What happens after the first reply.

After the first replyHow many times should you follow up with a lead before you stop?A written standard for following up on inbound inquiries: what counts as a follow-up, what each nudge must contain, when the sequence must stop, and how to close a lead with a reason instead of letting it fade.Policy desk · 11 min · 1 October 2026 Startups & support@Should a startup use AI for customer support? What founders who lost customers say, and the three missing controlsCustomers accept AI support when they can reach a person. The failures that cost startups customers trace to three missing controls, not to the model.Data desk · 11 min · 1 October 2026 New channelsGoogle's Business Profile AI agent: what "Handled by your AI agent" and "Needs your attention" meanMessaging came back to some Google Business Profiles in June 2026, with an AI agent answering first. Here is what the two message tabs mean, what the agent reads, and what is still left for you.Explainers · 10 min · 1 October 2026 Definitions & dataHow to audit your last 90 days of inquiries: the Lost Lead Audit method, template and agency versionA spreadsheet, a handful of Gmail searches and five outcome labels will show you how many inquiries you never answered, how many stopped after one reply, and what they were worth.From the Valet team · 11 min · 1 October 2026 Trust & AI handoffThe customer who fired a contractor over a bot: what "never stuck with a bot" has to meanA homeowner dropped a plumber because the AI receptionist could not pass him to a person. The public cases since then show where handoffs break, and what a working one has to promise.Field reports · 11 min · 1 October 2026 The owner's inboxThe contact-form leak: why 42.6% of website inquiries never get a reply, and how to prove you repliedSome form leads die in the spam folder. Most arrive safely and nobody answers. Here is how to check both in ten minutes, and how to prove a reply went out.People at work · 10 min · 1 October 2026 Tools & pricingWhat counts as an AI "resolution"? Seven vendors' billing definitions, side by sidePer-resolution pricing is only as fair as its definition. Here is what Fin, HubSpot Breeze, Zendesk, Help Scout, Gorgias, Tidio Lyro and Clio Grow AI each bill as a resolution or outcome, what you can verify yourself, and why Intercom declined to price per booking.Policy desk · 11 min · 1 October 2026 Startups & support@How to handle support@ as a solo or 3-person SaaS founder (without hiring, and without a bot that traps people)Sort support@ into four categories, give each a written promise and an automation level, and check the handoffs every week. Here is the runbook, plus the six ways it breaks.Playbooks · 12 min · 1 October 2026 Stores & orders"Will this fit?" "Do you ship to Ireland?" The 53% of pre-sales questions that arrive after hours, and what they're worthGorgias's 2026 data puts a number on the shopper who asks at 10pm and buys elsewhere. Here is a worksheet to price that wait for your own store, and three ways the next 12 months could change it as AI shopping assistants start asking the questions.Trends desk · 11 min · 1 October 2026 Weddings & eventsWhy couples ghost after you send pricing, and what the second message should sayThe usual answers are price, a hidden price list, or a slow reply. The bigger leak is that most vendors send one reply and stop, while the couple is still choosing between five of you.Strategy desk · 12 min · 1 October 2026 Definitions & dataThe lead response statistics everyone quotes, traced to their sourceMost of the numbers on speed-to-lead pages are 2007 to 2014 vintage, misattributed, or invented. Here is the citation chain for each one, and the newer studies with a sample size you can name.Technical desk · 13 min · 1 October 2026 After the first replyWhy leads go silent after you send a quote (and what it actually means)Silence after a quote is usually read as a verdict on price. The base rates say it is more often an unfinished conversation, and the thread itself carries signals that tell you which kind you have.Trends desk · 9 min · 1 October 2026 After the first replyWhen should you follow up after sending a quote: 2 days, 7 days or 2 weeks? (the 2-2-2 rule, tested)The 2-2-2 rule, the five-to-seven-day rule and the ten-day rule all count from the day you hit send. The clock that decides whether a follow-up lands is the customer's, and it runs at different speeds for different jobs.Strategy desk · 9 min · 1 October 2026 After the first replyWhat to say when a customer doesn't respond to your quote: 2nd and 3rd follow-ups that aren't "just checking in""Just checking in" fails