This post is going to be a bit different from the technical Google Analytics or Google Tag Manager posts I usually write. But, I want to share with you an example of why sites should be built with analytics in mind.
As an analyst, there is nothing more frustrating than analyzing a site that wasn’t built with analytics in mind. You can probably think back to a couple of projects where you told a client measurement of a site would be simple only to figure out the site was built with no default way of measuring success. I know I have dealt with many sites that have wonky URL structures, absurd URL parameters, a lack of actual pages on the site (“this site is so cool, it’s all one URL but the content changes!!!), and so on. I understand that building a cool experience sometimes means taking a significant leap from the norm. But why build a site that can’t be analyzed?
For those of you that have never had the joy of experiencing this kind of pain, let me give you an example. I had a client that was launching a campaign and so they had a creative agency build a microsite for the campaign. The microsite was going to be soooo cool and it would look sooo amazing! It had all of this content related to the campaign, a bunch of social elements, and gamification built-in!!!! The people were going to love it.
So they asked my company to measure the effectiveness of the campaign. One site, monitor a few metrics, present a pretty monthly report, sounds awesome, right? Ehhh.
We had this nice measurement plan thought out. We’re going to judge the site on a couple KPI’s like campaign registrations, but they also want to know which sections of the site are the most popular, get the most engagement, and so on. We needed to get everything set up and tracked ASAP, because 96.7% of clients won’t give you the time you actually need for planning measurement. And yes, that’s a scientific fact. So, obviously, before we complete the measurement plan, we needed to understand how the site works, what can and can’t be tracked, how the registration process works, and so on.
My boss at the time gives me the development site URL and I’m already thinking about my plans for the night because evaluating what to measure on the site should be a piece of (cheese)cake. Except when I landed on the dev site, I’m spitting out my La Croix sparkling water as I realize the entire site is one URL even though there are really many pages a user can go to throughout their experience.
For instance, there was an entire main navigation bar at the top of the site, just like any regular site, where users could click on a menu item to take them to another section of the site. Except that when the user clicked on a menu item, it took them to a different “page” but the URL never changed. So cool right?
Yea, cool I guess, until you have to tell the client which section of the site is the most popular. Or gets the most activity. A quick peek into Google Analytics would show that www.example.com was the most popular page, the only page, and the page that drove the most registrations. So yea, that’s not helpful at all. And the technical agency that built the site is busy high-fiving each other in their Levi jeans about how cool the site looks, because the appearance and functionality is the only thing they’re being evaluated on.
So now that my night plans have been spilled all over my desk, I’m clicking around the dev site cursing the developers and Google Maps searching where they are located in case I get the inclination to pay them a quick visit. Realizing the structure of the site, I’m inside my head thinking about everything that we CAN’T track because of how the site is built. Which section of the site is most popular? Most engaging? Yea, hey client, can’t tell you that because the entire site is one URL, which is extremely helpful for me (sarcasm at an all-time high now if you couldn’t tell).
I’m not going to explain everything we did to track the site, but just know that I was working with the technical agency up until the minute the site went live to make sure we could track everything. I lost an entire night’s sleep because of this and Andy without 8 hours of sleep is an Andy you won’t want to add on Facebook.
Long story short, the entire “measurement” of the site was fine (my report was so pretty!), but the days leading up until the site went live was a nightmare. The best quote I can share about the development agency that built the site, when I told them how I couldn’t evaluate the much of the site’s performance because of how it was built, was “Yea, we didn’t really think about that.” Oh really? Yea, I noticed. Don’t worry, it’s not like anyone cares to evaluate how the site is performing against its objectives (*face palm* – *slams laptop shut*).
As an analyst, you’re only as smart as the touchpoint (website, app, etc.) will allow you to be. So please, start forcing clients, technical agencies, and the like to build sites that can be easily measured. It’s for your own sanity. You’re welcome 😉
Have you run into similar issues? Share your experiences in the comments, I’d love to vent and hopefully laugh about past projects with you!