Your analytics dashboard reports a healthy bump in visitors, SRPs, and VDPs – sounds like you’re in for a great month. But by the time the 30th rolls around, showroom traffic was flat, and so were your sales.
That disconnect is the central problem of automotive digital KPIs: too narrow, no context, and not enough emphasis on how people shop for cars.
To start pushing our industry past the “best-practices” KPIs we’ve all been focused on for too long, String proposes a new, people-focused metric customized to shopper behavior on your website. We’re calling it highest-value visitors.
We put VDPs to the test against highest-value visitors to see which correlated to sales more accurately. We took over 1.8 million data points across dealer websites encompassing different markets, inventories, pricing strategies, website vendors, and PPC providers. VDPs came in at a respectable 0.565 (for reference, 1 is a perfect correlation.) Highest-value visitors scored a 0.698 – a 24% improvement. And when it came to being a predictor of sales, tracking highest-value visitors is 52% more accurate. And we're working on an even more comprehensive test to see if that gap widens further.
Why Current KPIs Need Rethinking
The problem with best practices is that they can stifle innovation. When an industry adheres strictly to one set of metrics to determine success, there is resistance to change, even in the face of improved data and a wider range of tools.
Pegging success to inventory searches and vehicle details pageviews can be misleading. Take, for example, an email blast that drops 2,000 new visitors on your search results pages, or a PPC campaign that takes shoppers directly to your VDPs.
In terms of pageviews, your dashboards will say that you’re crushing it. But automotive retail isn’t like shopping for new jeans – it’s a lengthy process, so measuring product pages won’t tell you the whole story. But user-level segmentation in Google Analytics will give you the CliffsNotes.
In July 2013, Google Analytics introduced user segmentation, which allows you to look at user behavior across multiple website visits. Add to that the fact that Google’s getting better at tying users across devices, and you have the tools to pinpoint your most valuable shoppers and see who’s bringing them to you.
Form submissions are your most valuable web conversions, but many dealers are seeing fewer of them. With user segmentation in Google Analytics, we can examine the behavior of users who submit forms and create a new segment that mimics their behavior. That segment will capture all of your high-value users, even if they didn’t submit a form.
What this segment tells you is that your most valuable visitors spend a lot of time shopping you. They average as many as 4 or 5 visits across multiple traffic sources. They take 8 minutes on your site, and their pageviews can climb into the double digits. Some of their sessions may be bounces. And they might not view as many VDPs as you’d think. But they’re preparing for a big purchase, so their behavior makes sense.
Building Your Segment
Here’s our walkthrough for building a high-value user segment.
1. Start by creating a segment that tracks the behavior of all those who submitted a form. To do that, open up the advanced segments menu and click add new:
2. Next, click on conditions, change the filtering to users instead of sessions, and then select Page from the dimensions dropdown. Since this is a Dealer.com site, put “confirm” into the field so that you’re looking only at users who have touched a thank-you page. If you have a Cobalt site, use “leadsubmission.” For DealerOn, use “thankyou.” Give the segment a name, and you’re ready for the next step.
3. Now that you have a look at the behavior of the users who have submitted a form, change the date range to 90 days so that you’re looking at the widest data set that Google Analytics allows for user segments:
4. With your blueprint, you're ready to build the high-value users segment. So create a new segment again, but this time you want to match the segment filters to the behavior that’s below.
- Start by dividing the number of sessions by the number of users (in the example I’m using, it’s 2988 sessions divided by 713 users, which comes out to just over 4 sessions per user.) So we’ll set the first filter to “count of sessions” is greater than or equal to 4.
- After that, hit the “+ Add Filter” button and create one for “session duration” (the blue one, not the green one – you’ll see two options when you start typing “session duration” in.) Set it greater than or equal to the Avg Session Duration, but in seconds – in my example it’s 7:20 which is 440 seconds.
5. You can stop there and already have a pretty good user segment, or you can add one last filter based around pageviews. Since the count of sessions is 4 or more, and the average number of Pages / Session is 7.06, we’re going to look at only those users who have viewed at least 10 pages: 7 from the average Pages / Session + 1 pageview for each of the additional 3 sessions.
Name the segment, and you’re all set.
Act on Your High-Value Users
You can apply that segment to any reports that you want to run in Google Analytics, including the Source/Medium report in Acquisition to see which traffic sources are sending the most high-value visitors.
It’s also helpful to look at which sources are sending the most New Visitors once you have this segment active – that way you can see who’s bringing the high-value visitors to your site for the first time.
You can also look at the Demographics, Interests, and Geo reports under the Audience tab to learn more about who these visitors are. And if your PPC provider uses Google Analytics for retargeting, they can use the segment you’ve created as a starting point for custom audiences to target aggressively.