for a reason that has little to do with tone. It hands the customer a chore. The follow-ups that get answered make replying take one word.Strategy desk · 8 min · 1 October 2026 After the first replyHow to revive old leads: re-engaging inquiries that went quiet 30, 60 or 90 days agoAn inbox full of quiet inquiries is mostly a record of messages nobody sent. A revival pass sorts the backlog by what actually decides whether a lead is live, reopens each one in its original thread, and closes the rest with a reason.Technical desk · 9 min · 1 October 2026 After the first reply"We lost the deal because we replied on Monday": after-hours and weekend inquiries, and who answers while you're doing the workA Friday-afternoon inquiry in a business closed at weekends waits about 66 hours for Monday. The numbers on what that wait costs, the principle we built Valet around, and a response-time policy you can copy.From the Valet team · 10 min · 1 October 2026 After the first replyHow fast should a small business reply to an inquiry email in 2026? Benchmarks by channelWhat secret shoppers, marketplaces and owner surveys measured between 2024 and 2026, channel by channel, and why the number to watch is the time to the first useful reply.Field reports · 8 min · 1 October 2026 After the first replyHow to write a quote email that gets a reply: ask for the booking, offer real timesMost quote emails end the conversation instead of continuing it. The evidence on what is missing, an annotated template you can copy, and where quoting is heading as AI assistants start reading quotes for customers.Trends desk · 9 min · 1 October 2026 After the first replyWhen follow-up automation backfires: drip sequences, the fourth email, and spam complaintsA drip sequence keeps sending until its last step. A customer who already said no, booked by phone or asked a real question still gets the next email. What that costs, where Gmail draws the line, and what follow-up with judgement sends instead.Field reports · 9 min · 1 October 2026 After the first replyOne hour late = 46% fewer hires: what the Fiverr / Management Science study means off-platformThe headline number comes from a marketplace with a visible response-time badge and buyers comparing sellers side by side. Here is what carries over to a consultant's Gmail or a photographer's contact form, what doesn't, and how to measure your own reply times.Data desk · 9 min · 1 October 2026 After the first replyThe Lost Inquiry Calculator: what unanswered emails, forms and marketplace leads cost you per yearEvery lost-leads calculator we found counts missed phone calls. This one counts written inquiries, shows its formula, and tells you where the borrowed defaults stop being true for your business.Field reports · 10 min · 1 October 2026 After the first replyHow to follow up without being pushy (what customers say they tolerate)Owners worry about sounding desperate; customers complain about vendors who won't stop. Read side by side, the two complaints point at the same defect, and it is not the number of messages.Technical desk · 10 min · 1 October 2026 Weddings & events"Our leads from The Knot / WeddingWire never respond": ghost leads, fake leads, or leads nobody followed up?Vendors say the leads are fake. The Knot says they are templated and ghosted. A U.S. senator has asked the FTC to look. None of them has published what happened to each inquiry, and that is the only record that settles it for your listing.Policy desk · 10 min · 1 October 2026 Weddings & eventsHow many times should you follow up with a couple who hasn't replied? (And when the lead is dead)Wedding advice runs from "one more time than you are now" to "as many times as it takes". Nobody says when to stop. Here is a stop rule, a list of lost reasons, and the three messages that sit between the first reply and the end.People at work · 11 min · 1 October 2026 Weddings & events"Why are wedding vendors so hard to reach?" Where your inquiry goes, when to move on, and the 2026 mystery-shop methodMost wedding vendors are one or two people who are at someone else's wedding when your message arrives. Here is where your inquiry goes, what mystery shops in other trades found, when to move on, and how the first wedding-specific shop is being run.Explainers · 9 min · 1 October 2026 Weddings & eventsWhy couples don't book after the venue tour (the dead end after "we'll think about it")The tour is the most expensive hour in a venue's sales process, and the step after it is the least managed. Here is what the evidence says, what it does not, and a hold-expiry workflow that turns "we'll think about it" into a date.Trends desk · 10 min · 1 October 2026 Weddings & events"50% of couples book the first vendor who replies": where that stat comes from, and what "first" has to includeThe figure has circulated in wedding-vendor education since 2018 with no survey, sample or date behind it. Speed still matters. But the first reply is becoming automatic, so "first" now has to mean the first useful answer.Trends desk · 9 min · 1 October 2026 Weddings & eventsWhen do couples send wedding inquiries? Nights, weekends, holidays, and who answers while you're at a weddingCouples write when they have time, which is when you are shooting, hosting or setting up. The Monday catch-up and the 'I'm at a wedding' autoresponder both fail for the same reason, and the fix is a weekend reply standard you can write down.Field reports · 10 min · 1 October 2026 Weddings & eventsCouples contact five vendors per category: a venue's reply-time math, and how to book without being the cheapestA five-way comparison gives every venue a fair share of one booking in five. Price is how most venues try to beat that share. The cheaper lever is being the easiest conversation in the comparison.From the Valet team · 8 min · 1 October 2026 Weddings & eventsWedding inquiry response template, and the two messages after it that nobody templatesThe first reply is the message every template pack covers and most software now sends automatically. The offer of a time, the nudge and the close are where bookings are decided, and they need rules as much as wording.Policy desk · 9 min · 1 October 2026 Weddings & eventsThe 10 questions every couple asks every wedding vendor: build the facts sheet onceCouples' inquiries are converging on the same handful of questions, and the next people asking them may be AI assistants. The vendor who has the answers written down is ready for every version of that future.Trends desk · 9 min · 1 October 2026 Weddings & eventsEngagement season for vendors: the January inquiry flood and the leads that die in itJanuary is when couples who got engaged over the holidays start writing to venues, photographers and planners. The first reply usually goes out. The second one is what the flood washes away.Policy desk · 9 min · 1 October 2026 Weddings & eventsWhat a good inquiry-to-booking rate is for a wedding venue (2026 benchmarks, and the number nobody tracks)The published benchmark is about 20%. The benchmarks disagree with each other below that line, and the step where most venues leak, answered but never offered a tour time, is the one nobody reports.Playbooks · 8 min · 1 October 2026 Weddings & eventsVenue tour no-shows: what a 70–85% show rate is worth and the reminder sequence that moves itThe show-rate benchmark every venue blog repeats comes from one vendor's worked example. The money behind it is real enough to act on, and so is the message almost nobody sends after a couple misses the tour.Policy desk · 8 min · 1 October 2026 Independent professionals"What are your rates?" by email: answering pricing and pre-sales questions without a call, a lost lead or an invented discountThe person asking for your rates is not trying to trap you. They are trying to find out if you are in range before they spend a call on you. Answer that question, then give them one easy next step.Explainers · 9 min · 1 October 2026 Independent professionalsDiscovery-call no-shows for consultants and coaches: real rates, why they happen, and what actually fixes themA no-show is the last visible failure in a booking path that broke earlier. This piece traces that path stage by stage, separates the sourced numbers from the recycled ones, and ends with a policy and recovery sequence you can copy.Technical desk · 11 min · 1 October 2026 Independent professionalsWho answers your inquiries when you're on holiday, sick or on parental leave? Coverage without stranding customers with a botSolo owners describe two choices: switch on the out-of-office and hope, or check email from the beach. Both price your absence badly. Here is the case for a third arrangement, where it fails, and a checklist to set it up.Strategy desk · 10 min · 1 October 2026 Independent professionalsClient ghosted after the proposal or quote: a follow-up plan for consultants, designers and small agenciesSilence after a proposal is the most common ending in independent work, and the least discussed. What the people who send proposals say about it, what the evidence says about the second and third message, and a plan that ends with a reason instead of a fade.People at work · 8 min · 1 October 2026 Independent professionalsHow to qualify leads before a call without an intake form that scares people off (and without a