On the television show Shark Tank, judge Mark Cuban once put a young entrepreneur on the spot by asking him about his web traffic and leaving the conversation at that. Where was the follow up, Mark? What kind of traffic was he getting (organic, paid, referral), and what is a good baseline for measuring whether this is a good or bad number?
This is key when looking at the performance of dealer websites—a single metric will hardly ever tell the complete narrative of how your website is performing. It’s asking the appropriate follow up questions that will glean insights into the true success of your digital marketing strategy.
What to Look For
One of the first numbers that draws my eye when reviewing auto website performance is monthly site visits. This metric, found in most analytics tools as “Visits,” is not wholly useful on its own—it can tell you a lot, but you need to know what contributes to this figure.
For example, let’s say monthly visits are down month-over-month by 20%. This is a catastrophe, correct? Actually, that information alone isn’t enough to draw any firm conclusions from. There are several follow up questions that immediately become relevant:
- Did the dealership change anything in their traditional advertising efforts during this period? If a store cuts back on traditional ad spend, the digital side of their marketing efforts will be affected.
- Is seasonality at play? Perhaps the current month is traditionally a weak month, or the prior month is typically a strong one.
- What are the traffic numbers year-over-year? While month-over-month differences can show short-term trends, a baseline for what is “normal” needs to come from a year-over-year comparison.
Let’s say traffic is still a concern after addressing the questions above. Traditional marketing efforts have been consistent, there are no apparent seasonal issues affecting traffic, and numbers are actually flat year-over-year. This could signal that the traffic metric needs help—but what to do?
First, we need to figure out where visits have declined, or else we won’t know what is really happening to our site traffic. We can do this by looking at marketing channels (I use Google Analytics).*
There are different types of traffic—organic traffic comes naturally from user queries on search engines, paid traffic comes in through display or text ads, and there is also referral traffic and social media traffic. Funneling monthly site visits through these channels can help provide a more clear understanding of exactly how traffic is flowing into your site.
This is absolutely the key takeaway. Whether traffic is up or down does not tell you anything—you need to know which type of traffic has changed in order to address the decline. In Google Analytics, look at “All Traffic” and then at Source/Medium or Medium to separate traffic into specific types.
Case Study
A case study: Awesome Dealer on the northeast coast. Awesome Dealer had a decline in traffic from December to January, so I looked at year-over-year traffic. Comparatively, traffic looked almost flat, although there was a decent 17% increase in overall traffic YoY. But that only tells us more people visited the site in January 2014 than in January 2013. It doesn’t tell us why—whether they came in through an organic search, clicked on a Facebook link, etc.
Since filtering traffic by channels tells a better story of what specifically is working for a dealership website than a broad metric like “visits,” I broke down Awesome Dealer’s traffic that way. This is what I saw for the top three mediums:
While traffic was somewhat flat overall, organic traffic was up significantly—more than 50% year-over-year. Organic traffic is the highest converting type of traffic, so this is always a good anchor to rely on. These numbers tell a much clearer story than looking at only monthly visits. As the marketing company for Awesome Dealer, we were happy to be able to report that our search engine optimization (SEO) efforts and our search engine marketing (SEM) efforts through cost-per-click campaigns were in fact yielding returns year-over-year.
If we would have just looked at “All Traffic,” all of this insight would be missing.
Follow-Up Questions
At this point it became important to discover what was pulling down our traffic numbers elsewhere. If traffic through the organic and paid channels was booming, where were the declines that were canceling out this growth, causing overall traffic to appear flat?
One source of decline for this dealer was traffic grouped under “Direct/None,” which is where Google groups its private keywords. When users are logged into a Google account, including YouTube, their statistics are lumped into this category for privacy reasons. There is unfortunately little insight that can be gained from this category.
Another source of decline that could be addressed for this dealer was traffic coming via e-mail. This, like a decline in overall traffic, needs to be assessed within the correct context. Because I didn’t have access to Awesome Dealer’s e-mail marketing plan, I couldn’t know for sure why e-mail visits were in decline from the prior year. I suggested that the e-mail campaigns perhaps lacked a strong call to action, and suggested that the dealer ensure they had a substantial, legitimate list of customers to reach out to as well as an e-mail distribution program that ensures optimal delivery rates.
Monitoring site traffic is not something to be handled with a single discussion, but rather something that needs to be tracked continuously along with other key performance indicators. A lone snapshot of a single metric cannot tell you the full story of how your digital marketing efforts are performing. Learning to ask the right questions can help provide context for those general, broad metrics, and will empower you to better assess the performance of your digital marketing efforts.
*Google Analytics recently rolled out its “Channels” tab under “Acquisition,” grouping traffic into default marketing channels, which is very helpful for this type of analysis. However, as a newer feature this won’t be available for year-over-year comparisons until July 25, 2014.