Your Data Doesn't Always Fit a Checkbox, and That's OK
By: Elle Hamson
Organizations come to CTA for help with voter files, syncs, and our data warehouse, PAD. For many teams, that's exactly what they need. But some of the hardest problems show up in the gap between "we have the data" and "our staff can actually use it day to day."
Those problems are usually specific: a particular survey structure, a long-running operational process, or access rules that shift from one engagement to the next.
Our Product and Engineering teams recently worked through two examples of this with our partners. On the surface, the projects looked very different. Underneath, the ask was the same: work with the data where it already lives, and make it usable in a way people can repeat.
From raw data to publication-ready crosstab reports
One partner was sitting on survey data in BigQuery and needed to turn it into weighted, formatted crosstab reports - the kind researchers and campaigners use to break down answer choices across questions, with the right handling of question types, response labels, and weights.
Instead of pulling data out and rebuilding everything in a spreadsheet, we built a step-by-step flow inside PAD. Users select the dataset and tables, map how questions and responses are structured, choose what goes on the rows and columns, and then fine-tune how answers appear - reordering options, grouping responses, hiding values that don't belong in the final output.
Once submitted, the system queues the job, tracks it from pending through completion, and serves up a downloadable report from cloud storage with a time-limited link so access stays controlled.
Report generation runs asynchronously, so big jobs don't freeze up the interface, and the queue processes on a schedule so multiple users aren't fighting over capacity. We also made it easy to reuse a setup: run the same configuration against a new table, or adjust it without rebuilding from scratch.
The result is straightforward: the data stays in PAD, and analysts get polished output without needing to wrangle a pipeline.
Operations, relationships, and access all in one place
The second partner wasn't looking for more reporting. They needed a system to manage their day-to-day work: jobs tied to clients, outside vendors, internal staff, and locations, plus session-level records tracked over time. They'd outgrown spreadsheets and needed something with structure, appropriate access, and tooling their own admins could actually manage.
We built a project-scoped application inside PAD that pulls those pieces together: it includes filterable lists that hold up at scale, forms that capture the details that actually matter to their workflow, and navigation that connects related records without making users dig.
On the access side, we built admin controls that allow designated staff to define data-access rules and enforce firewalls - scoping who can see which jobs or clients, setting expirations when engagements end, managing users, and keeping dropdown fields consistent as their language evolves.
It's not a generic CRM; it's their workflow, built on the same infrastructure as the rest of PAD.
What we take away
These were different problems - survey analytics in one case, operational recordkeeping and access control in the other - but our approach was the same:
Start with how your data and workflows actually operate: table shapes, who touches what, and who should/shouldn’t have access
Extend what you already run on PAD: the same login, hosting, and patterns your team already knows
Deliver something repeatable: not a one-off solution, but something people can use again tomorrow
If you’re dealing with a problem that doesn't map to an existing feature, that's not a dead end. It's usually where a good conversation starts.
We begin with discovery, get clear on what you need and what constraints matter, and scope to see what we can build together.