sales team)The standard advice is to add friction: a longer form, a shorter free call, a paid consult. It works, and it has a cost that rarely gets counted. The alternative is to qualify by answering, and to let prospects rule themselves out.Strategy desk · 9 min · 1 October 2026 Independent professionalsShould you charge for discovery calls? Free vs paid vs "answer first, then book"Free calls leak time. Paid calls filter people out, sometimes the wrong ones. A third option answers the written questions first, so only people who still want to talk reach your calendar.Explainers · 9 min · 1 October 2026 Independent professionalsBooking link vs email back-and-forth: why clients email instead of using your Calendly, and what to do about itA booking link moves the scheduling work to the client. Many clients decline it without saying so. Here is a standard for when to send a link, when to offer times in the thread, and how to confirm.Policy desk · 8 min · 1 October 2026 Independent professionalsReply speed for tutoring and music-lesson enquiries: why parents vanish, and what a good first reply containsParents and students who enquire about lessons are usually writing to more than one teacher. The first useful reply tends to win. Here is what useful means, with a template built from a music teacher's own.Explainers · 8 min · 1 October 2026 Independent professionalsAccountants and bookkeepers: quoting by email, the "too expensive" silence, and keeping the lead without discountingA 2017 AccountingWEB thread shows how a bookkeeping quote unravels by email: the price objection, the discount, the counter-offer. The fix is in what the quote says and what happens after it, not in the number.Field reports · 9 min · 1 October 2026 Independent professionalsCoaches: prospects who book, get the payment link and disappear. Follow-up with judgement, not pressureA discovery call goes well, the payment link goes out, and then nothing. This post traces where that pipeline breaks, which explanations usually turn out wrong, and the post-call message set that closes the loop without pressure.Technical desk · 10 min · 1 October 2026 Startups & support@An AI agent that answers vs one that finishes the job: what "acts" means for a small business, and the four controls that make it safeMost AI tools sold to small businesses answer the first message and stop. This piece defines what it means for an agent to finish the job, traces where answer-only tools lose the inquiry, and sets out the four controls that make acting safe.Technical desk · 11 min · 1 October 2026 Startups & support@Why inbound leads don't convert at a tiny startup: the demo-request and sales@ inbox nobody readsAt a seed-stage startup, inbound usually dies in the reply, not the lead. This runbook shows how to test your own demo-request flow, set up sales@ so every request gets a complete answer, and catch the failure modes before they cost a deal.Playbooks · 10 min · 1 October 2026 Startups & support@Demo-request follow-up without a sales team: the two-nudge sequence and when to stopSales playbooks recommend eight to twelve touches. A founder answering inbound demo requests needs something different: one complete reply, two nudges that each carry something new, and a clear point at which to stop.From the Valet team · 9 min · 1 October 2026 Startups & support@"If your articles are vague, Fin will be vague": the facts sheet and hard limits a store or SaaS needs before any AI answers customersThe public cases of bots inventing policies and discounts trace back to the source and the limits, not the model. Here is the one-page facts sheet, the authority table and the test to run before go-live.Data desk · 10 min · 1 October 2026 Startups & support@Why your FAQ page doesn't reduce support emails (customers ask about their order or account), and what doesThe FAQ answers the category. The email asks about one customer's case. Until the reply can apply your facts to that case, the same questions keep arriving, however good the FAQ is.Strategy desk · 9 min · 1 October 2026 Startups & support@"Intercom is too expensive for a startup": the cost-creep arithmetic (seats, outcomes, add-ons) and the four ways outThe bill that surprises founders is not one price. It is seats, add-ons and per-outcome fees stacking up just as a startup discount ends. Here is the arithmetic, what each alternative charges when the bot gets stuck, and the signals to watch.Trends desk · 8 min · 1 October 2026 Startups & support@Overnight and cross-time-zone support for a 3-person SaaS: what to automate at 3 a.m., what to promise, what never to let the bot decideA 3-person SaaS cannot staff the night