Read this case study to find out how Matrix improved their data by outsourcing data management to Hopewiser.
Matrix Telematics provide fleet management technology, ranging from hardwired/plug-in black box telematics, remote access CCTV and a direct to driver behaviour management solution. These devices provide key insights to optimising fleet performance, while reducing occupational road risk, which contribute significantly to road incidents and fleet running costs. By implementing these devices their clients have experienced savings in fuel, maintenance and reductions on insurance premiums.
Matrix was collating data from all different sources in an attempt to give a single view on the individual – especially regarding past accidents, defaulting and identity fraud. A “single view database” will be invaluable to the industry due to the ever-increasing volumes of data. It will allow their clients to offset risks in a better way.
Due to the many sources of data, Matrix needed to verify and standardise. They also receive data on an on-going basis so it is growing all the time.
Matrix decided that they did not want to buy in cleansing software, or manage the process in-house. Their preference was to use a full Bureau Service which means they can securely hand over the database for verification and standardisation.
The data was of extremely variable structure and quality due to the disparate sources. However, using AtlasBatch, an automatic data cleansing and enhancement tool, Hopewiser was able to verify and correct large quantities of address data. AtlasBatch read the database of existing addresses and cross-matched them against the Royal Mail’s Postcode Address File (PAF). It then assigned a unique personal reference number to each record, so that Matrix could anonymise the record, should this be needed.
The data was run through a second process, using AtlasDedupe, a sophisticated tool for intelligently identifying and eliminating unwanted duplicates at both individual and property level. Duplication of information is a common issue for anyone maintaining a centralised corporate database or multiple departmental data sources, which is exactly what Matrix had.
The ResultDuring this process, Hopewiser identified that over 97% of the data was duplicated. By identifying this, Matrix was able to hone the data down into a usable format that is now highly accurate. To identify this many duplicates manually would have taken months and months of hard work, but as Matrix had a very tight timescale for getting the job completed, Hopewiser was able to run 37 million records during a weekend and get the data prepared quickly.
Future ProofingIn addition there is an on-going requirement for an online service, which can capture and verify an address at the input stage. Matrix estimated that this could be in the region of 100,000 requests per calendar month. Matrix opted to supplement the data cleansing bureau work by using Hopewiser’s Address Lookup online service. Matrix chose Hopewiser as their preferred address lookup tool as they were able to guarantee the API could handle the volume of lookups, without any service issues. This means they can accurately capture address data at source so that there are fewer problems with new data further down the road, resulting in smaller data cleansing and deduplication requirements in the future.
“We have worked closely now with Hopewiser over the last few years, to ensure our data is the best it can be. Merging our disparate data across systems has been a real benefit to the business which means we can offer this as a service to our clients. The single view allows us to manage risks such as identify fraud in a far more efficient way. Outsourcing the data management processes to Hopewiser means that the job can be done in days as opposed to months. We have saved time and money using Hopewiser and have greatly enhanced the reliability of our data. We continue to save time by now checking every address as it is entered which means our data will be far more accurate in the future.” Mark Packman, CEO Matrix
, updated 16th August 2022.
Topic: Data CleansingDeduplication