The Problem: How to build digital advertising audiences based on contextual intelligence
It’s one thing to know where in the world (geographically-speaking) a mobile ping comes from. It’s another thing to be able to uncover the context or, rather, the real world environment surrounding that ping. Is it coming from a restaurant? A gym? A corporate office park? And then taking it one step further, where are those anonymized mobile pings traveling to and from?
Unfortunately, most location-based data sources available today—including open source data—simply don’t cut it. While they can provide the exact latitude and longitude, time stamps, device types, and device IDs associated with almost any mobile ping, they can’t tell the full story about where those pings really came from. That doesn’t really help advertisers at all.
So, the RainBarrel team tasked itself with developing a cutting-edge way to build digital advertising audiences at scale. They already had their own source for mobile pings but lacked the necessary context around those mobile pings to unlock new value for their customers. They just needed access to high-quality and accurate geospatial data to fill in those contextual gaps.
Making matters worse, as a Canada-based startup squarely focused on addressing the targeting needs of advertisers servicing Canadian consumers, the pickings were slim. “When it comes to data, Canada is an underserved market,” explained David Choi, RainBarrel’s Product Manager. “Most data providers are based in the U.S. and tend to treat Canada like an afterthought.”
While getting access to the right data was one hurdle to overcome, it was equally important for that data to be sourced in an ethical way. “For us, it’s an absolute priority to not only be privacy-compliant when building digital advertising audiences but also to be clear about how we put those audiences together,” underscored Travis Riedlhuber, RainBarrel’s Managing Director. “Transparency is the key to building high-quality digital advertising audiences at scale.”
The Problem-Solver: RainBarrel
RainBarrel is a proprietary Audience Graph based on commercially available geospatial data, allowing advertisers to target their messages to the right audience. As opposed to most other audience targeting methods available in the market today—typically defined by online user behaviors—RainBarrel takes precision targeting to an entirely new level by leveraging the power of location-based data to build digital advertising audiences rooted in offline user behaviors.
Say you run a luxury brand and wanted to target consumers interested in purchasing high-end goods. In the past, you might have built a digital advertising audience around users who regularly visited the websites of brands like Chanel, Louis Vuitton, and Gucci. Those online behaviors would lead advertisers to believe that those consumers make up the “luxury buyer” segment.
Unfortunately, these online behaviors are often aspirational at best; just because consumers visit a luxury brand’s website doesn’t necessarily mean they can or will make a high price point purchase. However, a consumer who regularly steps foot at least twice per month into a Nordstrom store is more likely to be an active shopper—based on offline behaviors alone—and, therefore, more valuable to a brand from a targeting standpoint. This is precisely how RainBarrel expertly uses geospatial data to close the gap between aspiration and intent.
In other words, they’ve unlocked the untapped potential at the convergence of digital audience segmentation and real-world actions to create a more effective way for advertisers to target, reach, engage, and convert audiences. And they’ve built this around three guiding principles:
- Privacy: All data used to build digital advertising audiences must meet the most stringent privacy standards and be in compliance with current GDPR and CCPA regulations. No personally-identifiable information is ever captured or shared whatsoever.
- Transparency: All information pertaining to the company’s 2,600+ audiences is made publicly available, including details around data recency, audience size, number of POIs, geographic distribution, methodology, documentation, and the list goes on.
- Reach: All audiences are built, first and foremost, to address the evolving needs of RainBarrel’s customers in real-time. This also means making the audiences easy to activate and compatible with today’s leading DSPs and digital advertising platforms.
When vetting our options, we had two choices. Either we built this capability in-house from scratch or we worked with a partner who could tick all of the boxes.
The Solution: SafeGraph Places and Geometry
“When vetting our options, we had two choices,” stated Riedlhuber. “Either we built this capability in-house from scratch or we worked with a partner who could tick all of the boxes.”
Doing this in-house would be a significant undertaking that would ultimately require hiring up to two resources dedicated to building and maintaining a constantly updated database of contextual location information. But the fruits of that labor would likely be piecemeal at best—tackling one customer request at a time—because the internal resourcing model could scale quickly.
So, the RainBarrel team looked into various open source and private data provider options. It became clear early on that open source data simply didn’t have the right quality they needed. Nor was there anyone they could reach out to for support; it felt like a proverbial “black box.” After running several tests, the team also found several inconsistencies with the data and had no line of sight into details like operating hours or whether a POI is now permanently closed.
And for other (non-open source) data providers, either there wasn’t the breadth and depth for Canada they needed or the right level of transparency. That is, until they contacted SafeGraph.
“When we compared SafeGraph to the polygons we pulled from other data sources, we found that SafeGraph had the largest count of POIs—at least 50% more than other providers,” stated Choi. “This was an eye-opening moment for us because it made us realize just how underserved the Canadian market really is.” As an added benefit, the SafeGraph Places dataset, updated monthly, is enriched with contextual tagging—at the brand, category, and industry (NAICS) levels—while the SafeGraph Geometry dataset hangs its hat on spatial hierarchy (i.e. being able to identify when there’s a coffee shop or a bank located inside of a big box retailer, for example).
“What really stood out to us was how SafeGraph’s focus on transparency aligns with our mission,” reiterated Riedlhuber. “Thanks to the detailed documentation they provide, we know exactly how their datasets are built from the bottom up and, as a result, can use this information effectively to explain to our own customers how we’re building our audiences.”
When we compared SafeGraph to the polygons we pulled from other data sources, we found that SafeGraph had the largest count of POIs—at least 50% more than other providers.
The Result: Removing the limits for creating digital advertising audiences at scale
RainBarrel just hit a major milestone: launching the company’s external-facing web application for customers. “When we were first building our launch plans, we had hoped to go-to-market with anywhere from 250 to 500 public audiences,” said Riedlhuber. “But today, we provide over 2,600 public audiences—something we couldn’t have done without using SafeGraph data.”
In addition to creating a virtually infinite number of public audiences at scale, they now have a turn-key way to leverage the granularity of SafeGraph data—POI details, category tags, NAICS codes—and combine it with data from other sources—mobile device ping data, Canadian census data, and population density data—to build custom audiences addressing specific customer needs. This has given them a leg up against the competition; they can now offer their customers exclusive access to audiences that aren’t currently available from any other provider.
“We can now build audiences in a matter of 30 minutes, something that would have likely taken weeks to do without access to the right data before,” reiterated Choi. “Moving at the speed of our clients—and not just responding to the whim of demand—is critical for scaling our business quickly and staying one step ahead of rapidly evolving audience targeting trends.”
When we were first building our launch plans, we had hoped to go-to-market with anywhere from 250 to 500 public audiences. But today, we provide over 2,600 public audiences—something we couldn’t have done without using SafeGraph data.
The Future: Adding more contextual details and expanding beyond Canada
Aside from the big launch of their web application, the RainBarrel team is eager to expand their reach—first into the U.S. and then moving into other markets around the world.
Additionally, they are constantly on the lookout for other privacy-compliant contextual data sources that can help them add even more detail to the audiences they create. “We want to know what else we can learn about the behaviors actually happening at these locations,” concluded Riedlhuber. “That way, we can get even more granular with our targeting to help our customers find untapped audiences that are more likely to engage and convert.”
And finally, the team is looking into the potential of using new mobile identifiers—versus cookies and MAIDS—as a way of expanding the size and scope of their audiences.
Our shared commitment around transparency and privacy is what put SafeGraph above and beyond in our books.
