I want to make data-based decisions about my site in a cost-effective way.
This post walks through a simple example of:
How did you calculate Per Visit Goal Value?
In other words,
What number do you enter for goal value in Google Analytics?
Key Performance Indicator
The site in this example has two goals:
- Sell Hats
- Email Newsletter Sign Ups (not covered in this post)
My Key Performance Indicators would be
- % Hat Sales per Visit
- % Email Newsletter Sign Ups per Visit
By making each of these activities an Event or a Goal within Google Analytics, I can measure them. If my marketing efforts are successful, each should rise over time.
If I assign dollar values to each event and goal and segment my data, I can see what channels or content drive the most value and then use that information to make decisions about how to hopefully increase % Hat Sales per Visit.
Here are a few examples from the site on which this example was based:
- Keep the photo gallery of available colors and the size guide page. These have the highest Page Value of any pages on my site (except for the two product pages). In the future, I should be careful how I change these pages, because they work so well.
- Q&A sites and animal hat-related sites generate more sales than Facebook, which leads to two decisions:
- Participate in more Q&A sites and get more links from animal hat-related sites. Set up campaign tracking for these.
- Facebook has a big audience. Diagnose our presence there and refine our measures of success there. Facebook may be a better source for engagement.
A Simple Example
Below I walk through the steps I took to calculate Event and Goal Values.
Understand How my Business & the Site Work
My business sells one hat that comes in two styles:
- “Rats” for $10
- “Cats” for $20
Each style comes in a variety of colors and sizes.
Each product is ordered in a combination of size and color from its own page on the site. For example, my customer might choose a grey Cat hat in size small.
Customers choose a color and size combination then click the “add to cart” button. Clicking on the “add to cart” button takes the customer to a third-party shopping cart (i.e., it’s not on my domain).
In the shopping cart, she can:
- choose a quantity,
- check out (i.e., completes her purchase),
- shop more on my site (hopefully!) or
- abandon her shopping cart (sad).
Understand Google Analytics Tracking Capabilities
I cannot put the Google Analytics tracking code on the third-party shopping cart, so I cannot tell:
- the quantity bought,
- whether or not my customer completes her purchase, or
- what additional items she buys (though I can get that information from my shopping cart application).
Within Google Analytics, I can track when each of the two “add to cart” buttons are clicked. Google Analytics’ event tracking is perfect for that. I can count “Rat” and “Cat” hat clicks and assign a value for each click.
I can then create goals that are tied to the events, “Rat” and “Cat”, and assign values to those, too. Because events and goals are included in so many of Google Analytics’ standard reports, I get a lot more information for very little extra work which lets me do more value-added or fun things with my time!
I can set a goal that is triggered by an event, however a couple of limitations exist:
- Google Analytics counts only one unique goal per visit, i.e., one “Rats” goal will be measured even though multiple “Rats” events may have happened. In other words, a customer may purchase the same product in two different color and size combinations and that would count as one goal and two events. What value would I assign to the goal in that case?
- I can’t track the unfortunate instances where customers put products in their cart and then do not purchase them (i.e., these customers either subsequently remove them from their cart or they do not complete their purchase at all). In these sad cases, events and goals would be recorded, but there would be no revenue.
These constraints mean I cannot simply use my price or gross product as the event or goal value. I have to find a work around.
Calculate the Event & Goal Values
I have an easier time understanding concepts with tangible examples. Maybe you do also.
Configure Google Analytics
I need to configure Google Analytics to track sales as best I can. In this case, I:
- Set up Event Tracking to track clicks on each of the two “add to cart” buttons: Rats and Cats.
- Configure two Goals: Rats and Cats that are triggered when the respective Events happen.
I am not assigning values to Events and Goals at this time. I only want to count them.
Here’s what Google Analytics measured:
(The best way to find out how to configure Google Analytics for Event and Goal tracking is to Google that.)
Note: I would do all this in a “Test” profile in Google Analytics before copying it to my “Main” profile (where I do all my analysis).
Get Sales Data
During the same time period as in Step 3a, these order were placed:
This is the data I downloaded from my shopping cart application for those orders:
Comparing actual customer behavior in the first table above to what my shopping cart tells me the second table and what we know Google Analytics will measure (Step 2) illustrates that that setting the Event Value or the Goal Value equal to the selling price would be inaccurate.
In this example, if I set the Event or Goal Value equal to the selling price, I would undercount for Borat’s order and overcount for Pat’s order.
Calculate Event & Goal Values
Here’s how I did calculate the event and goal values:
Note: Event values must be set as integers in Google Analytics or your event won’t be tracked, so round those.
How Much Data?
This example has far too little data. With a lot of data over a long period of time, this general calculation should work as long as future customer behavior mirrors past behavior.
I just completed the third iteration of this same calculation. I started with one month of data, re-did it with six months of data and just re-did it using a full year of data. That’s important because Christmas is a big selling opportunity. With a full year of data, the event and goal values each changed by about 5-15%.
More data is generally better, but we need to make decisions every day. The decision to maintain the status quo because the data or analysis is not perfect is a decision that has real effects.
The data and analysis – including the time period used – need to be just good enough to make decisions. Not perfect!
In my years of financial securities analysis, even the most sophisticated valuation models were only guides for decision making
That said, past performance is no guarantee of future performance. I would caution against using years and years worth of data. Surely, my business, industry or website were very different five years ago, than they have been to the past year, so using five years’ data may lead to poorer estimates.
More Than Two Products?
Or a lot of products? Google Analytics and many consultants have suggested using e-commerce tracking even if you do not have e-commerce on your website.Google+