The objection isn't the problem. The objection is the close. A homeowner who says "I need to think about it" at minute 47 isn't ending the conversation — they're handing the Comfort Advisor the exact lever to close. The advisor who pulls up a 10-second AI-drafted rebuttal on the tablet, in shop voice, with the customer's specific deal context, closes 18-25 percentage points higher than the advisor reaching for memorized scripts. This lesson is the top-10 objection prompt library for the 2026 kitchen-table seat — "I need to think about it," "let me get other bids," "can you do it for cash," "my buddy can do it cheaper," "I'll wait until summer," "I need to talk to my spouse," "the other quote was $4K less," "is this in stock," "will my utility rebate cover this," "how long has this company been around." Ten objections. Ten prompts. Ten-second answers on a phone, in front of the customer, in shop voice. The advisor reads the rebuttal off the screen with eye contact intact. The customer never sees the screen. The close holds.

Why AI Rebuttals and Not Memorized Scripts

Memorized scripts fail at the kitchen table for the same reason generic AI prose fails — they sound rehearsed. Homeowners in 2026 have been pitched by enough contractors that the rehearsed script triggers the "salesman alert" within 12 seconds. Trust collapses; close rate drops 8-14 points compared to advisor-personalized rebuttals. The fix is not to abandon scripts — the fix is to abandon generic ones. The AI-drafted rebuttal lives in the same place memorized scripts used to: on the tablet, on the index card. The difference is that the AI rebuttal is generated against the homeowner's specific deal context — name, ticket size, financing tier, rebate stack, system age, discovery-conversation references, spouse-decision dynamics — and produces a 10-second answer that sounds like the advisor's own voice.

The AI rebuttal is also faster than memorized recall on novel objections. The advisor who has memorized 6 rebuttals can handle 6 objections fluently. Objection #7 — "my cousin says you're about to be sold to private equity" — collapses the memorized advisor. The AI-equipped advisor types five context fields, gets a 4-sentence rebuttal in 8 seconds, and reads it off the tablet with calm composure.

The mechanic is the five-part prompt structure from L2 Chapter 1 — Role, Context, Task, Format, Constraint — with the template pre-built per objection and the context filled at the kitchen table. The advisor types the homeowner's specific deal context in 8 seconds and gets a rebuttal in 4 more seconds. Total: 12-second turnaround. The homeowner sees the advisor pause, glance at the tablet, look back up, and deliver. The pause reads as thoughtfulness.

Tablet Mechanics Without Breaking Eye Contact

Three rules govern AI rebuttal use at the kitchen table or trust collapses faster than any rebuttal can save the close.

Rule one — the tablet is tilted toward the advisor, never the customer. The homeowner never sees the AI screen. Tablet angled 30-40 degrees from the customer's line of sight. When the advisor types or reads, the homeowner sees the back of the tablet. Protects two things: prevents the homeowner from reading the AI prompt and recognizing it as AI-generated, and preserves the eye-contact rhythm. Advisor: see the homeowner, glance at tablet, see the homeowner, read rebuttal, see the homeowner.

Rule two — the advisor never types during the conversation flow. Typing is for 8-second pauses the advisor takes to "consider" the objection. Advisor responds verbally first ("That's a fair concern — let me think about how best to address that"), then types the context update during the homeowner's natural pause, then reads aloud. Typing during the homeowner's speech is the dead giveaway of device-preoccupation. Rhythm: listen, acknowledge, pause, deliver.

Rule three — the rebuttal is read aloud verbatim, never silently paraphrased. AI language is shop-voice-tuned through the prompt constraint; paraphrasing introduces advisor anxiety vocabulary that reads as uncertainty. The 4-6 minute personalization layer happens before the appointment, not during. At the table, the AI's draft is the script. Trust the constraint discipline — "no hard-sell," "no scarcity," "no 'partner' or 'family,'" "in [shop name]'s direct Southern voice" — and the output reflects it.

Objection One: "I Need to Think About It"

Most common kitchen-table objection. Closes 15-20% on memorized scripts; closes 45-55% with the AI permission-to-think rebuttal that re-anchors to payment and surfaces what specifically they're thinking about.

