When Web Analytics and Email Marketing Collide
Merging your Web analytics and email marketing data isn't a new concept, but it is an under-explored one. Combining both data sets is a powerful way for marketers to harness the information and to develop a better understanding of their customers. The result is that marketers learn valuable insights into their customers' history, leaving them better able to predict future actions.
During November's Ad:Tech, I had the chance to sit in on the Merging Web Analytics with Email Marketing Metrics to Increase Performance panel moderated by David Baker with panelists Bill Nussey (Silverpop), Jay Kulkarni (Theorem), and Margie Chiu (Avenue A | Razorfish). Typically I liveblog the sessions, but this one hit on such an important topic and included such a wealth of information that I felt it deserved a longer, more thought-out response.
David Baker started the session by revealing a few rather surprising statistics. He noted that though 56 percent of marketers say the Web is the "hub of their marketing strategy", only 15 percent have a dedicated Web analytics staff. I found this startling. For me, basing the bulk of your marketing strategy around something that no one is in charge of monitoring or quantifying is a dangerous business proposition. If it's important enough to focus your marketing strategy around, it's important enough to allocate resources to.
Though the numbers aren't yet overwhelming, smart marketers are waking up and merging Web analytics with email marketing data. By doing so they are finding that they are left better equipped to understand their audience and increase overall profits. In fact, last year a JupiterResearch study found that targeted emails pulled from Web analytics data generated nine times the revenue as those that didn't.
But before you can start leveraging both sets of data, you need to determine what's important.
Step 1: Collecting the Data
It was once virtually impossible to get email vendors and analytics vendors to share data, but a new wave of partnerships between email service providers (ESPs) and analytics vendors have made integration far less complicated. In fact, many vendors are even working together so that marketers can spend more time working on their sites and less time playing office politics.
However, before you even approach vendors, you have to standardize your key business metrics to determine and define what you're actually looking for. There's so much data out there that if you don't know what you're after, you'll end up learning a lot about nothing. Defining what data and reports you're after will help you determine how to get where you want to go.
Your first step in this process should be to determine whether or not you're already running some kind of email and/or Web analytics program. If you're not, it's time to get started. According to the panel, there are two factors to keep in mind when selecting an analytics vendor. First, be sure the analytics program will allow you to track customer information without knowing their e-mail address up front (since customers won't always give one). Of course, the program should allow you to update this information once you do know it.
Second, make sure your email program allows you to keep your customer database within a separate Web analytics tool. If it doesn't, there may be no way to easily combine your data later on.
If you're already running an email analytics program, get with your team members and figure what's going on - what information are they collecting and what are they doing with it?
For example, know what flavor of email analytics are you running. During the session, Margie Chiu detailed the three stages:
- Basic (Jello): Aggregate level tracking of email response and site-level conversion. It's the just-add-water approach to email analytics.
- Intermediate (Pie): Response tagged and tracked by both email and site analytics system. Not so much about merging as it is about having multiple views of the media.
- Sophisticated (Baklava): Tracking and targeting by site and email activity behavior. May also include other tactics, including banners and offline methods.
Obviously, the more advanced the system, the more information you have to work with. The most basic thing your email campaign should be doing is tagging all your email links for site analytics purposes. We do this ourselves and it's a great way to understand what users are interested in, what they're letting pass by and how often they're engaging.
Once you have email under control, get together with your site analytics team and repeat the process. What level of information are you tracking and how are your team members managing the large inflow of information?
Step 2: Merge both sets of data
By combining both your email analytics and Web analytics data, you'll be able to construct a more complete understanding of your customer. You should see who converted, who bailed during the process (and where), where they came from, how often they visit, what page they were looking it, what part of your site they are most likely to spend their time, etc. Once you know all this, you can use it to better understand them and appeal to their needs.
For example, by studying the data you may find that a large number of your customers made it all the way to the shopping cart before abandoning the site. This may signal a possible conversion trap that you should follow up on. In the meantime, in order to entice customers back, send them a follow-up email acknowledging you noticed that they didn't complete the purchase and offer them a discount to finish. By offering a small price break, you let customers know you're paying attention to their experience and encourage them to reenter the conversion path.
Of course, you must be aware of the "creepy factor". The creepy factor means waiting at least a few hours or even a whole day between them abandoning your site and sending them a follow-up. Sending a customer an email about the product they were considering buying just 5 minutes ago may send them running away scared.
Rule #1 of Internet Marketing: Don't scare the customers.
If they do come back and complete the conversion, you may also want to consider sending them a questionnaire that (in part) asks what made them abandon the site in the first place. This is an excellent way to identify other obstacles to conversion.
Step 3: What are you doing to do about it?
Suddenly you find yourself with wealth of information. Now what should you do it with? The best thing you can do with this information is to use it! Realize that in your hands are the tools you'll need to make an educated guess as to what customers are likely to do in the future.
Take a look at your site's data and identify how customers found your site, where they reside and how often they visit. Once you know this information you can use it to better target your email campaigns and identify where they are getting lost in the conversion path. It could be that you're trying to push products they are not interested in or that you're presenting a conversion path that is too long or too difficult for users to navigate through. Whatever the specific reason for the abandonment is, knowing it will allow you to plug up the holes.
Breaking down the conversion funnel will show you where customers choose not to continue on and where tweaks can be made to keep them on-site. You want to clearly see each customer's individual path to understand their mindset as they make their way through your site.
Which pages receive the most traffic? How many customers arrive on the home page and don't click through anywhere else? Or if they didn't arrive on the home page, which page did they land on? If they clicked through, where were they headed? By studying where users click, you're able to see if they're going where you actually want them to go.
You always want to study the path to see where most customers are dropping off. If you're seeing a lot of visitors abandoning your site after they've already entered the checkout stage, you should hear sirens violently going off. These are the committed customers you cannot afford to lose. What made them leave after they already showed a personal investment? Did your site not provide enough information? Did they feel bombarded with other unrelated products? Locate the conversion trap.
If you find a drop-off in your conversion funnel, there are two things you want to do. First, get rid of all the unnecessary steps between when a user lands on the site and when they complete their conversion. The more pages and forms you require them to go through, the less likely it is they'll make it to the other side.
Second, make sure there is a recognizable call to action to pull customers through the site. If you lead the way, they are more likely to follow the path and not get sidetracked.
We recommend using the collected information to construct detail customer profiles. Segment your audience based on interests, conversion percentages, high engagers, or by whatever other specifics your data is showing you. Once you have your groups, construct varying emails with group-specific triggers.
For example, if you find that a large percentage of your customer base is most interested in pricing information, show that before you present them with all the options and different models. By giving them what they want up front, you tailor your site to the customer and present them with a positive user experience.
Rule #2 of Internet Marketing: Happy customers come back.
The best part about combining email and Web analytics data is that it provides you with historical data you need to predict what customers will do next. You'll be able to determine which paths lead to conversion, which pages with get the best response and how customers will respond or interact to different features. The combination of all this insight will help you, the savvy marketer, to fill in the gap that come with one type of tracking data and uncover new customer profiles.
One last piece of advice from the panel: Don't wait for the perfect time to launch a campaign - there will never be one. Just go!