In this piece, I describe some of the key aspects of an organisation that has been on the course of setting up an entire country's animal welfare operations for success by enabling better data collection practices! If you would like to know why I care deeply about this, head over to [[Fur & Figures - Navigating the World of Animal Shelter Analytics]] for some context. ## High-level view of Shelter Animals Count [Shelter Animals Count](https://www.shelteranimalscount.org/) is an organisation in the U.S., advocating for enhanced data management in animal shelters. ### Overview of Shelter Animals Count Shelter Animals Count is on the [purpose of establishing a comprehensive national database in the US, dedicated to animal welfare](https://www.shelteranimalscount.org/leading-national-animal-welfare-organizations-launch-shelter-animals-count/). Their mission, as detailed in [About the Data](https://www.shelteranimalscount.org/about-the-data/), is to foster a data-driven approach to saving animal lives, by providing reliable and accessible information. Shelter Animals Count motivates and incentivises shelters to contribute their data, fostering a collaborative environment for improving animal welfare across the country. ### Key Databases and Resources #### Intake and Outcome Database - The [Intake and Outcome Database (IOD)](https://www.shelteranimalscount.org/intake-and-outcome-database-iod/) plays a pivotal role in tracking the journey of animals through the shelter system. - Database with minimum requirements expected of an animal welfare database as accepted by several other similar orgs in the US. - Basically, just a tool for data collection. Formerly called the "Basic Data Matrix" - [PDF sheet with further explanations](https://www.shelteranimalscount.org/wp-content/uploads/2022/07/SAC_IntakeandOutcomeDatabase_IOD_070522.pdf) > At shelter level, animal level data (ALD) is collected. It's basically collecting all relevant info of an animal which will have an `animal_id` assigned to it. This `animal_id` is not unique as duplicates indicate the animal moving to and from the shelter in multiple times. #### Community Services Database - The [Community Services Database (CSD)](https://www.shelteranimalscount.org/community-services-database-csd/) supports shelters in managing and optimising their community-oriented services, offering a broader perspective on animal welfare beyond the shelter walls. - Started being used only since 2021 - [PDF sheet with further explanations](https://www.shelteranimalscount.org/wp-content/uploads/2022/07/SACCommunityServiceDatabase_CSD_R070522.pdf) #### What kind of species are tracked? The following graphic on [the website](https://www.shelteranimalscount.org/intake-and-outcome-database-iod/) indicates that data collection is not limited to just felines or canines, as is the case in most shelters.  ### Challenges and Paths to Resolution Data accuracy is crucial in animal shelter management[^1]. The article [Navigating Data Entry Errors](https://www.shelteranimalscount.org/ald-blog-series-navigating-data-entry-errors/) sheds light on methods to correct conflicting data around intake types and establish a single source of truth, which is essential for reliable data. Similarly, [Overcoming Lack of Standardized Data Entry](https://www.shelteranimalscount.org/ald-blog-overcoming-lack-standardized-data-entry) discusses the challenges of non-standardised data entry in shelters. This is definitely interesting as every shelter could have their own interpretations of metrics and methods of data collection. By standardising definitions and encouraging consistent data tracking methods, the organisation aims to enhance data quality across the board. ### Utilisation of Data for Insights #### Dashboards Shelter Animals Count provides comprehensive [Data Dashboards](https://www.shelteranimalscount.org/data-dashboards/)[^2], offering insights into industry trends and key metrics like intake, outcome, and community services, visually represented for easy understanding and analysis. The dashboards are done in Tableau. #### Reports and Studies The organisation publishes [Reports](https://www.shelteranimalscount.org/reports/) that summarise key findings and trends in animal welfare. Additionally, their [Scientific Studies](https://www.shelteranimalscount.org/data-in-action/) section highlights how data is being used in research to advance animal welfare practices. ### Tools and Guides for Data Management #### Animal Welfare Glossary The [Animal Welfare Glossary](https://www.shelteranimalscount.org/animal-welfare-glossary) is a valuable resource for building a data dictionary or catalog, ensuring that all stakeholders have a common understanding of the terms and metrics used in animal welfare. #### Data You can get the data for a fee from here - https://www.shelteranimalscount.org/request-the-data/. If you are getting data as a "data enthusiast", you can get a CSV file with aggregated ALD data at the state level from 2019-2022. ## In Conversation with Samantha Hill - Director of Data & Tech at Shelter Animals Count Samantha Hill is Director of Data and Technology at Shelter Animals Count (SAC). She spearheads the vision of collecting, standardising and providing shelter data at a national level. Recently, I had the opportunity to pick her brain. > In her words, SAC perform the task of **specialised data curators** who make animal shelter data publicly available for the rest of the nation. ### Key Points - Collecting shelter data in a standardised way across the nation offers key benefits - Useful for national fundraisers to identify key areas of growth and focus - Shelters are able to use the data to understand their own needs from an organisational standpoint - Accurate data => Accurate references for the media to help spread important awareness - Key challenges to the vision - Data collection - Knowing which attributes to collect is a question that can't be completely answered at the time - Nascent practices - Shelters that collect data have limited guardrails and processes in place to ensure accurate data collection - And a good amount of shelters do not collect data itself (at least not digitally) - Data standardisation - Every shelter has their own "right way" - Adding more attributes to collecting animal level data does not naturally scale at the moment across shelters - There are about 30 different shelter software companies in the US and all have their own ways of data collection => Makes it harder for the SAC software to standardise this - A prospective solution that the team is actively working towards is creating individual APIs to connect to shelter software companies - Tracking an animal - Animal level data is considered to be of most important value - Every record is expected to have a unique `animal_id` . For SAC, this id field should help identify an animal throughout its lifetime as it passes through different shelters (if this happens). If not, then the animal will be double counted in the national statistics - It's hard to ascertain this with animals ### (Very) High-level overview of SAC's data architecture  *Note: The above is only a high level representation. It is not to be considered as the complete data architecture of SAC* ### Footnotes [^1]: As it is in any industry that has got to do with data! [^2]: I absolutely love the dashboards ;)