Template (locked). Role: Comfort Advisor at kitchen table for [ticket size] replacement. Homeowner just said "I need to think about it." Brand voice direct, no pressure. Context (advisor fills): [name, ticket size, financing tier from soft-pull, credit stack net cost, system age, spouse dynamics, comfort/budget priorities, last discovery concern]. Task: 30-second permission-to-think rebuttal that re-anchors sticker to monthly payment, surfaces what specifically they're thinking, offers same-evening callback if spouse input needed. Format: Single paragraph, 70-90 words, ending with one open question. Constraint: No pressure. No invented APR or credit amounts. Offer no discount. No "today only" or "limited time."

AI output (Mr. Henderson, $14K HP, Wisetack 8.99%/84mo, post-stack $146/month, 18-year-old system, spouse Linda absent, discovery concern was upstairs bedrooms cold): "Mr. Henderson, that makes complete sense — this is real money and the right call is to think it through. What I'd want to make sure is clear before you do: the actual out-of-pocket after the federal credit and the SoCal Edison rebate is $9,100, financed at $146 a month over seven years. The 18-year-old system you're running now is failing on the upstairs cooling problem we talked about, and the repairs you've put into it over the last three years averaged $400 each. What specifically would help you sleep on this — the monthly piece, the timing, or wanting Linda in the room before you decide?"

Open question surfaces the actual blocker. "Monthly piece" — advisor pivots to GreenSky longer-term option. "Linda in the room" — advisor offers same-evening callback at 8:30 p.m. "Timing" — advisor surfaces install scheduling and failure risk on the 18-year-old system. Objection becomes the close.

Objection Two: "Let Me Get Other Bids"

Closes 22-30% with the AI compare-the-math rebuttal that respects the homeowner's diligence, surfaces apples-to-apples criteria (SEER2, warranty, financing tier, rebate stack, manufacturer brand, installer certification), and offers to be the reference point. Pre-AI close rate on this objection: 35-40% but typically by reducing margin via discount tactics.

Template constraint critical: do not bash competitors. Bashing is the second-most-common close-killer in 2026. Homeowners read it as insecurity. Shop voice: "we are confident in our work — compare us on the math."

AI output: "Mr. Henderson, that's the right thing to do — a $14K decision deserves more than one perspective. What I'd ask you to compare on, when the other quotes come in: the SEER2 rating (this proposal is 17 SEER2; some quotes will be 14-15 to hit a lower price), the warranty term and what's covered (this is 12-year parts and labor; some shops are 5-year parts only), the manufacturer (this is Trane; check what they're quoting and whether the installer is a certified dealer), the financing tier (we have Wisetack at 8.99% over 84 months; some shops quote longer terms at higher APRs to make the monthly look smaller), and the rebate stack (this proposal captures the $2,000 25C federal credit and the $1,500 SoCal Edison utility rebate; not every quote includes these — that's $3,500 of out-of-pocket you'd lose). Could you walk me through what the other quote includes on those five points when you get it?"

Rebuttal commits the homeowner to a future call. 60-65% call back within 72 hours; of those, 70-80% close at the original or near-original proposal. The shop's apples-to-apples discipline anchors the comparison; competitors trying to win on price expose their lower SEER2, shorter warranty, or missing rebate.

Objection Three: "Can You Do It for Cash"

"Cash" in 2026 means two things — sometimes literal cash payment without financing (legitimate), sometimes off-the-books without sales tax or warranty documentation (illegal). The AI rebuttal handles the legitimate case warmly and the illegal case firmly.

AI output (legitimate cash): "Absolutely — we accept cash, check, or credit card on any deal. If you're paying cash, the proposal is the same price; we don't offer a cash discount because our pricing already reflects our actual cost of doing the work properly with full warranty. If you're asking whether we can structure the financing differently — say, half cash and half financed — that's something we can do; Wisetack will approve a smaller financing amount in 30 seconds. What's driving the cash question — wanting to reduce the financed amount, avoiding monthly payment, or something else?"

