If It Looks Like a Bot, and Smells Like a Bot…
Many of us have heard about, or experienced robot (bot) traffic on websites. There are many forms of bots hitting our sites. There are search-spidering sites, inventory scrapers, as well as other bots we’re not quite sure of the intent. However, it is the bots posing as paid referral traffic that are perhaps the most concerning.
That’s right, there are still companies pitching dealers display, retargeting, or email campaigns, with promised “auto-intender” traffic, that upon closer look seems to be traffic generated from bots. For me the most challenging aspect is that often this bogus digital marketing is offered by reputable, very large companies. When looking at both the questionable source, and poor quality of the inbound traffic, I have to assume these reputable companies are perhaps reselling offerings from non-reputable companies.
When analyzing the quality of inbound traffic to a dealer’s website using Google Analytics (GA), the first smoking gun around bot traffic usually appears as disengaged traffic. My defined segments for disengaged traffic are flagged by: very low time on site, less than 2 pages viewed per session, and large percentages of traffic from outside the U.S. Once referral sources begin to show these patterns, I dig deeper into country, technology, devices, and operating systems to hopefully snuff out bogus traffic.
One recent example came from one of my dealer customers who had purchased a $3,000 monthly package for display advertising from a reputable media company. The company promised display ads to auto-intenders, people who planned to purchase within the next 30 days. My initial inspection of the data first showed very low engagement, which triggered me to layer on bot filters. Filtering for bots can take many forms, but at a high level it is both looking for traffic that does not mirror the overall website profile, and traffic that does not follow typical shopping behavior. As I began to look closer at the traffic from the media company, the devices being used by the “shoppers” looked very strange. Google Analytics showed that 95% of the traffic from the media company’s campaigns were coming from an Android phone. Certainly, with the number of iPhones, Windows Desktops, etc. in the U.S., a larger percentage of traffic would come from those devices?
So, for a comparison, I then pulled the overall traffic for the site. Sure enough Android only represented around 38% of the traffic. So what caused this strange anomaly? The most logical source would be Phone Farm traffic.
What are Phone Farms? They are banks of cheap phones purchased by individuals, along with software that plays videos, or initiates clicks on these phones as a way to earn money. Since Android phones are the least expensive, they are the most commonly employed devices for phone farms. The website beer-money.org has famously been promoting phone farming as a way for people to earn extra income, at the unfortunate expense of advertisers. Now, I have no proof of this, but at the very least this dealer was getting very low quality paid (and suspicious) traffic, that is providing zero return on investment.
When I showed the dealer these results, they were understandably unhappy. The dealer considered pursuing the advertiser for a refund, but instead just did not renew the 3-month agreement. On the bright side, the dealer was extremely happy they did not opt for 6, 12, or 18-month agreement this large media company was pushing for!
The lesson here is to inspect your traffic, and don’t rely on your vendors’ reports to measure success. Take control of your own website dashboard using Google Analytics, and you will be money ahead!
If you are an automotive retailer, I would love to discuss helping you with digital marketing audits, including auditing your paid search traffic. I can help you get more transparency with your digital market investments, saving you wasted spend and improving your return on advertising spending. Details here: www.generationsdigital.com
You can also follow me here: