Fortunately, building a data infrastructure for tracking analytics and KPIs across multiple Shopify stores and third-party services is easier than it’s ever been. Unfortunately, it’s still not easy. Or quick. It takes time to discover which metrics and KPIs are most important, and some further steps in the process still require the intervention of developers or data scientists. Just like crops from the soil, data is dirty and needs to be cleaned and treated before it can be consumed for all the good nutrients inside. If it’s not, your outcomes will be unreliable.
By the time you finish reading this article, you will know what you need to do to wrangle your business and marketing data from multiple Shopify stores and third-party services, as well as how you can do it, or at least—in cases where implementation depends on your individual needs and data setup—what to consider when doing it. Let’s dive in.
Note: The focus of this article is business and marketing analytics. For more information on how to integrate other types of data across Shopify stores, like product data or warehouse and inventory data, see our blog on Shopify multi-store data integration.
- Determine what metrics you should track
- Set up an analytics tool and data layers via Google Tag Manager
- Extract, transform, and load your data into a data warehouse
- Send your data to a BI tool and create dashboards
1. Determine What Metrics You Should Track
It’s extremely important that you take some time to think about what metrics you want to track before you start integrating your data. There is a seemingly infinite number of metrics out there, however, and many of them can only be tracked by combining individual data items generated by your Shopify stores with items generated by the other cloud-based applications you use. So, how to identify clearly what’s most important for your business? Only you can answer this question, but here are some pointers to get you started.
#1 - Calculate rates and averages instead of just tracking totals. It may be exciting to see your total number of email subscribers or social media followers go up, but the truth is that these numbers usually go up over time anyway. In the long run, they are just vanity metrics that don’t really indicate anything about growth.
Instead, focus on rates and averages. Keep in mind that many of the rates and averages you will want to track, like return on ad spend (ROAS) and net promoter score (NPS), will need to be obtained by combining data from Shopify with data from other platforms.
#2 - Assign time intervals to your metrics. While it may not make much sense to track your total number of email subscribers over time, it does make sense to track the number of new email subscribers per week or month. When the numbers across intervals start to go up, you know that you are growing.
#3 - Define clearly which department, team, or individual is responsible for each metric. Things that are measured and monitored tend to improve, but only if clearly assigned. When multiple entities are collectively responsible for a given metric, the success of one could overshadow the shortcomings of the others, making it hard to detect any inefficiencies. In such cases, it’s best to break the metric down into smaller metrics that each entity can be accountable for individually.
Dashboard courtesy of dried meat marketplace Maso Here. Read our case study to learn how they determine ROAS for 3 eshops.
For ecommerce businesses, good rates and averages to track are:
- Return on ad spend (ROAS)
- Customer acquisition cost (CAC)
- Customer lifetime value (LTV)
- Customer retention rate (CRR)
- Average order value (AOV)
- Net promoter score (NPS)
Keep in mind that, in the world of big data, less is often more. Collect only the data you need so your dashboard doesn’t become too cluttered.
2. Set Up an Analytics Tool and Data Layers via Google Tag Manager
In the days of old, adding these codes/tags to web pages was something that you had to ask developers to do for you. Now, thanks to Google Tag Manager (GTM)—a central console that lets you control and deploy all your tags—you can do much of the work yourself. (Keep in mind, however, that even though GTM is made for business users, you still may need a developer to add a special “container code” to your pages. Or you can add a pre-made container code using a Shopify app, which we will suggest below. All other tags can be easily added through the GTM interface.)
Think of the data layer as the “front line” of your data collection process. Every tool you connect to your website accesses it, and it ensures that all tools get the data they need and that the data is the same across them.
Example image courtesy of cf.agency.
Be sure you install GTM and data layers for each of your Shopify stores. How exactly to do this will not be discussed in depth here, as there are plenty of online resources to help non-technical users through the process, like this GTM + data layer tutorial for Shopify or this data layer code block, which can be copied and pasted directly into your Shopify order processing page.
3. Extract, Transform, and Load Your Data into a Data Warehouse
The above steps will get you properly collecting traffic and behavior data, and feeding it into Google Analytics. But this is not the only data you’ll need to build comprehensive reports. You will also need:
- Sales data from each of your Shopify stores
- Advertising data (e.g. from Google Ads, Facebook Ads)
- Customer service data (Zendesk, Freshdesk)
- Financial data (Stripe, Rebound Returns)
- Other customer data (Salesforce, Mailchimp, etc.)