If homeowner clarifies off-the-books cash: "Mr. Henderson, I can't structure the work that way — every install we do gets permitted, gets warranty documentation filed with the manufacturer, and gets the sales tax remitted properly. That's not flexibility, it's our license. What I can do is walk through whether the financing payment math works for your situation or whether structuring it as a partial cash payment changes the picture. Is the warranty piece something you'd be willing to give up, or is the concern really about the financing terms?"

Firm-but-warm "I can't" preserves the relationship while protecting the license. About 30-40% of "cash" objections are legitimate financing-restructure conversations that close; 60-70% are off-the-books requests that walk. Better to know in 60 seconds than 60 minutes.

Objection Four: "My Buddy Can Do It Cheaper"

The buddy objection is the trust objection in disguise. The homeowner has a relationship with the buddy, has heard a price, and is comparing on a single dimension (price) against a relationship advantage. The advisor cannot win on relationship; the advisor wins on what the buddy can't offer.

AI output (buddy described as "works on cars and does some HVAC on the side"): "Mr. Henderson, your buddy probably can do the work cheaper — guys working on the side usually quote 30-40% less because they're not carrying the licensing, the insurance, the manufacturer dealer certifications, the EPA 608 universal refrigerant license, the warranty backing, or the permit pull-throughs that we are. If something goes wrong with the lineset two years from now, your buddy's repair will cost you out-of-pocket; if our install has a problem, our 12-year parts-and-labor warranty handles it. The other piece is the federal credit — to claim the $2,000 Section 25C credit, the install needs to be done by a licensed contractor with the manufacturer certification statement. Your buddy can do excellent work; he just can't deliver the credit, the warranty, or the permit. What matters most to you on this — upfront price, long-term coverage, or federal credit eligibility?"

Buddy stays respected. Math separates the paths. 50-60% close at the original proposal once warranty + credit math is surfaced; 30% sequence (buddy does adjacent non-licensed work, shop does the licensed HVAC install); 10-15% close with the buddy. Honest math, not buddy-bashing.

Objection Five: "I'll Wait Until Summer" (or "Spring" or "Next Year")

Trades off perceived cost-avoidance against actual failure-risk and rebate-window risk. Most common 2026 mistake: pressuring scarcity ("rebates are running out"); homeowner reads pressure as desperation; close collapses. AI constraint locks pressure out.

AI output (homeowner saying "I'll wait until spring"): "Mr. Henderson, the spring timing makes sense from a comfort perspective — you're not running the system as hard. Two things I'd want to put on the table for that decision. First, the 18-year-old system you have now: you mentioned repair history is running $400 a year and the upstairs cooling is failing. If the system gives out in February or March, install crew schedule is roughly 4-6 weeks out for a standard replacement, and you'd be paying out-of-pocket emergency repair plus the replacement. Second, the federal Section 25C credit and the SoCal Edison rebate — both reset on program-year cycles. The 25C resets January 1, so installing this year captures the 2026 credit ($2,000 heat pump + $600 panel); the SoCal Edison program-year ends June 30, after which rebate amounts and eligibility may change. Neither is 'last chance' — both are program-year considerations. The clean question: are you waiting because the monthly payment is tight, or because the timing feels rushed?"

Clean question separates the two waiting motivations. Monthly-payment waiting routes to GreenSky longer-term option. Timing-feels-rushed waiting routes to a scheduled follow-up call in 30 days. 35-45% close in the same conversation; 30% schedule a follow-up that closes within 60 days; rest genuinely wait, which is a legitimate outcome.

Objection Six: "I Need to Talk to My Spouse"

Highest-stakes kitchen-table moment in 2026. Spouse-decision dynamics are the single biggest predictor of close rate. AI rebuttal handles three scenarios — spouse-absent, spouse-present-disagreeing, spouse-veto-implied.

