along the lines of what we had been doing the last few weeks. We changed the match type for Drupal to exact match. We also adjusted the keywords in the responsive design campaign to be a broad match modifier in an attempt to improve the quality score and save some money (they were a bit vague before this). We also paused the web development campaign. We thought about doing this earlier, but figured we would wait because though it was not getting a lot of clicks, it was not wasting a lot of budget.
Some Interesting Metrics
At this point, we have had a handful of conversions of varying quality. We also have had noticeable improvement in our campaigns since April 1, when we first started to institute these changes. These changes have resulted in a 10% increase in impressions and a 13% increase in showing on the first page.
We have a few more ideas for improvement too. These include a new Ebook to answer web design questions, sending people directly to a service page with a contact form on the side, and changing the hours the ads show to be only from 5am-1pm (when most of our traffic occurs).
The main thing to take away from this is the idea of constant refinement. It’s interesting to note how much of an effect some of these changes had. The more we learn, the more we can make smart decisions, and hopefully the more success we have. According to the Google reps, the next six weeks is where we will really start seeing success. Check back in six more weeks for an update to see if what we’ve been working on pays off!