Jeff Conlon

A plumbing client switched to an AI receptionist and watched monthly revenue drop from $1.16M to $814K. The dashboard looked flawless. Zero missed calls. Here is how a 17-point drop in booking rate quietly erased $346,000 in 30 days:

Five software tools to bill one client at our agency. Proposals in one, signatures in another, recurring billing in a third. If an invoice looks off, tracing it takes four open tabs. Agency margin rarely bleeds from ad campaigns. It leaks in the back-office plumbing.

A plumbing client replaced front desk dispatch with an AI receptionist to save on payroll, and revenue dropped from $1.16M to $814K in a single month. Booking rate fell from 62% to 45%, and ad spend stayed flat. Saving money on payroll does not mean much when substantial revenue walks out the door.

I agree with this. I also think if you don’t have agents running most tasks in your agency by the end of 2026 you’re cooked. You won’t be able to compete. We’re entering the era of “super agencies” who generate 10X the results of past agencies and do it at scale.

A client retired a delivery truck and saved about $5K a month. The team deployed an AI agent to handle route and dispatch scheduling, smoothing a queue run by feel. I would rather fix dispatch bottlenecks than buy another truck.

Agencies build 23 data views across multiple boards, but clients only care if the phone rang. Our team rebuilt it down to a single page: reach, engagement, and leads. Now they actually open it. If clients need a walkthrough, the report is the problem.

The single-channel agency model is built on reporting phantom wins to protect retainers. Your PPC buyer reports 84 leads at a $62 CPL. They hit their monthly goal and send an invoice. Then you check the CRM. 19 leads were outside your territory, 31 never picked up the phone, and dispatch booked 8 calls. Effective cost per booked job: $650. The ad buyer didn't lie. Google Ads recorded 84 conversion pixels. When agencies don't tie compensation to closed revenue, they optimize for cheap clicks that choke your front desk. If your marketing doesn't reconcile against your bank deposits, your CPL is just fiction.

Most marketing agency retainers are a tax on business owners who don't know how ad platforms work. An agency charges you $3,000 a month to manage $15,000 in Google Ads spend. They set up 6 ad groups, turn on automated bidding, and check the account for 20 minutes every other Tuesday. If performance drops, they recommend increasing budget so their percentage-of-spend fee goes up with it. If your agency gets paid more when you spend more, they have zero financial incentive to make your marketing efficient. The agencies that survive the next three years won't be media buyers tweaking bid caps. They will be operators who plug into your CRM, fix your intake leak, and tie their fees to pipeline revenue.

If I took over marketing for a $5M HVAC contractor in Dallas tomorrow, here is the exact 4-quadrant plan for Month 1. 1. Win the Game (Q1) Kill broad match on Google Ads. Bidding $80 a click for generic terms like "air conditioning repair" is burning cash. Lock spend down to high-intent emergency keywords and Local Services Ads. Audit your GBP categories and setup direct SMS review links. Stop paying search auction tax on calls you should own for free. 2. Build a Brand (Q2) Carve out 20% of the media budget and deploy it into Meta and YouTube video within 7 miles of your highest-margin neighborhoods. Run 30-second raw clips featuring your technicians, trucks, and flat-rate pricing. When an AC compressor fails in July, you want homeowners searching your brand name, not opening Google to compare five competitors. 3. Return and Retain (Q3) Pull every customer in ServiceTitan who had an install or tune-up between 12 and 24 months ago and hasn't called since. Run an automated 2-way SMS campaign offering an off-peak system check before the summer rush. That simple reactivation can book 30 to 45 calls in a week without spending a dollar on ads. 4. Operations & Intake Standardize your dispatch queue. Inbound system replacement inquiries route directly to a comfort advisor or dedicated sales tech within two minutes. If your CSR scripts a $12,000 heat pump replacement like a routine $89 capacitor swap, your marketing spend is just paying for lost opportunities. Single-channel agencies pitch more ad spend. Full-funnel operators build a machine that compounds.

A 111-year-old commercial contractor came to us with a 14-year-old website pitching $99 residential service calls. A single-channel SEO or PPC agency would have immediately pitched a $5,000 monthly retainer to drive more traffic to that page. Nobody asked what actually drives the business until we audited revenue. Commercial jobs across hospitals, schools, and facilities make up 90% of their billing. Residential is under 8%. Before you spend a dime on ad auctions or organic rankings, your positioning, intake, and capacity have to point at the revenue you actually want. Anything else just burns cash buying the wrong calls.

A PPC buyer looks at a $160.93 blended CPL and puts it on a slide to prove his retainer is working. Our automated audit flagged that exact number last week. Best week on record across the account, four straight weeks under $250. A single-channel vendor stops reading right there. When we connected the ad data to service line intake, one segment was bleeding: water heater leads were sitting at $448 against a $325 target. Full-funnel marketing means refusing to let a cheap emergency drain click cover up an unprofitable replacement campaign.

The truth is that SEO and organic traffic as we know it is over. We have a client who lost 90% of their organic traffic after the rise of AI (the business is doing fine though). The weird part right now is that we don't know what AEO really looks like. We're flying the plane while we build. SparkToro and Gumshoe are doing amazing work with these types of studies because it illuminates part of the AEO field that we're on. But the rest of the field is still dark. Over time we'll have a better understanding of that game we're playing but right now it's all R&D.

