Product descriptions:
Turning verified rental data into actionable fraud and ROI insights
Snappt is an applicant trust platform for the rental housing industry, helping property management companies verify identity, income and rental payment history so they can reduce fraud and bad debt. Snappt’s property management customers have been asking for richer analytics for years: fraud trends, aggregated insights and hard proof of Snappt’s ROI across their portfolios. Yet competing roadmap priorities and limited engineering cycles, coupled with Tableau’s per-user licensing costs, meant Snappt couldn’t justify building a full self-service analytics solution. By using Databricks AI/BI to embed dashboards directly into the application, the Snappt team launched production-ready reporting in about a week, giving hundreds of property admins real-time visibility into fraud, aggregated analytics, product usage and bad-debt risk while keeping engineering focused on core features.
Manual reporting and the cost of scale
Snappt serves property management companies that oversee large multifamily portfolios, from 100-unit buildings to thousands of units. During the rental application process, Snappt verifies identity, income and rental payment history, combining machine learning (ML) models and human-in-the-loop review into a single applicant trust report that leasing teams can use to make decisions with confidence.
As adoption grew, customers began requesting more in-depth analytics on top of their verified data. Owners and operators wanted to understand questions like:
- What is my renter applicant income and identity fraud rate by property, region or brand?
- How much bad debt is Snappt helping me avoid across my portfolio?
- What aggregate insights do we know about our applicants?
They also needed better evidence of value for the building owners who ultimately pay for Snappt. Teams were constantly being asked to prove the ROI of Snappt by showing how much fraud they had caught and how many potential evictions and bad-debt dollars had been avoided.
Early attempts to meet this demand relied on internal Tableau dashboards and manual reporting. Customer success managers would screenshot charts, assemble QBR decks and manually slice data for each customer request. This was time-consuming for Snappt’s analytics team and frustrating for Snappt’s customers, who wanted the freedom to “slice and dice” the data themselves.
The need for customer-facing analytics was clear, yet the effort required to build it kept it below the cut line on the roadmap.
“Every quarter, embedded dashboards have come up in our roadmap conversations,” Briana Ings, Chief Product and Technology Officer at Snappt, said. “It always looked like a big rock. We assumed we would need a dedicated squad for months, so it kept losing out to core features like new data connections.”
Snappt quickly ruled out embedding Tableau. With over 22,000 potential leasing agents and hundreds of users, their existing per-user licensing model resulted in unpredictable, high costs for users who might log in only once a month. The alternative — building a custom analytics solution on AWS — would require significant application engineering effort and increase infrastructure burden on their core product databases.
Snappt needed a way to provide property owners and administrators with rich, trustworthy analytics directly within the product, without requiring a large engineering initiative or overhauling their existing stack.
Delivering embedded analytics with Databricks AI/BI in a week
Snappt already powers their fraud detection models and core analytics, using the Databricks Data Intelligence Platform as their central data lakehouse and ML environment. Learning about Databricks’ ability to embed AI/BI Dashboards for external users was the breakthrough.
For Snappt, this unlocked a different way to think about in-product reporting.
Embedded AI/BI Dashboards for external users enabled Snappt to authorize thousands of end users through a single integration, while still enforcing strict row-level security, so that each property or admin only saw data they were allowed to see. Because the dashboards run on top of Databricks SQL and utilize their data in Unity Catalog, they would not need to move data or place an extra load on production databases. Unity Catalog and built-in Git integration provided governance and change control that directly aligned with Snappt’s SOC-compliant development processes.
Equally important, the team could treat reporting as a data product rather than an engineering-heavy feature. Anyone who understood the data model and SQL could build and iterate on dashboards without tying up application engineers.
To test the idea, Briana built a proof of concept herself.
“I spun up a dashboard in Databricks and coded a simple standalone app one night after my kids went to bed,” she said. “It took about four hours to get from idea to a working prototype.”
Within about five days and only a few hours of engineering work, the embedded dashboard was live in production as a new tab inside the Snappt portal.
Their initial launch focused on customers who were asking for analytics the most: admins and regional leaders who oversee portfolios of properties. Out of roughly 850 admin users, Snappt has already enabled close to 200 for the new dashboards and is rolling it out in a controlled, high-touch way through customer success and sales conversations. Over time, they plan to extend access to their broader base of 22,000 leasing agents.
The embedded experience currently includes several focused views:
- Fraud Insights to track income and identity fraud rates, bad debt prevented by Snappt and how fraud trends change over time as fraudsters shift away from buildings using Snappt
- Aggregated Insights, such as top employers, median income, median age, segments like students and young professionals, and where applicants are moving from
- Geographic Insights with maps of fraud rates and income by county and state, as well as employer and applicant trends by geography
- Product Usage to show which features are activated across a customer’s portfolio, so leaders can identify underutilized capabilities that could reduce risk
- Bad-Debt Calculator that combines verified fraud and usage metrics with customer-specific assumptions on eviction timelines and costs to estimate bad debt avoided
Behind the scenes, their AI/BI Dashboards are fed by data pipelines that aggregate data from multiple operational systems into a unified lakehouse, alongside outputs from Snappt’s Databricks-based machine learning models.
Driving retention, customer stickiness and product strategy
The impact of embedded analytics has been felt across customers and internal teams.
For customers, the dashboards turn verified data into clear, visual insights that are easy to explore without waiting on a CSM. Property admins can compare fraud rates by building and region, understand aggregated applicant demographics and quantify the financial impact of Snappt in terms that matter to owners.
By correlating product usage with retention outcomes, the Snappt team found that customers who activate newer features experience a roughly 30% improvement in retention compared with those who do not, even after accounting for other factors. While the team is careful about causality, that signal is strong enough to shift focus. Customer success now prioritizes feature activation as an early indicator of long-term health.
The dashboards also help Snappt’s go-to-market teams articulate differentiated value in a competitive market. Initially, Snappt lagged some competitors on surface-level reporting features. With the new embedded dashboards, they now offer deeper fraud and demographic insights than others can match, as their analytics are grounded in verified income, identity and payment data rather than self-reported fields from application forms.
Most importantly, all of this has been achieved without standing up a dedicated engineering squad for reporting. New metrics and data cuts are built and tested directly by Snappt’s data team in Databricks, then surfaced in the embedded dashboard without requiring changes to application code. That keeps core engineers focused on evolving the applicant trust platform while still delivering a high-quality analytics experience to customers.
“If your core product is not reporting, you do not need to build a whole reporting stack from scratch,” Briana explained. “Because we already use Databricks as our data lake and for machine learning, using AI/BI to power embedded dashboards was the most efficient and scalable path. It let us deliver something our customers had been asking for, with a fraction of the effort we originally expected.”
Looking ahead, Snappt plans to roll the dashboards out to more users, continue replacing internal Tableau reports with AI/BI Dashboards and deepen their use of the Databricks Data Intelligence Platform for both aggregated analytics and machine learning. The result is a more data-fluent organization, stickier customer relationships and a roadmap where in-product analytics no longer requires a massive trade-off against core innovation.