Disclaimer: This is a personal post on a personal blog
. I’ve been working in the analytics industry as a consultant for a few years and have gotten a feel for some of the challenges that business face. These are my personal ideas and do not reflect those of my employer.
Now that all the excitement has started to subside around Google’s Universal Analytics announcement last week I wanted to take some time and talk about how the system can actually be used.
I think it’s important to remeber that Google Anlaytics is a platform. The current Google Analytics does a great job of measuring the digital world.
The goal of Universal Analytics is to give a business insights into their entire world. This might be online marketing, sales from a store, catalog orders, app usage, traditional ecommerce tracking, etc.
To understand the platform let’s walk through how a real business might use Universal Anlaytics.
Note: Whenever I give a talk I like to infuse it with a personal touch, which often includes a reference to Vermont, a place I love dearly. And that’s what I’m going to do here.
There’s a great company in Vermont called Gardner’s Supply. As the name implies they provide solutions for gardeners. They’ve got an innovative product line that’s created right here in Vermont.
Dislaimer: I have NO connection to Gardner’s Supply other than I’m a customer and a big fan.
Gardener’s Supply sells via physical store, catalog, online and phone. They have a desktop site and a mobile version of their site. Like many commerce-based businesses they have a shopping cart that supports a login.
This is key. Whenever I make a purchase, regardless of the mechanism, they can connect my purchases together into a customer record.
They’re what you would call a multi-channel retailer.
They also have an amazing app for iOS. In fact, I think their app is what makes them so brilliant.
Mobile App Strategy
The mobile app, named Garden Minder, is a garden management app. It reminds you when to water, when to plant, etc. All you need to do is set up your garden in the app.
If you’re new to gardening you can input the physical location of your garden and it will recomend a number of gardens based on your culinary taste and climate.
Just think, if you’re a new gardener, the app will help you get started by suggesting the right plants. It will also make sure that you’re caring for your garden by alerting you when it’s time to water, harvest, etc.
If you’re a seasoned gardener then the app will help you plan your garden and remind you of important tasks.
In case you’re wondering – I’m an average gardener. But I’m getting pretty darn good ;)
These guys are providing a valuable service to their target audience: people interested in gardening. It builds trust with consumers and always puts the Gardener’s brand front-and-center for gardeners.
Oh, and by the way, if you register with Gardener’s Supply using the app you’ll get a coupon for use on your next purchase.
A benefit for everyone. All you need to register is an email address.
You save some money and the business get a better way to connect to their target audience.
Don’t forget, you can buy via physical store, catalog, online and via phone. When you put all of these pieces together you get true multi-channel retailer.
How Universal Analytics Will Help
So how does all of this relate to Universal Analytics?
The goal of Universal Analytics is to make the user, the customer, the center of the data. In the case of Gardener’s Supply they coud use the user ID feature of Universal Analytics to align all of their data, across all of their platforms, with users, not session.
ASSUMPTIONS: I have NO idea what systems Gardner’s supply uses or how they current measure their business. Everything below is a guess. But a logical one.
I’m going to assume the most complex scenario, six different data sources: store (point-of-sale), desktop website, mobile website, mobile app, call center and catalog. Each of these systems is a completely different system and the only shared data is an inventory/order management system.
If we’re going to implement Google Analytics it means that data must be sent to Google Analytics from each system.
Let’s look at how Universal Analytics might be integrated into each system.
Point of Sale system:POS systems are a fancy way of saying cash register. They’re the systems that handle in-store purchases. With the new Measurement Protocol transactions from these systems can be tracked in Google Analytics.
Each time a customer completes a transaction in the store the transaction must be sent to Google Analytics along with the customer ID. Obviously this will take a bit of coding. When Google Analytics gets the data it will aggregated with other data for this user.
I can also send a pageview or event to track the transaction as a goal completion.
Remember, we’re not tracking this individual. We’re just using the ID as a key to loop all the data together. We don’t have access to this ID anywhere and the user remains completely anonymous.
Desktop Website: This is probably the easiest tracking to install. We can use the new
analytics.js library to track website activity. We can also implement the user ID feature to sessionize the data by user across all data sources. This means that we’d need an encrypted customer ID present wen the page is built so the ID can be passed to analytics.js.
Mobile Website: Very similar to the desktop website. Use the new
analytics.js library to track mobile website users. Also implement the user ID feature to sessionize the data by user across all data sources. Again, the encrypted customer ID will need to be present when the page is sent to the browser the ID can be passed to
NOTE: Tag Management: I should also mention that this would be a good time to implement a tag management solution. If you’re new to tag management, it will help you better manage all of your advertising and analytics tags. There are a lot of options out there, including Google Tag Manager, which is free. So don’t complain about cost :)
Mobile App: This case is very similar to tracking a website. We can use the new iOS mobile SDK to track visitor use of the app.
Remember, when you’re tacking an app with Google Analytics you need to define screens and events. It’s a fundamentally different data model because apps are different than websites.
We’ll also need to implement the user ID feature in the app tracking to join the app data to the desktop web and mobile web data. Again, the user ID will need to passed into Google Analytics tracking.