Single-channel agencies treat strategy like a slide deck. Once you agree to it on a Zoom call, their job is done. We approved an ad fix last Friday. Everyone agreed, the thread was archived. I opened the account seven days later and nobody had touched a single setting. Same structure, same budget caps, same CPL drifting up. In a full-funnel operation, strategy only begins when settings change inside the ad auction and data flows clean into dispatch. If nobody audits execution a week later, you don't have a marketing strategy. You have an agency hiding behind meeting notes.

One media buyer running 66 ad accounts isn't marketing operations. It's an automated churn factory. On a standard 40-hour week, that buyer has 36 minutes per account. That isn't even enough time to clear negative keywords, let alone check if leads are converting into booked revenue inside the CRM. The buyer stares at dashboard CPL to survive the week. The client manager runs damage control on check-in calls. Nobody audits the dispatch board, nobody tests intake speed, and nobody looks at landing pages. Month 1 looks fine because lead volume is cheap. By month 3 the client realizes zero booked jobs came through and fires the agency. If your ad buyer doesn't have the bandwidth to connect ad spend directly to closed revenue in the CRM, you aren't managing accounts. You're just renting out ad dashboards until the churn catches up.

Ouch! 😂

The bigger a contractor's website gets, the worse its conversion rate becomes. Most owners assume ranking higher and stacking more traffic makes an agency's retainer more efficient. The SearchLight dataset showed the exact opposite across 89 contractors. As a site scales, the new traffic isn't people looking for AC repair. It's DIYers reading how-to articles on clearing a drain. In that audit, blog posts pulled over a third of all organic visits and generated exactly one bookable form lead in thirty days. Meanwhile, homepages drove 68% of bookable forms, and localized city pages converted at twice the rate of generic pages. SEO agencies love reporting massive traffic graphs. But unless that traffic hits high-intent service architecture routed straight to dispatch, you're paying for readers who will never put a truck in a driveway.

The $40B local lead generation market is built on search friction that AI is about to erase. For twenty years, lead brokers built fortunes by buying the top three ad slots on Google, trapping the homeowner in a directory form, and selling that same lead to five local contractors for $150 each. The contractor ate the margin compression. The homeowner got spammed by five sales reps in two minutes. The middleman printed cash. That model collapses the moment homeowners stop searching inside closed directories. When an AI assistant parses local license data, reads real Google Business Profile sentiment, verifies technician capacity, and books directly into dispatch, renting leads from a broker is over. If your growth strategy relies on buying leads from a directory, you don't have a business. You have a subscription to someone else's customer list.

If I took over marketing for a $3M plumbing shop in Phoenix tomorrow, here is the exact 4-quadrant plan for Month 1. 1. Win the Game (Q1) Cap Google Ads to emergency terms only. Audit the GBP: fix primary categories, clean NAP across directories, and turn on review requests via SMS. One dialed GBP pulled 643 direct calls in 5 months with $0 media. Stop overpaying for basic search volume. 2. Build a Brand (Q2) Take 15% of the ad budget out of search and put it into Meta video ads targeted within 5 miles of your highest-margin zip codes. Show the techs, the trucks, and pricing transparency. Warm the 97% of homeowners who don't have a broken pipe today so they don't search Google when they do. 3. Return and Retain (Q3) Export your past 5,000 customers from ServiceTitan. Segment everyone who hasn't had service in 10 months and send a 2-way SMS offer for a pre-season water heater flush. This fills your dispatch board for two weeks without buying another click. 4. Operations & Intake Wire lead tracking directly to closed CRM revenue, not form fills. Retrain dispatch: replacement calls get routed to a senior sales tech within 60 seconds, not a general CSR script. Single-channel guys buy more clicks. Operators build an engine that compounds.

We audited a shop spending $28,000 a month on paid search and organic local SEO. Their emergency capture was fine. Their middle-of-funnel conversion was a disaster. The landing page dropped visitors into a generic form, the CSR took 45 minutes to call back, and replacement calls were scripted like drain cleanings. Half their ad budget was subsidizing bad intake. Fixing the capture channel does nothing if the operational handoff drops the ball on the five-yard line.

A plumbing client replaced front desk dispatch with an AI voice receptionist to save on payroll. Revenue dropped from $1.16M to $814K in thirty days. Ad spend stayed completely flat. The PPC agency checked Google Ads and saw steady conversions. The AI vendor checked logs and saw zero dropped calls. Both declared victory. Meanwhile, booking rate collapsed from 62% to 45% because the bot choked on dispatch nuances and sent high-ticket replacement calls into a dead queue. A single-channel vendor washes their hands at the click. An AI tool vendor washes their hands at the transcript. If your marketing agency doesn't track what happens after the phone answers, you are subsidizing automation that costs you six figures.

Paying lead middlemen works until the search surface shifts under your feet. Right now, aggregators charge you $140 to buy a shared lead they sold to four other contractors. They make money purely on search friction. When a homeowner asks an AI assistant to find a plumber, the model crawls your Google Business Profile reviews, verifies your license on your domain, checks live dispatch availability, and books the call directly. No directory middlemen. No third-party paywall. The contractors who will get wiped out by the AI search shift aren't the ones who missed a prompt hack. It will be the ones who relied on lead brokers to rent them customers instead of building direct brand equity in their own market.