MyPlanAdvocate: The Medicare Team That Received a 262x ROI With Bland and Cut Unqualified Calls by 500%
MyPlanAdvocate (MPA) is a Medicare brokerage that runs entirely on the phone. With more than a hundred licensed agents and thousands of inbound calls every day, the way those calls are routed, qualified, and processed has a direct impact on revenue. As their VP of Growth Jake Peters, the man behind all the extreme growth the company has seen, explains, “Everything in our business is telephony based. It is all a telephonic sales structure.”
Last year, the AI vendor who was powering their telephony process abruptly exited the market. The reason, as Jake later learned from engineers inside the company, was Bland. After seeing Bland’s approach and realizing they couldn’t compete with a flexible and scalable platform like Bland, the vendor chose to shut down their business entirely. That left MPA without the system that had become absolutely essential to their operations. They needed a replacement immediately. Something they could control, iterate on, and trust with the most sensitive part of their business.
The Challenge
MPA buys large volumes of inbound Medicare calls in real time. In their industry, they only pay for those calls if they last longer than 90 seconds. After that point, every caller who turns out to be unqualified becomes pure cost. Before Bland, 25 to 30 percent of calls they paid for weren’t even eligible for a sale.
Their frontline agents could qualify callers, but humans vary in approach. Some push too hard to get a transfer. Others take too long. Everyone brings a slightly different workflow. As Jake put it, “The front end agent is incentivized on the number of transfers they produce… whereas the Bland agent doesn’t care.”
On top of that, after every successful sale, agents were required to read 15 minutes of mandatory disclosures. It was time-consuming, repetitive, and exhausting, and with 4 to 5 sales per agent per day, those ten-minute blocks added up fast.
When their previous AI qualification provider shut down, Jake said this: “It was basically impossible for us not to have this thing in our business at this point.”
The Solution
MPA explored two options: building on ElevenLabs and building on Bland. Given the urgency of their need, they tried both simultaneously. They built an agent on ElevenLabs, but it struggled to stay on script. The agent would handle calls in unpredictable orders, which made it difficult to reliably disqualify callers in under 90 seconds.
Bland was different. Jake built the first working version of his Bland agent, Emily, in just one day. Coming from a marketing automation background, the visual builder felt intuitive. “Everything felt familiar… connecting the steps and branching logic made sense,” he said.
Once the initial version was live, Bland’s trusted implementation partner, Bland Labs, stepped in. Jake describes it this way: “The agent was operating fine, but it was like 80 percent of the way there. The Bland Labs’ team took it from 80 to like a 99.” Together they refined timing, pronunciation, edge-case handling, and the exact sequence needed to get callers qualified or disqualified before the 90-second mark.
Emily now answers roughly 2,500 inbound calls a day. She confirms their medicare eligibility, checks service areas, confirms intent, and gracefully ends calls that aren’t a fit. “Most people think Emily is a real person,” Jake told us.
On the backend, MPA introduced Mason, an AI agent who reads all legally required disclosures once a package is purchased. Instead of agents spending fifteen minutes repeating the same scripts, Mason handles the entire flow. Agents are now able to go straight to their next sale.
One caller even expressed disbelief at how quick the handoff was. After Mason finished and transferred back, the agent simply confirmed everything in 24 seconds. The caller said, “Oh, that’s it?” And yes – it was.
The Outcome
The impact was immediate and significant. Emily achieved a conversion rate that was 200% higher than their human agents.
Just as importantly, the cost savings were dramatic. Before Bland, MPA paid for about 25 to 30 percent of unqualified calls. Now it’s under 5 percent. At 5,000 calls a day, often $20 to $30 each, that difference alone adds up to over $1.2M in savings per year.
Then there is the time impact. Mason’s automation gives each agent back roughly 40 to 50 minutes per day. That’s a full extra sale per agent, every day, leading to an additional $40M in revenue each year. “You can clearly see the impact we get.”, said Jake
Their human agents are happier too. Nobody misses qualifying low-quality calls or reading long disclosures. “We actually use Emily and Mason as recruiting tools now too,” Jake said. New agents join because they know they’ll spend more time selling and making commissions. Bland is helping them gain and retain top talent at scale.
Surprisingly to Jake, MPA’s older customer base has really embraced the AI experience. “I was relatively surprised how well our demographic has adjusted,” Jake shared. People routinely thank Mason, and most don’t even realize Emily isn’t human.