and should not hand it to a bot that decides refunds. The answer at 3 a.m. is a fact, or a reply time the customer can plan around.Strategy desk · 9 min · 1 October 2026 Startups & support@When should a startup make its first support hire, and what to run until then? (the $45–65k question)A first support hire costs roughly $45,000 to $65,000 a year. Before you post the job, find out which emails a person actually has to answer. The founder's inbox already holds that list.People at work · 10 min · 1 October 2026 Startups & support@Build your own support@ bot (n8n, Zapier or Make plus ChatGPT) or buy one: what it costs in hours, and the four things it still won't doThe model is the cheap part. Upkeep, refusals, handoffs and the record of what was said are where a home-built support@ bot costs hours, and the public build reports are a sample of about four.Data desk · 8 min · 1 October 2026 Startups & support@The support@ audit: what a seed-stage SaaS inbox contains (pre-sales vs support vs billing), and why the resolution rate is the wrong first numberVendors quote AI resolution rates between roughly 30% and 76%. That spread is a measure of inbox mix and billing definitions, not of the tools. Nobody has published what a seed-stage support@ inbox contains, so here is how to measure your own.Strategy desk · 9 min · 1 October 2026 Startups & support@Plain vs Pylon vs Help Scout vs Crisp vs Chatwoot vs Fernand vs Front for a 5-person startup: what each charges for AI, and what happens when the bot is stuckSeven support tools, priced from their own pages on 1 October 2026 for a five-person team. Seat prices are easy to compare. The AI unit, what counts as one, and what happens when the bot cannot answer are not, and only two pages define them.Policy desk · 10 min · 1 October 2026 Startups & support@The agent that remembers: account context and "what was promised" in AI support, and why most AI forgets at the handoffA small business is held to what its assistant told a customer, yet most AI support drops that record the moment a person takes over. This piece separates what an agent must know from what it may know, and gives you a promise ledger to keep.Policy desk · 10 min · 1 October 2026 Stores & ordersHow to handle "where is my order" emails in a small store: answer from the order, act on it, stay until deliveredA tracking link answers the question a store expects. The email asks a different one: what is happening to my parcel, and what will you do about it. Here are the base rates, the five states behind the question, and the reply each one deserves.Trends desk · 8 min · 1 October 2026 Stores & ordersThe 48 hours after "order confirmed": why the status page didn't stop the emails, and the three silent states behind themStores answer the post-purchase silence with a status page or a stream of automatic emails. Both are broadcasts. The emails keep coming because the customer is asking about one parcel in one of three states the page cannot explain.Trends desk · 8 min · 1 October 2026 Stores & orders"Can I still cancel?" and "I put the wrong address": the fulfilment-cutoff clock and allowed actions for a one-person storeA cancel or an address fix costs a click before the label prints and a return trip after it. For a store run by one person, the email that asks usually arrives while that person is asleep.People at work · 8 min · 1 October 2026 Stores & ordersRefunds, store credit and returns within policy: the money-actions rules an agent can execute without inventing a refundAn agent that touches money needs the same thing a new cashier needs: a written policy, a float it may hand out, and a manager's key for everything else. Here is how to write those three things for a store or a SaaS product.Explainers · 11 min · 1 October 2026 Stores & ordersHow fast should an online store reply to customer emails? 2026 benchmarks, and the "9pm test"One reply-time target for every email is the wrong unit. A question asked before checkout, a cancel request before the label prints and a sizing complaint after delivery each have their own clock. Here are the 2026 numbers and a test you can run on your own store tonight.Explainers · 9 min · 1 October 2026 Stores & orders"We'll be back within 24 hours": a teardown of small-business auto-replies, and what the first reply should do insteadThe auto-reply is the most-sent email in many small businesses, and almost nobody reads it twice. Here is what each of its stock sentences tells a customer at 9pm, and a first reply that answers instead.People at work · 9 min · 1 October 2026 Stores & ordersWho answers support@ when a one-person store or startup takes a holiday? Pause