How to enrich first party data
By Claes Nygren
In this two-part series, we’ll discuss the importance of owning your own data—or first-party data—and how to enrich that data when you own it. In part one, we explain what first-party data is, how to collect it, and why owning your own first-party data is important. In part two, we’ll explain what it means to enrich data, how this can help your business, and how you can start the process of enriching your own first-party data. Let’s talk about how to enrich your first-party data.
Once you have your first-party data, what do you do with it?
The most important thing you can do for your customer is to know what they want. And there is no better way to give your customer what they want than by knowing who they are. One way you can do that is by data enrichment. Data enrichment is the process of complementing the first-party data you collect with more meaningful information, allowing you to better define who your customer is and what they’re looking for. Here’s what we mean.
As an online retailer, you may have basic visitor information such as the following:
- IP address
- Device type
- Date/time of site visit
- Site activity/user flow
- Purchase information: Product SKU(s), Gift vs personal purchase
Through various methods, these data attributes can be enriched with associated non-identifiable information like these:
- Geolocation and related info: Population, geography, government type, language(s), community type (urban, suburban, rural)
- Age range
- Income level
- Occupation and educational background
- Marital status and family size
And with these enriched data sets, you can answer powerful questions:
- Do gift purchases follow seasonal or time-related trends?
- Are certain products selling better in one region over another?
- Does age or education affect how customers navigate my site?
- Which product is a customer most likely to purchase?
Make sure that this information can never lead back to the original customers, since that can violate GDPR and privacy rules, and also reflect poorly on your company. A way to ensure the anonymity of this information is to disconnect the source while not losing the integrity of the information. This is so important to ensure that this data doesn’t end up in the wrong hands, hurting both you and your customers.
Retailers big and small can serve their customers better by enriching their data and using it to guide their decisions. Many data companies offer robust solutions, and if you’re browsing those options, it can feel daunting. Instead of feeling like you have to run in the data enrichment marathon without training, we’d like to encourage you to get up from the couch and walk instead of run. Over time allow that walk to develop into a slow jog, and before you know it, maybe you’re running a half marathon. Assemble wants to help you find ways to enrich your data over time sustainably.
How do you enrich your data?
Regularly looking at your data will help you know what to pay attention to as a retailer. As you examine your traffic, you’ll start to notice small things that begin to tell you more about your audience.
The first step is to associate traffic with an IP address to get a geographical location; now you know where your customers spend their time. From there, you can tap into public data sets to pull demographic information related to those areas. (In the U.S., census.gov or bea.gov are great resources.) You can use software or enlist the help of an engineer to gather this information and connect it to your first-party data.
Side note: Find a trusted partner to work with
We’d highly recommend working with someone to either choose software or build your enrichment program internally. When it comes to software integrations, tech debt is a real thing. You don’t want to start using a system that doesn’t integrate with your current tech stack or creates duplicate (or dirty) data. So make sure you get the right person to help you find a solution to enrich your data.
After you have a geographical location associated with the IP address, you can associate other info with that data. Your IP provides geolocation (like a country code). Geolocation can help determine the region’s population and information like distance to the coast, mountains, etc. Third-party data sources can then help you understand income distribution and other demographic data for a particular zip code, enabling you to explore information that may affect your product viability in that area.
How to enrich and analyze first-party data?
Once you have your data, it’s essential to be able to interpret it in a way that provides you with actionable information. To do this, it is helpful to first know your goals. Do you want to increase sales of a particular product line? Do you want to maximize customer ratings or net promoter scores? Do you want to decrease the average time to purchase? Perhaps you simply want to observe user behavior and hypothesize why certain events are happening. Knowing your priorities going in will make your data exploration much easier. With your data set and high-level goals in mind, you can analyze your data and put it to use.
Interpret the data through visualization
Humans are naturally visual creatures. You can use many tools such as Excel, Tableau, Power BI, or others to slice and aggregate your data and convert raw numbers into visual representations. Graphs and easily digestible tables will not only help you see what’s happening but also help share your findings with your team. Custom dashboards will allow you to run analyses routinely and monitor trends over time. You can use these visuals to uncover insights that guide impactful decisions like building location-specific landing pages, drafting targeted ad copy, and expanding or limiting product offerings.
Confirm your suspicions with statistics
Even with a beautiful graph, data can be misleading if not backed by statistical testing. Statistics can help you answer questions like, “Is customer satisfaction really higher in our Austin branch than in Los Angeles?” or “Are colder temperatures correlated with increased jacket sales?” Your analysis and visualization software can likely run these statistical tests for you, and if you’re not sure which tests are most appropriate, consulting a data scientist or statistician may be useful, especially before making any big decisions.
Predict customer behaviors with machine learning
Not all insights from enriched data sets can be visualized in two-dimensional dashboards or reports. Machine learning algorithms can analyze multiple parameters in parallel in order to predict outcomes (purchase or no purchase, expected product rating, etc.) and help you create dynamic, personalized experiences for your customers based on their expected preferences and behaviors. As you collect more data over time, the accuracy of these algorithms will increase and enable an ever-improving customer experience.
Try and try again
Don’t worry about being 100% accurate. In fact, you never can be. Data is inherently noisy, and customers change over time. The trends you observe in one month may not hold in the next. Data collection, enrichment, and analysis is all a cycle, and it’s important to treat this as a learning process that allows you to make small adjustments over time that will help you better understand your customers and grow with them. As you have probably guessed, radical change does not happen overnight, but the results you will see in the long term will help you improve your business exponentially.
Why is data enrichment important for retailers?
It all goes back to knowing your customers and providing products they want and need. The questions you’re asking should help you find ways to serve your customer better. Using your data to inform your business decisions will improve customer experience, decrease shipping times, ensure stock availability, and provide products to your customers when they want them. And serving your customer well leads to better business.
Need help enriching your data?
If you’d like to enrich your data but don’t know where to start, we’d love to talk. While Assemble may not be the right solution for your business, we will help you find the right tools or people to make your data enrichment journey a success.
This article was written in collaboration with Eric Tucker and Rachelle Cummings.