Measurement Marketing Case Study: Internet Alchemy

Intro + Case Study Context

Internet Alchemy has done a really solid job focusing on their sales over the past 5-7 years. However, with increased pressure from a booming industry, they couldn't put off knowing their numbers to the extent they needed to anymore. 


Enter: My contract date starting around November of 2017. 


I was hired to focus help this team become more data-driven, to know their numbers and grow their numbers. Don't get me wrong here, they knew some of their numbers. But no where near to what they needed to for their business to grow how they wanted it to. 

Phase 1 - Planning

The planning phase is one that everyone cuts corners on or skips all-together. Which will seriously come back to bite them as they get further down the road and deeper into the funnel analysis; it just becomes a mess. 


We spent some serious time on the planning process. We had 2 large websites to go through and a fairly small site to look at. 


Like with all sites, we started by mapping our own journey through screen-recordings and LucidCharts: 

We mapped the entire funnel:

This way we not only know everything that goes in within the link POD System (Purpose Offer Details of every step), this is a very useful asset the company now has for their team to not only communicate better between departments, but actually know the how the funnel works and understand it better. 


You wouldn't believe how many people that were on the team, for years, who didn't know of the actual process the users go through with the team's funnel. And in now knowing, they are able to understand and do their job better as well. #winwin

Phase 2 - Building

If you do a great job in the planning stage, the building stage is a lot easier. There was even more work to do here, however. 


The next step was to get Google Tag Manager involved, as this is our go-to way to deploy any kind of code. We had multiple different codes to deploy here. 

Facebook: 

- Base Pixel, Conversion Tracking Pixels, and Event Pixels. 

Google Analytics: 

- Base UA code and setting up all the events, some of which we turned into goals as well. 

- We also installed standard e-commerce on site, and enhanced e-commerce on another site. We needed to leverage the Data Layer Push with GTM in order to get the transaction details we needed to the data layer. 

Google Optimize

- We do spilt testing with Google Optimize, so we needed to install the base code of that here, in addition to the "Page Flickering" code so the page wouldn't "flicker" for the split testing. 

Custom Javascript

- We wrote a custom Javascript variable in order to figure out the "Days in System" when a user opts in as a lead. And then also for a sale, as we have multiple products. The CRM we were using wouldn't tell us that out of the box, so we doubled down to figure this one out, as we wanted to know that data point. 


We did a lot more with Google Tag Manager, but to not get to verbose here, I'll just say the rest involved a lot of QA'ing of the overall process, making sure our events matched up with how we wanted to view them in Google Analytics, we also setup user ID tracking, to be able to tie contact IDs in our CRM to the actual users coming to our site. That was a lot of Data Layer work as well...

Phase 3 - Reporting

Google Data Studio is our choice of reporting. It does some serious calculations on the back end that are seriously impressive. We feel right at home with the Google Stack, anyway, so we might as well keep with it. 


We build a custom API to pull our CRM data and integrate that with a Google Sheet. Data Studio then reads from the Google Sheet in order to compare sales with visits/goals/etc to get the true, full picture of the online sales spectrum. 


One way we love to use Data Studio is so that its custom for each and every business we work with. Seriously, those templates you can start with or find from anyone else out there are not going to work. Did you use their sales strategy and closing lines, too? Didn't think so...


From the LucidChart we built them from the start (this is something we always do), we simply took screenshots of the step we wanted, and then imported that into Data Studio, and then put the metrics by the step as scorecards. And then in the middle, we calculated the conversion rate with a calculated metrics in Data Studio. 

Every time someone asks me what our conversion rate is for a certain client, I say, "Which conversion rate? We have 28!"

Naturally, they are probably asking about the core conversion rate. But I like the philosophy of Occam's razor - to no assume. And everyone defines conversion rates differently, even if you agree upon the same step! (conversions/reach vs. conversion/impressions) Which one is right for Facebook Ads?! Very different metrics...

Funnel Tracking, 2.0:

Phase 4/5 - Forecasting + Optimizing

Clients are often surprised when we can then forecast and predict the leads and sales we are going to make in the coming week/month/quarter/even year. 

This shouldn't be surprising if you are in tune with your numbers. and have trustworthy data collection and reporting in place. 

When you can forecast your sales, you can plan your cash flow, ad spend, and ultimately, ROI much better with less surprises. 


In terms of optimizing, we have been split testing much more than just landing pages. We have also run a few multi-variant tests as well. One of the things we really like to test are offers. You can never have a high enough converting offer - which is why the word "optimize" shouldn't have a past tense 😉 


We have used both VWO and Google Optimize for the split testing. I am personally a bigger fan of Google Optimize. The page flicker is basically non-existant, where as the VWO one is seen at times. And since I'm so big on Google Analytics, I'd much rather see the info ported natively into Google Analytics, instead of using up multiple precious custom dimension slots. 

In Conclusion...

This is one client that I have poured a lot of work into. I wanted to keep this proposal separate from the rest, as I have close ties with this client and didn't want to share a lot of their inside metrics, even though I got permission to do so. Its just out of respeonct to them and to the empire of a brand and team he has built. 

I respect that greatly and that is why it lives behind a password. 


This client was a perfect fit for myself and skill sets. I'm very much looking for more of these types of a clients 🙂 

Ready to Elevate Your Business?

Let's see if we're a good fit for one another.