At Loqate, we often talk about how we use data curation to produce the most comprehensive and precise premise-level address data in the world. In this blog post, our product manager Chris Harman, explains what we mean by data curation, how the process works, and how it increases address match rates.
Data curation is a term to describe the process and tools we use to analyse address data from different countries so that we can generate a golden global address record for our customers. For us, this means taking in data from multiple sources and combining that with industry-leading technology, human expertise, local knowledge, and cultural context.
The first step in our data curation process involves taking in data from multiple reputable data suppliers from across the globe. We aggregate data from several postal, geospatial and local sources to give us a consistent and reliable single view of a location. We work with 36 high quality data suppliers covering premise-level location data across 249 countries and territories in 6,500 languages. On average, four million additional premises are added to these datasets every six weeks.
As new commerce markets emerge and the world evolves, so too does our data. We’re always researching and evaluating, rebuilding, and finding additional suppliers. But we’re also extremely selective about the data suppliers we work with and will reject data if the quality isn’t good enough. We only take on a data supplier if it’s going to benefit our customers and give us the additional uplift and coverage we need.
What sets us apart from our competitors is our unique Persistent Data Hierarchy (PDH) process. This method of data analysis is what enables us to carry out comparative sampling when looking at new data suppliers and determine which data is best when combining multiple sources. In doing this, we’re able to give our customers access to the most accurate, pinpoint, premise-level data.
For example, let’s say we have two of the same addresses coming in from different data sources. That’s duplicated data, which we obviously don’t want. However, one is spelled incorrectly and is missing the right diacritics on certain words. Thanks to our unique PDH process, we are able to detect that inconsistency and select the address that is 100% correct and accurate. It’s this level of finite detail that no other company on the market can achieve.
Our data curation process would be nothing without our data engineers who spend over 3,000 hours a year researching, studying and evaluating global addresses to make sure we’re bringing you the best data quality.
They’re also the driving force behind our intelligent machine engine, inputting the algorithm, developing the logic, and determining the mapping.
Global addressing is complex, and there are always exceptions to the rule that can’t be picked up by technology alone. For countries that have complicated addressing infrastructure, it requires our engineers to analyse results and determine what is the best outcome for the customer.
The completeness of our global data coverage, the capability of our AI parsing engine, and our in-house global addressing expertise, means we’re able to deliver the most comprehensive and precise premise-level address data for every country and territory in the world.
Together, these factors significantly improve address match rates and uplift precision in existing, as well as hard-to-reach address markets.Compared to competitors, on average, we offer:
We give businesses across industries the unrivalled precision and reliability they need to make location data-driven decisions, deliver superior customer experiences, help prevent fraud and enable cross-border commerce – globally and at scale. See our technology in action with a free demo.