Your dead estimates leak more cash than your missed calls
Industry data from May shows contractors miss 60-80% of after-hours calls, and that gets the headlines. The quieter leak is bigger. On estimates above $5,000, owners stop following up after 72 hours about 41% of the time, and booked-job rate on those bids drops by more than half after day three. A 20-truck plumbing shop writing 30 estimates a week at a $4,800 average ticket loses roughly $30,000 a month to bids that just go cold. That number is not theoretical. It is sitting in your CRM right now, tagged “sent” with no activity.
The good news is that almost every field service platform you already pay for shipped automation to plug this exact leak in the last 60 days. Most owners have not turned it on. The May 2026 product releases from Housecall Pro and ServiceTitan changed the in-the-box capability set on this workflow, and the math has moved. If your last serious look at automated estimate followup was in 2024, the picture today is different in your favor.
The five-touch stack that wakes stale bids back up
A working estimate followup stack has five touches and one rule: every touch has to be channel-appropriate and voice-on-brand. Touch one fires at T+24 hours by SMS. SMS open rates on small-business outreach run near 98%, compared to about 20% for email, so you lead with text. Use a short message that names the job and a specific next step, not “checking in.” Touch two fires at T+72 hours by email with the proposal PDF attached and a one-line FAQ. Touch three lands at T+7 days as a second SMS, this time with a soft deadline.
Touch four fires at T+14 days by voice. Big-ticket bids get a real human callback. Mid-tier bids get an AI voice agent calling on the owner’s behalf. Touch five drops at T+30 days as a closeout email asking if the work has been deferred or awarded elsewhere, which is the data point that lets you actually price your win rate. Without touch five, you cannot tell a price problem from a followup problem, and most shops have a followup problem dressed up as a price problem.
That five-step pattern is not new. What changed is who runs it. Two years ago you needed a dedicated CSR plus a separate marketing tool license to make it work. Today the same sequence lives inside your field service software, fires on the same trigger your dispatcher already uses, and costs almost nothing to switch on. The reason nobody runs it well is that the trigger has to come from the CRM event, not from a calendar reminder a coordinator forgets to set.
Why Housecall Pro and ServiceTitan changed the math in May 2026
Housecall Pro shipped its May 2026 platform update last week. Two pieces of that release matter for estimate followup. The first is what HCP calls its “AI Team,” a set of bots that book jobs, handle admin, and now run smarter campaign triggers off CRM events including estimate sent, estimate viewed, and estimate stale. The second is the rebuilt mobile app, which lets a tech mark “estimate sent” from the truck without going back to the office, so the trigger fires the same day instead of a day late. Pricing for HCP’s MAX plan with the AI add-ons sits around $239 per month per office user as of this writing.
ServiceTitan moved at the higher end. Atlas, their conversational AI assistant, is now live in dispatch and is paired with Dispatch Pro AI re-evaluating every job assignment every 10 minutes during business hours. Atlas also drafts auto-generated proposals, which matters because if your proposals leave the office two days faster, your followup window opens two days earlier. ServiceTitan pricing is custom, but most shops above 10 trucks land in the $400 to $600 per office user per month band before add-ons. Jobber sits in between, and its 2026 release added route-centric scheduling and a more aggressive review and followup campaign trigger.
For an owner running fewer than 30 trucks, the picture is simple. You probably already pay one of these three vendors. Each of them now ships a viable estimate-followup engine in the box. You do not need to add a Birdeye or a Podium for this specific workflow. You need to log into your existing platform’s automations panel and turn on the campaign trigger that fires on “estimate sent, no activity” at 24, 72, 168, and 336 hours.
The voice AI piece nobody is pricing correctly
The fourth touch in the stack is the one most shops skip, and it is the one that prints money on bids over $5,000. A voice AI agent calling on day 14 hits an estimate that has already been viewed, with a contact who has already received two SMS and one email touch. That is a warm call, not a cold one. NextPhone, LeadTruffle, and AnswerForce all sell into trades. Pricing in 2026 runs from about $0.18 per minute on usage plans up to flat $399 to $899 a month on unlimited plans for a single shop. A single recovered $8,000 install pays for the entire year of the unlimited plan.
The trap is that owners hear “voice AI” and either dismiss it as gimmicky or try to bolt it onto inbound call answering, which is a different problem with a different stack. Use voice AI for outbound followup on high-ticket bids first. The script is short, predictable, and has a clean handoff to a human if the customer wants to move forward. AnswerForce specifically has documented experience with HVAC, plumbing, and electrical work, which means the agent does not stumble on the words “condenser” or “service panel.”
One operator tip from the morning’s research: tie every review and every followup touch back to the specific tech who did the work. Housecall Pro and the Birdeye-style review platforms both support employee-level attribution now. That single setting turns reviews and followups from a marketing report into a coaching tool, because you can see which techs leave estimates that close on their own and which leave estimates that go cold.
The decision to make before June 1
The decision for this month is not whether to add a new vendor. It is whether to spend 90 minutes inside the field service software you already pay for, turn on the estimate-followup campaign that ships in the box, and assign one person to own the trigger setup. If you run Housecall Pro, log in to Automations, build the four-step sequence above on the “estimate sent, no activity” trigger, and confirm the new May 2026 mobile app is letting techs fire the trigger from the truck. If you run ServiceTitan, ask your CSM specifically about Atlas-drafted proposals and the Dispatch Pro AI bundle, because if Atlas is shaving 48 hours off your proposal-out time, the followup window opens two days earlier and your close rate moves before you change anything else.
If you run Jobber, turn on the review and followup campaign triggers and pair them with the route-centric scheduling that landed in the 2026 release. The labor side of this matters too. Industry reports out this month put time-to-hire for a skilled tradesperson at 56 days, two days longer than time-to-hire for a programmer, and the Department of Labor is now embedding AI training into every Registered Apprenticeship program. Plugging the dead-estimate leak is one of the few plays that buys you revenue without adding a single hire. That is the actual move for the next 90 days. You do not need to find a new tech. You need to stop letting the bids you have already written quietly die.
We at AU build the playbooks and the integration work for trades operators who want to switch this on without burning a week of their own time. If that sounds like the right help, the AU contact link is in the bio below.
About the Author
Trevor Kaak is the founder of Atlas Unchained, a portfolio of products and services helping local businesses run leaner with AI — from custom websites to vendor-bidding marketplaces to vertical SaaS. He writes about marketing, automation, and the craft of building software for operators who’d rather work on their business than in it.