AI output (spouse Linda absent): "Of course — Linda should absolutely be part of this conversation. What I'd want to do is set up a 30-minute follow-up where I can walk through this proposal with both of you. I can do same-evening at 8:30 p.m. if she's available, or first thing tomorrow morning, or this Saturday — whatever works for your schedule. Before I leave: is there anything specific Linda would want to see in the proposal that I can prepare for the follow-up — system specs, financing details, the federal credit math, the install schedule, or warranty terms? And one quick question — is the spending piece the main thing for Linda, or is it more about the timing and the install disruption?"

AI output (spouse Linda present and disagreeing softly): "Linda, I'd love to hear your concerns directly — what are you thinking about that's different from what we've talked through so far? [pause] Mr. Henderson, that's helpful context. The piece I want to make sure is clear for both of you is the post-stack effective payment — after the federal credit and the utility rebate, we're talking $146 a month over seven years, against a current heating bill of $310 with a system that's costing roughly $400 a year in repairs. The decision doesn't have to happen tonight; what I can do is walk through whichever piece is creating the disconnect. Is it the monthly piece, the timing, the install scope, or something else?"

Acknowledge the dynamic; never try to override it. 60-70% of absent-spouse objections close on follow-up; 40-50% of present-disagreeing close after explicit-conversation pivot; 15-25% of veto-implied close (the rest walk, and the advisor respects the no). Forcing the spouse-decision is the failure mode; surfacing the dynamic honestly is the close.

Objection Seven: "The Other Quote Was $4K Less"

The specific-dollar comparison objection is apples-to-apples discipline applied live. The AI rebuttal asks for the other quote's specifics and walks the math.

AI output: "Mr. Henderson, $4K is real money — let me make sure we're comparing apples to apples. Five things I'd want to know about the other quote: the SEER2 rating on the equipment, the warranty term and what's covered, the manufacturer brand, the financing tier they offered, and whether the federal credit and utility rebate were included in the quoted price. Some shops quote $10K equipment that's 14 SEER2 with a 5-year parts-only warranty against our $14K 17 SEER2 with 12-year parts-and-labor; the apples-to-apples math closes the gap. Some quotes don't include the $2,000 federal credit and the $1,500 utility rebate — that's $3,500 of out-of-pocket you'd lose. Could you walk me through what the other quote includes on those five points, and we'll see whether the $4K is real or whether the comparison shifts when we compare line-by-line?"

40% of homeowners realize they can't answer SEER2/warranty/rebate and concede the comparison wasn't apples-to-apples; 30% produce specifics and the advisor walks the math live (typically closing 60-70% when the math favors the shop's proposal); 20% confirm a legitimate competitive advantage at the other shop (advisor either surfaces a warranty/rebate/cert advantage or honestly loses); 10% close on the spot because the question itself signaled trust.

Objection Eight: "Is This in Stock"

Became dramatically more frequent in 2024-2025 and persists into 2026. AI rebuttal handles legitimate timing concerns and price-anchored stalling. Advisor needs current inventory data from supplier integration; AI surfaces live data.

AI output (Trane XR17, Carrier Enterprise shows 5 units in stock): "Yes — the Trane XR17 we're proposing is currently in stock at our regional Carrier Enterprise distribution; we have allocated supply for installs within the next 2-3 weeks. The lineset, refrigerant (R-454B), and the panel-upgrade materials are all in our truck inventory or available next-day from RE Michel and Johnstone. The install crew schedule is the only constraint; we can do a Thursday or Friday next week, or roll into the week after if your schedule prefers. Is the stock question about timing or about whether we're proposing equipment we can actually deliver?"

Open question separates timing from credibility. If timing — advisor surfaces install scheduling immediately. If credibility — advisor offers to show the supplier portal screen briefly (tilting tablet to share visibility, then returning to advisor-only orientation). 75-85% of in-stock-questioned proposals close within the conversation when AI surfaces live supplier data; the legacy 6-month-wait fear collapses against verifiable real-time inventory.

Objection Nine: "Will My Utility Rebate Cover This"

Specific case of "is the deal as good as you say." AI rebuttal surfaces the morning rebate digest data live — current funded status, exact amount for the homeowner's tier, program-page URL. This is the moment where the AI's funded-status flag from Lesson 4 most directly protects the advisor.

