The six data sea monsters can combine in many awful ways to create harrowing problems for marketing departments and leadership. For a refresher on what these monsters are, check out this post.
As you gather ’round the turkey this Thursday, consider saying a brief thanks that none of these problems are plaguing your marketing department. (Or if they are… we can help.)
“83% bounce rate?!”
This company was migrating to a new website, and was about 75% through the migration. However, there was no clear owner of the entire process or the data involved. Thus, the web dev team wasn’t telling the digital marketing/CRO team when they were moving forward with the migration, and the digital marketing/CRO team wasn’t telling the web dev team when they were running tests on the website.
So when the digital marketing team ran a test on a portion of the website that had already been migrated, the results were predictably awful. This lack of communication between the teams is one of the hallmarks of a non-data-driven culture.
Furthermore, the digital marketing team was using 10 different tools to run their tests, which meant that up to 10 platforms were hitting the website at any time. These tools each had unique ways of measuring and reporting the test results, so any benefit the tool stack might have offered was offset or completely negated by the confusion and work involved in interpreting the tools’ results.
“We don’t have keyword-level data.”
We had worked with members of this company’s digital marketing team to build out keyword-level data and connect it to revenue, so we were astonished to hear this. However, the two people we had worked with—the company’s “stewards of data”—had left the company. When they left, not only did they take all their data-related knowledge with them, they left a void in the company for data-driven conversations.
None of the remaining members of the digital marketing team knew about:
- The systems the two stewards had been using
- How to access those systems and harvest that data
- How to drill down in the data to get keyword-level information and tie it back to revenue
- The importance of regular conversations about data
Fortunately, we were able to show the rest of the team where to find their keyword-level data and how to use it, but this experience underscored how dangerous it can be to have only one or two people who understand company data.
“Our CPL is great; we just need to grow.”
In this case, the company wasn’t viewing their data with the necessary critical eye. Their average cost per lead (CPL) was around $110, well within their target range. However, that average CPL included all their keywords, brand and non-brand.
In general, brand keywords target existing customers while non-brand keywords target new customers. This company’s paid search portfolio revealed that 52%of their total spend went to brand keywords. (We recommend no more than 10%.)
CPL for brand keywords can be a vanity metric, especially when calculated as part of the overall CPL. Brand terms’ costs are lower due to low competition and search volume; non-brand terms are more expensive. The average CPL for this company’s brand keywords was $75, well below their average CPL of $110.
By comparison, the average CPL for non-brand keywords was $175—but even with the higher costs, these keywords were utterly essential for growth. By leaving only 48% of their paid search budget for non-brand keywords, the company was starving the keywords targeting new customers and thus actually sabotaging its growth.
A stronger understanding of paid search data and a culture with regular conversations about interpreting data would have prevented the digital marketing team from sinking so much ad spend into brand keywords.
“Impressions are down. What’s going on?”
This power company was concerned about impressions, believing that the quantity of people who viewed their ads was a crucial metric. To maximize impressions, they had cast a wide net with their keywords, hoping to reach as many people as possible. So they were dismayed to see that their ads weren’t getting as wide of an audience as they had hoped.
That said, the company was quite happy with the click-through rate (CTR) of some of their keywords. While their average CTR was only .09%, the CTR for many of their low-impression keywords averaged .60% – almost 7 times the aggregate CTR.
When we examined their keywords to address the company’s issues with impressions, we found that their wide-net approach to keywords had led to deceptive metrics with potentially devastating results.
Many of their low-impression/high-CTR keywords had nothing to do with what the company offered:
- Keywords related to gas prices had high CTRs, but were used almost exclusively by people seeking gasoline for their cars, not gas heating for their homes
- “Transformers” as a keyword attracted fans of the movies and toys, not people interested in electrical transformers
- More mystical keyword choices included “Power Rangers,” “Power Ball numbers,” “Monster Energy Drink,” and “juicers”
Yes, these keywords were delivering clicks and impressions, but they weren’t the impressions and clicks that would turn into customers and revenue. Worse, the company was spending almost 25% of their paid search budget on these non-relevant keywords.
We suspected that so many inexplicable keywords had come as a result of using a keyword planning tool, then adding all keywords in a group without reviewing each keyword individually. A more in-depth understanding of paid search data and tools would have prevented such a significant waste of spend.
We hope this trip to the Davy Jones’s Locker of digital marketing hasn’t been too harrowing. Remember, there are steps you can take to correct these problems and prevent future ones.
You might have noticed that many of these problems are based on a lack of data-driven culture. Next week, we’ll show you how to build just such a culture, from hiring on up.