Proof of Concept => Has anyone build dynamic *but trackable* landing pages with EE?
Development and Programming
“Test. Test. Test.” is the mantra of a good marketer, correct? And yes I’m on board with that. With multiple variables to test, and some of those being random, the tracking can get pretty hectic.
I’m bidding on a project to develop (PPC) landing pages for a client who offers multi-service. That is, one template could be used many many times. I’m wanting to propose that this be done with EE. Duh? 😊 with flexibility as well as the ability to randomize being the big pluses.
But I’m at somewhat of a loss for tracking component - which is key. How can I capture the value of the parameters such that once there’s a conversion, we’d know exactly what the lead was looking at, and thus be able to attribute success to a parameter value, or a combination there of. I’m leaning towards Google Analytics for e-commerce such that each “value” would be a “product”. I’d have to figure out a way to pass those things forward (in a SESSION() maybe?) but I’m confident something is doable. One way or another I’d figure it out.
I’ve used G.A. but not for e-comm. It makes sense to me but I figured I’d throw it out there for the punters and see what comes back in terms of other ideas, validity of this idea, etc. Anybody?
Thanks!
I’d suggest using Google Analytics. Utilizing their funnels+goals and/or custom events can be powerful and can be setup fairly quickly. The only downside to GA is that it is not real-time, and if you’re not well-versed in setting up events, funnels, etc. it can be a slow development/validation/debug process.
One of the biggest pain points in website tracking is managing the data collection, access, storage, graphs, etc. which GA (and other packages) do easily.
Hey ajlny
Thanks for the reply.
Yes, I’m aware of GA. Love it!! But I don’t think goals and funnels, etc. are what I need here.
Maybe I should have given an example? (Let’s say) there’s 4 variable elements on the landing page. For example, a call to action button (color and copy), a headline, some copy, images, etc.
In addition, to that the traffic to these pages is being driven in via various “campaigns”. For example, AdWord (with numerous keywords, and ads, etc. each with a uniquely tagged URL), direct mail, email blast, QR codes, etc.
Someone clicks on one of the calls to action and the client “makes a sale”. The question is, what combination of elements was the most effective? Is thre some magic headline that works for some ads and not for others?
In short, think A/B testing but on steroids. In theory, all that data is available. In theory all that data can be analyzed. What I’m trying to sort out is how capture the data. ALL of it 😊
Yes, I believe GA will probably play a role, but maybe not. Not unless there’s some way to pass multiple “parameter”. URL builder only allows for 5 or 6, right? As a result, I’m considering using GA’s e-comm functionality and each parameter can be a “product” in what’s finally “ordered”. I haven’t used GA for e-comm so I’m not too sure if that will work. That’s why I’m here brainstorming.
OK, now what do you think? 😊
It sounds like there are a discrete number of variables (campaigns, elements) and a discrete number of instances of each (5 colors, 5 different headlines, etc.).
You could still use something as basic as custom events to track the combinations that resulted in a sale. The mapping of content=>tracking code would need to be stored elsewhere, but I could imagine something like this:
Campaign: * DM: cd1, cd2, cd3 * Email: ce1, ce2, ce3 * AdWords: ca1, ca2, ca3
Element: * Color: Red (e1), Blue (e2), Green (e3) * Headline: x (h1), y (h2), z, (h3)
The trick would be keep all the values (and mappings) stored outside of GA and adding the logic to create the final event.
For a sale, User A generates: ce1-e1-h2, User B generates: ca3-e2-h4, User C generates: ce3-e3-h4. This could be exported and analyzed in Excel.
Additionally, you could even track information in a funnel fashion and track not just the final conversion, but the steps to conversion, i.e., which combination is not just most effective in generating the final sale but getting from page=>shopping cart, cart=>payment, etc.
This approach is more similar to using Omniture (passing raw data and letting other apps do the final processing, etc.).
Would this approach help?
Thanks. Yes, that’s about it, more or less. I think I better do some reading on Custom Events before I take this any further.
FWIW, in this particular case there’s no funnel per se. It more or less landing page and then “buy”. With “buy” being goals such as join mailing list, request a quote, download white paper, etc.
Thanks for your help. This was the 50,000 foot pass. It seems to be doable. When I get down to 25,000 or 20,000 feet I might have some more ideas to bounce around. If anything comes to mind between now and then please add it to this thread.
Cheers Mark