BTW – if you want potential customers to self identify you need to give them some reason. That’s where MARKETING comes into play. How can you provide value to someone that is not a customer?
At Gardener’s Supply they’re offering a discount if you register with them. It’s up to you, and your business, to figure out how to provide value to your customers.
Call Center: This is where the implementation starts to get complicated. Using the Measurement Protocol the call center can send conversion data back to Google Analytics. When a customer calls the call center the measurement protocal can send almost any type of data. It’s up to the business to define the data.
Here are some of the ideas that I have for using Google Analytics in a call center:
- Track general calls into the call center. This could be done using a pageview or an event.
- Track the types of calls. Can be done using the name of the pageview or the Label in the event. How you discern the call could be difficult. I know some companies that are using speech recognition software to mine call center conversations for more data. You could integrate Google Analytics into this system to automatically track the type of call based on certain keywords.
- Track customer purchases via call center. Probably the most important thing to measure from the call center. Could be done with a pageview or an event for the goal conversion and commerce tracking for transactions.
- Track product returns/refunds via call center. Here’s another area where the measurement protocol will really help Not only can you track transactions, but also returns and canceled orders. Just send Google Analytics the altered order using the same order ID and the value should be updated in Analytics.
Catalog System: Not sure what we need to do here. But if there is another system, separate from the call center that processes catalog purchases, then the data needs to be sent to Google Analytics.
Luckily, most catalog retailers have a unique ID on the catalog, or the order form, that is used to join the data. When a catalog transaction is received the system would need to send a Google Analytics ecommerce transaction along with the customer ID to join the transaction and revenue to other customer data.
You could also use an event to tell Google Analytics that a catalog was sent to the visitor. This would be a really interesting way to measure the influence of catalogs sent to users.
Importing Other Data with Dimension Wideneing
Ok, so all of the above Google Analytics tracking customizations (using the Universal Analytics Measurement Protocol) make the data user-centric. Now let’s look to Dimension Widening, another feature in Universal Analytics, that can help add more context to the data.
With Dimension Widening we can import other data sets and join them with the existing data in Google Analytics.
I think there are two main data sets that would help here:
First, product data. There may be instances where we need to translate a value (like SKU) to something more human-readable. Just think about the content reports in Google Analytics. You can import simple data, like product name, or you can import more complex information, like the margin for each product or the shipping time and cost.
Second, customer data. We don’t need to import a lot of customer data. We’ll already have transactions regardless of where they come from. But let’s think about this specific business.
What customer data would provide deeper insights into their behavior?
This is where you need to know the industry, the business and the customers.
I tried my best to come up with some interesting attributes that I would want to see if I ran a gardening company.
- Hardiness Zone: Anyone that does gardening, or loves plants, knows that the US is divided into hardiness zones. These zones help people understand what plants are best suited for their location. I think this is a FANTASATIC example of adding business specific data to Google Analytics. If I was a marketer at Gardener’s Supply I would want to understand my customer behavior by hardiness zone. It would make me a better marketer.
- Garden Size: Another piece of information that would make me a better marketer. If you’re a gardner in New York city, then you’ll probably want information about container gardening or rooftop gardening. But if you’re a gardener in Vermont, with 1 acre of land, you might want to know about more industrial types of products.
- Socially Connected: Many companies are tracking friends and followers. I don’t really care which network we’re connected on, but I would love to know if we’re connected in my analytics data.
- E-Newsletter Subscriber: A simple identifier to track who is a subscriber to the email newsletter.
- Catalog Subscriber: This is an interesting one. I could also get this data using the measurement protocol and tracking who receives a catalog.
All of this anonymous data can be joined using a customer ID.
Don’t forget, one of the most exciting types of data to import is cost data. That’s something that you can actually start doing now.
Using Universal Analytics to do Analysis
Ok, so now that we have all of this data, what do we do with it!?
That’s a bit of a tough question. Many of the Universal Analytics announcements focused on the platform. The reports are coming.
But one thing we know that is coming soob is User Segmentaiton. This feature, actually called Unified Segmentaiton, will let you segment your data based on users or sessions.
Using User Segmentation you’ll be able to perform some pretty standard segmentation tasks around different channels, products and marketing campaigns.
- Segment customers based on first purchase value
- Segment customers based on first purchased product line
- Segment customers based on first purchase marketing channel
- Segment High-value customers
- Segment customers that have not bought during the gardening season
- Segment customers that have bought more than X times
- Segment customers based on hardiness zone
- Segment customers based on garden size
You can probably tell that I’m really excited about the user segmentation. I think this is going to be a tremendos value to analysts and marketers.
If I can understand the behavior of people with different sized gardens, I can market to them more effective.
If I know the hardiness zones, I can market to them more effectively.
If I know which devices they use during different parts of the day, I can provide a better advertising experience.
So many opportunities.
I know this post may be a little thin on details, but I wanted to give you all some ideas how you can use Universal Analytics.
I look forward to hearing how you might use Universal Analytics to measure your business.