AI output (SoCal Edison heat-pump rebate, verified funded GREEN this morning, $1,500 at homeowner's tier): "Yes — and we verified it this morning. The SoCal Edison heat-pump rebate is currently funded; the amount for the 17 SEER2 dual-fuel system we're proposing is $1,500 at your tier. The program is currently in budget; I've got the program page right here [advisor tilts tablet briefly to show URL] showing the rebate amount and the equipment qualification. The rebate processes about 6-8 weeks after install — Edison sends the check directly to you. Combined with the $2,000 federal 25C credit and the $600 panel upgrade credit, the total credit-and-rebate stack on this proposal is $5,100, which drops the net out-of-pocket from $14,200 to $9,100. Is your question about the rebate amount, the timing of when you receive it, or whether the program might run out before the install?"

Three-option open question surfaces the actual concern. Timing — advisor explains post-install rebate processing. Amount — advisor confirms the verified tier. Program-exhaustion concern — advisor explains morning digest discipline and URL-verify before delivery. Objection becomes a verification moment that strengthens trust.

Objection Ten: "How Long Has This Company Been Around"

Longevity = trust in literal form. The homeowner is asking whether the warranty will be honored, whether the company will be there in five years. AI rebuttal surfaces the shop's actual history, manufacturer certifications, BBB rating, Google review velocity, and (where relevant) the PE-backing context for a Wrench Group / Authority Brands / Apex Service Partners portfolio shop.

AI output (12-year independent shop with 1,400+ reviews): "We were founded in 2014 — 12 years in business at this location. We have 1,400+ Google reviews averaging 4.8 stars, an A+ BBB rating, and we're a Trane Comfort Specialist dealer (top 5% of Trane dealers nationally for installation quality and customer satisfaction). The warranty on this proposal is 12 years parts and labor, which means our team handles the labor for the full warranty period — if for any reason we weren't here, Trane's parts warranty still covers; we're not the gating party on the equipment warranty. We've installed roughly 2,800 systems over 12 years; about 18% of our work each year is repeat customers or referrals. The longevity question is fair — is there anything specific you'd want to verify before close, like recent reviews, the BBB record, or a reference from a customer in your neighborhood we've worked with in the last six months?"

30-40% accept the stated information and close; 40% ask to see a verification item; 20% close because the verification depth signals shop confidence; 5-10% don't close (a legitimate honest outcome). For PE-backed shops, the AI rebuttal addresses the portfolio backing directly: "We're part of [portfolio name], which adds long-term financial stability and ensures the warranty is backed at the portfolio level — meaning our warranty is structurally stronger than a single-shop guarantee." Transparency about the structure wins more often than obscuring it.

Building the Prompt Library and the Monthly Objection Audit

The 10 objections are the 2026 floor. The shop's prompt library grows from there — the eleventh objection (the cousin-says-you-got-sold-to-PE story), the twelfth (the rebate-program-just-ended specific), the thirteenth (the brand-doesn't-make-that-model rumor). Each new objection that closes gets a prompt template built around it; each objection that stalls gets a prompt template iterated until close rate moves.

The named workflow: first Wednesday of each month, the service manager pulls every kitchen-table close from the prior month from Rilla's transcript log. AI tags each transcript by objection encountered, rebuttal used, and outcome (closed / scheduled follow-up / lost). Output is a heatmap: which objections most frequently appeared, which rebuttals closed best, which rebuttals stalled. Service manager reviews with advisor team in 45 minutes; new objection patterns get new prompt templates; stalling rebuttals get constraint iteration. The library version-bumps; bulletin-board catches feed back into the constraint field; next month's audit measures whether the iteration moved close rate.

Compound effect across 12 months. Year one of objection-library discipline at a 4-advisor shop: starting close rate 42% on $5K+ tickets, ending close rate 58% (Rilla coaching + objection library + post-stack effective payment math compound). The objection library alone — independent of Rilla and the credit stack — contributes 6-9 percentage points of the lift via faster, shop-voice-tuned, deal-context-specific rebuttals. At a $14,200 average replacement ticket × 1,920 annual leads × 8 percentage points incremental close = ~$2.18M annual incremental closed revenue. The library is the highest-leverage prompt asset the shop builds across L2, after the credit stack itself.