To make meaningful use of all this data, you’ll want to send everything to a data warehouse first. Why? Two reasons.
Number one. Sending large amounts of data directly to a dashboarding application like Google Data Studio might cause the app to crash. Moreover, dashboarding apps don’t store data, they simply visualize it. And most third-party services only allow storage of native data for a limited period of time.
Number two. You can’t flexibly combine data from third-party services by plugging them directly into a dashboarding app. In fact, many services can’t even connect to dashboarding apps at all.
So, without a cloud-based data warehouse like Google BigQuery or Amazon Redshift, which can store essentially limitless amounts of data that can be queried on demand, you won’t reap much benefit from using a dashboarding app.
For more information on how to choose a data warehouse that suits the needs of your organization, jump to section 3 of our blog on how to build a data infrastructure for an ecommerce business.
After you’ve chosen a data warehouse, you can start funneling your data to it via pipelines, which allow you to automate the transfer of data from sources to destinations at regular intervals.
With ETL (extract, transform, load) platforms like Dataddo, you can funnel datasets from any number of Shopify stores or other cloud-based services to any number of destinations without coding experience. Here is how it works:
Step 1: Create a source by selecting Shopify from Dataddo's list of connectors.
Step 2: Select your dataset.
Step 3: Select your metrics and attributes.
Step 4: Configure your snapshotting preferences by choosing your date range, sync type, snapshotting frequency, time, and timezone. You can also load historical data in this step.
Step 5: Start creating flows.
Once you have flows set up, Dataddo will extract and transform the data you’ve selected from its sources at regular time intervals, and load it into your data warehouse. What does it mean to transform data, you ask? Good question.
Data, when extracted from its sources, is unstructured, i.e. just lines of code. In order for this data to be readable by humans or machines, or to be stored in data warehouses, which only accept structured data, it must undergo transformation. Dataddo performs three kinds of data transformations:
- Data mapping. Think of data mapping as the conversion of data from lines of code to line items in a spreadsheet.
- Data harmonization. Even after datasets are mapped, they are still hard to consume because they often contain duplicates, items in different formats, or informational gaps. Harmonization cleans up these messes.
- Data exclusion. This transformation eliminates sensitive personally identifiable information (PII) from your datasets before your data is funneled to its destination, keeping your company operations within the law.
The first two transformations happen automatically “under the hood,” while the third can be selected optionally by the user. But all three are important to be aware of because they are essential to the maintenance of good data quality, especially for ecommerce businesses collecting data across multiple Shopify stores and cloud-based applications.
Moreover, they save time on any additional transformations you might need to do after your data hits the warehouse, such as data joining, data grouping, or window functions. These, however, will require the involvement of a data scientist. Though further transforming data once it gets to a data warehouse is a common and beneficial practice, it may not be necessary depending on the data needs of your organization.
4. Sending Your Data to a BI Tool and Creating Dashboards
A Google Data Studio dashboard.
Regardless of whether further transformations take place in your data warehouse, your data will need to be sent to a dashboarding app/business intelligence (BI) tool for interactive, visual tracking of analytics metrics and KPIs. In all likelihood, you will be using a data warehouse that is natively compatible with your BI tool (for example, Google BigQuery and Google Data Studio), in which case sending data from one to the other will be straightforward. In case you are not (for example, Mongo DB and Google Data Studio), an ETL platform like Dataddo can be used to connect them, but we recommend going direct.
BI dashboards are typically designed for use by non-technical business users and offer a wide range of visualization components that can be customized for the needs of any audience. Just like the metrics you track, only you can determine how best to visualize your data. For tips to guide you along your way, see section 6 of our blog on how to build a data infrastructure for an ecommerce business.
The Results Will Be Worth It
The data infrastructure described above is completely flexible, as each of its elements can be scaled up or down as the needs of your business change. Even if it takes you some time to determine your key metrics, understand and install GTM and data layers, or learn how to create dashboards optimized for the needs of your organization, the results will be worth it.
See how it’s done in practice by reading our case study on UK brand Lounge Underwear, which successfully built a data infrastructure across 9 Shopify stores.
See how Dataddo can help you get the most out of Shopify
Funnel all your ecommerce and marketing data to your dashboard in just a few simple steps—for better analysis, without the hassle.