ROI Case File No.476 'The Battle to Reach the Right Person'
The Battle to Reach the Right Person
Chapter 1: Two Hundred Calls, Two Appointments
"I only have a few hours per week available for outbound calls. Two hundred calls, two appointments."
John Smith, Sales Director at Globex Corporation, opened his planner as he spoke. Last week's schedule was full—recurring meetings with existing customers, internal reporting sessions, handoff coordination. Time for cold calls existed only as leftover space.
"When did you start doing outbound?" I asked.
"January this year," John answered. "Inbound leads have been declining, so I decided to go out myself. I'm using ABM to prioritize targets—focused calls only. Even so, I often get stopped at reception."
"What does 'stopped at reception' look like?" Claude asked.
"The receptionist picks up first," John answered. "They check the department, transfer me. I introduce myself again at the transfer. If the right person is out, I leave a callback request. Callbacks often don't come. One call averages seven minutes—and sixty percent of the time, I never reach the actual person."
"Of those seven minutes, how many involve actually talking to the right person?" Gemini asked.
"If I connect—about three to four minutes," John answered. "So three to four minutes are consumed by reception handling and waiting. For the sixty percent that never connect, all seven minutes are gone."
"You mentioned trying a sales outsourcing firm," I continued.
"I did," John answered. "But tracking progress reports, verifying content, sending feedback—the management overhead was more than expected. I concluded it was faster to do it myself, and here we are."
"When did you start thinking about AI calling?" Claude asked.
"Three months ago," John answered. "But I don't know how to structure it. There's also resistance to handing off all my calls to AI. I want to design where AI ends and I begin."
"That design is today's topic," I said.
Chapter 2: The Five Experiences of HEART
"This case calls for HEART."
Claude wrote five letters on the whiteboard: H, E, A, R, T.
"HEART stands for Happiness, Engagement, Adoption, Retention, and Task Success—a framework for designing user experience across five dimensions," I explained. "It was originally developed for product design, but the same logic applies to sales workflow design. Here, the 'users' are both John and the customers receiving the calls. If we don't design the experience for both sides, AI calling will only function for one of them."
"Let's start by measuring the current cost," Gemini said, opening ROI Polygraph. John's call logs and schedule data were entered.
"Monthly sales work costs are in," Gemini read aloud. "12 hours of calling per week × 4 weeks = 48 hours/month—wait, let me correct that: 3 hours/week × 4 weeks = 12 hours/month. At ¥5,000/hr for sales staff: ¥60,000/month. The 60% non-connect rate on 7-minute calls—roughly 50 files/month—consumes approximately 6 hours. That's ¥30,000/month achieving zero return."
"The number may look small," Claude continued. "But within John's available 12 calling hours per month, ¥30,000 worth—50%—is disappearing before reaching anyone."
John said quietly, "Half my calling time was being consumed before anyone heard me."
"Now let's design with HEART," I continued.
[Happiness — Whose Stress Do We Reduce?]
"John's stress is reception handling and waiting," Claude said. "The customer's stress is being hit with an unexpected call. AI calling reduces John's reception-bypass cost—but it can also increase customer stress. From the first design step, we think about how to raise Happiness for both sides simultaneously."
"Two methods for getting AI through reception," Gemini continued. "First: make contact by email or DM before calling—both reception and the target person receive the call with some context. Second: have AI identify the target department during the auto-attendant stage and connect automatically. Combine these two, and John only moves after the right person picks up."
[Engagement — What Happens When the Right Person Answers?]
"From the moment the right person picks up, John takes the call directly," I designed. "AI handles reception bypass; the instant the target connects, the call transfers to John. That's the heart of Engagement design. If AI starts talking to the prospect, the customer senses something wrong. Human conversations should be heard by humans."
"What if the transfer is delayed?" John asked.
"There's a fallback," Claude answered. "AI leaves a callback number and ends the call. A callback list is generated for times when John can engage. The design ensures John never misses an actionable moment."
[Adoption, Retention, Task Success — Designing to Keep It Running]
"The remaining three together," Gemini continued. "Adoption: for the first week, John reviews all AI call results personally—which receptionist blocks went through, which context led to connection. This feeds script improvement. Retention: review appointment conversion rate weekly; if targets are missed, revise the script. Task Success: set a monthly appointment target of 6—up from the current 2—and verify achievement in three months."
"Let's run the investment plan through ROI Proposal Generator," Gemini proposed.
AI calling implementation costs and appointment-gain effects were laid out.
- Initial cost: AI calling system + script design — ¥500,000
- Monthly cost: System usage — ¥40,000/month
- Monthly effect: Reception-handling cost reduction = ¥30,000; increased appointments from 2 to 6/month (at ¥5M deal value, 10% close rate = expected ¥2M revenue increase)
- Payback period: Cost-reduction basis only: ¥500,000 ÷ ¥30,000 = approx. 17 months; including revenue effect: within month one
"Including revenue effect, payback is in month one," Gemini summarized. "But more appointments only matter if they close. AI calling's goal isn't to generate appointments—it's to increase the time John spends talking to the right people."
John said quietly, "Framed that way, my resistance to AI handling calls fades. I'm giving it the time spent fighting reception, and getting back the time spent talking to real people."
"Exactly," Claude answered.
Chapter 3: Designing Before Dialing
"Let me lay out the plan," I said, standing at the whiteboard.
"Week one—script design. Articulate the full flow from reception bypass to target connection. Extract three successful reception-bypass patterns from John's history and use them as the AI's base script. Week two—small-scale test. Run twenty calls from the target list. Review results and refine the script. Weeks three and four—full deployment. Target 100 calls per week; track connection rate and appointment conversion weekly."
"Extracting successful patterns myself sounds difficult," John said.
"Let me ask you one thing," I said. "When you've gotten through reception in the past—what did you say?"
John thought for a moment. "I find it works better when I say I'd like to speak with whoever handles your DX initiatives, rather than naming a specific department."
"That's pattern one," Claude said. "Can you think of two more?"
John looked through his planner and named two more phrases. Claude wrote them down.
"The script has a skeleton," Claude said. "You can begin test calling."
Chapter 4: The Day the Prospect's Voice Was the First Thing He Heard
Four months later, a report arrived from John.
One month after AI calling launched, the target-connection rate rose from 40% to 68%. Monthly appointments went from 2 to 7. John's direct calling time fell from 12 hours to 4 hours per month; the freed time went into proposal preparation and existing customer follow-up.
"With no more time spent fighting reception, I can use that time to prepare for the conversation," John wrote in his report.
In month three, there was a week where monthly appointments exceeded 8. John reviewed that week's script and identified "leading with a theme" as the highest-performing pattern. The script was updated to front-load that phrase.
The day the prospect's voice was the first thing he heard.
"Of the seven minutes per call, three to four minutes were disappearing into reception and waiting. That time isn't a conversation with a customer. AI handles those three minutes. Humans handle the conversation that starts when the right person answers. The five experiences of HEART—Happiness, Engagement, Adoption, Retention, Task Success—are a map for designing how AI and humans divide the work. John gave away the time spent fighting reception and took back the time spent talking. The two appointments from two hundred calls grew to eight."
Related Files
Tools Used
- ROI Polygraph — Visualizing call hours and reception-bypass costs
- ROI Proposal Generator — Simulating appointment-gain effects from AI calling
