Faster delivery of your data exports
One of the most common challenges faced by non-profit human services agencies is taking all the data that they generate each day and finding ways to use it to tell the story of their impact. Every day in iCarol systems across the world, thousands of pieces of data, are recorded. There are the calls, chats, texts, and other interactions where specialists record detailed information about those contacts, what they need and what services were directly provided, or the referrals made. Each referral record may contain over a hundred fields of information updated by resource specialists on an ongoing basis. There’s shift information, who worked and when, and how many hours each volunteer logged in a given period of time. Speaking of volunteers, in addition to hours worked there are the trainings and certifications they’ve achieved, new skills obtained, contact information, and the list goes on. That’s just a small sampling of the kinds of information being captured, and when you multiply just the things listed so far by nearly 70,000 users worldwide, it really adds up!
Much of your storytelling can be achieved through the dozens of charts, graphs, and other tools available in the Statistics area of iCarol. All of the most common data points our customers require are typically found there, and especially once you add in the available standard filters like location or time frame, or more custom filters from your call report fields, the information produced from these in-system reports is very powerful. Still, there are reasons you may need to dig deeper and export detailed files. Some take the raw call report data and place it into external programs for further manipulation or cross tabulation, or prefer to write their own queries. In the cases of fields where a text answer is provided instead of selecting from a dropdown or checkbox list, an export is needed to review the information that can’t be quantified.
Using iCarol you can export much more than your call report data. Detailed information on your shifts, volunteer information, client profiles, and your full database of resource records are all available to be extracted. Some centers have policies or contracts that require local copies of contact data or a hard copy of the resource database be downloaded on a regular basis. These are just a few of the instances where a data extract may be required.
Of course we take great care to ensure that this data is secured and protected, but also accessible to you when you need it. While we act as its caretakers, the data is owned by our clients and it’s important that they be able to download it for offline use outside of iCarol when needed. As you can imagine, these large amounts of data take time to retrieve and process. Our systems also contend with multiple simultaneous requests from various organizations across the globe. As iCarol grows and more and more non-profit agencies use it and add to this mountain of data, it’s our job to ensure that our high performance standards continue to be met.
Early this year, members of our technical team endeavored to apply a fresh approach to the processing of this data. We made changes to the service that generates those data exports in an effort to reduce the wait times between making the request and receiving the data. The team was able to accomplish this by optimizing queries, re-architecting the exporter service being used, deploying additional services, as well as making changes to the exporter schedule.
So with all this talk of data, it’s time to share some of our own. Following these changes to our export service, we decided to compare some numbers occurring in February of two different years: 2016, prior to any changes, and February 2017 after the new approach had been put into place. The table below shows this comparison. The “Time to complete” accounts for the number of minutes that elapse between when the extract request is made, and then the email notification of completion gets emailed. It’s also worth noting that there were over 100 more exports requested in February 2017 than in February 2016, and yet as you can see our average time overall and for the individual types of exports all decreased by significant percentages.
As time goes on we’ll continue to monitor this performance and look at comparisons using a larger sample size to get a fuller picture of the progress. This first look, however, is very promising and we’re hoping this has had a noticeable positive impact on our clients as they request these exports.
Tags: Data Exports