The Verify Discipline and the Failure Modes the Library Defends

The Cardinal Rule applies to every AI rebuttal — verify before delivery. The 30-second pass with five checkpoints runs on every AI rebuttal that touches a customer commitment. Most common verify catches: hallucinated APR or financing payment math (Reg Z exposure if delivered), fabricated rebate amounts (truth-in-advertising), invented warranty terms, incorrect system specs (SEER2/AFUE/HSPF2 mismatch with the actual proposal), brand-voice drift toward "partner," "family," scarcity, or first-name address against the shop's voice constraint.

Three failure modes the library specifically defends against. One — generic rebuttal drift. Advisor reaches for a saved template without filling context fields; AI produces generic SaaS-prose; advisor reads it; homeowner senses machine; close collapses. Defense: prompt template requires all five context inputs (name, ticket size, financing tier, credit stack, discovery concern); incomplete inputs produce "context missing" flag rather than fluent generic output. Two — Reg Z exposure on advisor-paraphrased financing language. Advisor reads AI rebuttal silently and paraphrases the financing payment math, introducing a small numeric error. Defense: read the AI output verbatim; constraint discipline locked portal-sourced APR. Three — eye-contact collapse during typing. Advisor types during the homeowner's speech, breaks rhythm, signals device-distraction; trust drops. Defense: typing happens only during natural pauses or after the advisor's verbal acknowledgment of the objection.

The "AI Caught a Hallucination" bulletin board from L1 applies. Across a year of rebuttals, the board surfaces systematic AI weaknesses (rebate-amount hallucinations, APR fabrications, warranty-year inventions) and feeds them back into the prompt library's constraint field. The library hardens; verify catches drop in frequency; close rate compounds.

Key Takeaways

• 10 objections, 10 prompts, 10-second answers. "I need to think about it," "Let me get other bids," "Can you do it for cash," "My buddy can do it cheaper," "I'll wait until summer," "I need to talk to my spouse," "The other quote was $4K less," "Is this in stock," "Will my utility rebate cover this," "How long has this company been around."
• Three tablet mechanics protect eye contact — tablet tilted away from customer (always); advisor never types during homeowner speech; rebuttal read aloud verbatim, never silently paraphrased.
• "I need to think about it" closes 45-55% with AI permission-to-think rebuttal (vs. 15-20% on memorized scripts). Re-anchors to monthly payment, surfaces what specifically they're thinking, offers same-evening callback for spouse input.
• "Let me get other bids" closes 60-65% within 72 hours via call-back when AI surfaces apples-to-apples discipline (SEER2, warranty, manufacturer, financing tier, rebate stack). Never bash competitors — math wins.
• "Can you do it for cash" splits 30-40% legitimate / 60-70% off-the-books. Firm-but-warm "I can't" on off-the-books preserves license while preserving relationship.
• "My buddy can do it cheaper" closes 50-60% via warranty + 25C credit math. Buddy stays respected; the licensing + manufacturer dealer cert + permit + credit eligibility separates the paths.
• "I'll wait" closes 35-45% in the same conversation when AI surfaces failure-risk on aged system + program-year credit/rebate considerations honestly (no scarcity language).
• "I need to talk to my spouse" routes three scenarios — spouse absent (60-70% close on follow-up), spouse present disagreeing (40-50% close after explicit-conversation pivot), spouse veto implied (15-25%, respect the no).
• "The other quote was $4K less" wins the apples-to-apples conversation — 5 comparison points. 40% concede the comparison wasn't apples-to-apples; 30% close on math; 20% legitimate competitive loss.
• Monthly objection audit on the first Wednesday — AI tags every Rilla transcript by objection / rebuttal / outcome; service manager reviews heatmap in 45 minutes. Library compounds; year-one contribution: 6-9 percentage-point close rate lift (~$2.18M annual incremental at 4-advisor team).