The Challenge: Reaching Past Your Own Audience
Intuit QuickBooks already had what most brands are still building: an in-house SEO team, active social channels, and a paid native advertising platform. By any traditional measure, their content was getting found. But AI search changed the scoreboard.
When a small business owner asks ChatGPT or Perplexity how to handle overdue invoices, the answer draws on sources the model trusts. Owned channels and paid placements only go so far in earning that trust. So QuickBooks added a new lever: earned media distribution through Stacker's network of publishers.
Since April 2026, the team has distributed 17 stories, earning 3,920+ publisher pickups, roughly 272,700 estimated views, and a 223% average citation lift in AI-generated answers. Erica Chappell, Content Amplification Lead at Intuit QuickBooks, said: "With how rapidly GEO is evolving, we wanted multiple levers to pull."
The Approach: Three Strategic Choices
QuickBooks is not the only Intuit brand to work with Stacker. TurboTax and Credit Karma came first, giving the team a proven model to follow. Three choices shaped the program:
- Lead with proprietary research. QuickBooks built stories on its own data about small business life, from cash flow pressure to community support. Original data gives editors a reason to publish and AI models a reason to cite.
- Scale past paid. Stacker ran as a top-of-funnel channel alongside QuickBooks' paid native placements. Earned pickups extended reach beyond their paid efforts.
- Plug into existing measurement. Stacker's API connected directly to Intuit's GEO performance dashboard. Pickup and citation data flowed in automatically, with no manual reporting.
That last point matters more than it sounds. With something as novel as GEO, communicating impact upwards and into systems or dashboards that already exist is immensely important.
The Results: A Majority Voice in AI Answers
| Metric | Result |
|---|---|
| Stories distributed | 17 |
| Publisher pickups | 3,920+ |
| Estimated views | 272,700 |
| Cost per pickup | $8 |
| Share of voice vs. tracked competitive set | 56% |
| Average citation lift | 223% |
| Average response presence rate | 31% |
Share of voice is measured against QuickBooks' tracked competitive set. It shows how often QuickBooks appears in AI answers relative to those competitors. Nine of the 17 stories have matured enough for full GEO tracking, and their numbers show what earned distribution does for AI visibility.
Two Standout Stories
Two stories illustrate different paths to AI citation success:
"7 ways small businesses can support the LGBTQ+ community year-round" achieved a 79% share of voice against its tracked competitive set and a 726% citation lift. An evergreen, values-driven topic gave AI models a clear, credible source to return to.
"Report finds more small businesses are carrying overdue invoices than last year" reached a 62% share of voice. This story leveraged proprietary QuickBooks data about invoice aging, a topic directly relevant to AI search queries about small business cash flow.
The contrast is instructive. Values-driven content and data-driven content both performed well, but for different reasons. The first answered a question AI models were already being asked. The second provided data no other source could offer.
What Other Brands Can Learn
The QuickBooks program offers several transferable lessons:
- Original data is non-negotiable. Every successful story in the program was built on proprietary QuickBooks research. AI models cite sources that offer something unique. Generic content, no matter how well-distributed, does not earn the same authority.
- Earned distribution multiplies citations. A story on the QuickBooks blog is one signal. The same story on 230+ publisher sites is a pattern AI models learn to trust. The math is simple: more credible placements mean more citation opportunities.
- Measurement must be built in. QuickBooks connected Stacker's API to its existing GEO dashboard, making it possible to track citation lift and share of voice without manual reporting. If you cannot measure AI visibility, you cannot optimize for it.
- Cost efficiency is achievable. At $8 per pickup, the program delivered earned media at a fraction of the cost of paid placements, with the added benefit of building AI search authority that compounds over time.
- GEO is not SEO with a new name. The strategies that work for Google Search rankings do not automatically work for AI answer citations. Earned distribution across trusted domains is a distinct mechanism that AI models reward differently than link-based authority.
The Broader Context
The QuickBooks case study aligns with findings from Stacker's 2026 AI Trust Report, published October 5, 2026, which analyzed 3.9 million citations across six AI platforms. That report found that brand sites were cited in 29% of AI responses, matching or exceeding editorial sources. The QuickBooks program is a working example of how to earn those citations.
It also stands in contrast to the AthenaHQ State of AI Search report from September 29, 2026, which found that the average brand is invisible in 84% of target AI search responses. QuickBooks is in the minority that has done the work to be visible.
Read Next
- The 2026 AI Trust Report: Which Brands AI Actually Trusts — The full landscape analysis behind these citation patterns
- Google's Gemini Coworker Agent Gets Its Own Email — How persistent AI identity changes enterprise work
- Agentic Payments Fraud: Blind Spots in Fraud Controls — When AI agents make payments, fraud detection needs to evolve
- GEO Becomes a Venture-Backed Market: Inside OptimizeGEO's $80M Round — What 100+ brands reveal about GEO market maturity