vs keep selling with cover (the holiday math)Pausing the store and leaving an auto-reply on support@ look like the two safe options. Run the numbers for a 14-day absence and both leave money and customers waiting. Here is the arithmetic, its limits, and the checklist for a third option.Data desk · 9 min · 1 October 2026 Stores & ordersHow much does customer support cost a small online store per order?Tickets per 100 orders, times the cost of one ticket, plus the part most budgets leave out: the 55% of AI-touched tickets that come back to a person. A calculator you can fill in with your own numbers.From the Valet team · 9 min · 1 October 2026 Stores & ordersBest AI customer service for a small Shopify store (2026): verified prices and the gap each leavesEight options priced for a 300-order store from pages read on 1 October 2026. The usual advice is to pick the lowest price per resolution. The better question is what happens to the 55% of conversations the AI hands back.Strategy desk · 10 min · 1 October 2026 Stores & ordersGorgias vs Zendesk for a small store: what each costs at 300 and 1,000 orders a month, and what both leave on your deskTwo worked monthly bills, at 300 and 1,000 orders, from list prices read on 1 October 2026. Then the column neither pricing page has: the tickets that still land on you.From the Valet team · 9 min · 1 October 2026 Trust & AI handoffCan a quote from my AI be binding? What BMW Toronto taught every business that prices by emailA chatbot at a Toronto BMW dealer offered a customer about CAD $7,000 more than the dealer meant to pay. The number came from the dealer's own records, in the wrong place. Here is what that means for any business that quotes in writing, and a policy you can copy.From the Valet team · 9 min · 1 October 2026 Trust & AI handoffWhat to ask an AI vendor before it talks to your customers: 12 questions (limits, handoff, receipts, disclosure, undo, insurance)The demo shows the AI writing a good reply. The questions that decide whether you keep your customers are about what happens when it should not reply at all.People at work · 10 min · 1 October 2026 Trust & AI handoffIs my business responsible for what my AI tells a customer? Air Canada, BMW Toronto, Cursor "Sam": the case timeline, and whether you have to honour itA passenger, a car trader and a software developer each took an AI at its word. In every public case so far, the business ended up owning the answer.People at work · 8 min · 1 October 2026 Trust & AI handoffAI handoff policy template: when the bot must stop, who gets it, the context packet, and the reply-by promiseAn AI assistant's handoff settings ship on a factory default. One test showed what that default did to a £3,900 price-match request, and what a written policy needs to say instead.People at work · 9 min · 1 October 2026 Trust & AI handoffShould AI reply to customers automatically or draft for approval? The approval matrix by action typeApproving every draft is slow, and full autopilot will one day promise something you never offered. The answer is a rule per action, written down before the first customer writes in.People at work · 8 min · 1 October 2026 Trust & AI handoffThe handoff gap: one-third of AI-handed-off tickets never get a human reply. How to measure yours, and what "resolved" should meanAI support vendors report resolutions. Customers experience handoffs. This runbook measures the gap between the two in your own helpdesk in about 90 minutes, and gives you six ticket states you can verify without the vendor's word.Playbooks · 9 min · 1 October 2026 Trust & AI handoffWhat an AI mistake with a customer actually costs (and what making it right costs): a calculatorAn AI mistake costs a small business in two ways: the customers who leave when they get trapped, and the price of making a wrong promise right. Here is a worksheet for both, the public cases behind it, and the protocol we would follow.From the Valet team · 9 min · 1 October 2026 Trust & AI handoffDoes telling customers it's AI hurt bookings? What the 2025–26 evidence saysThe most-cited number says disclosure cuts purchases by four-fifths. It comes from one 2019 experiment on outbound sales calls. Here is what that study measured, what the 2025–26 evidence adds, and what it costs when customers find out late.Data desk · 8 min · 1 October 2026 Trust & AI handoffDo I have to tell customers they're talking to AI? The 2026–27 map in plain English, with copy-paste disclosure linesIn the EU, yes, since 2 August 2026, unless it is obvious. In the US it depends on the state, and the main laws start in 2027. One plain disclosure, made at the first message, covers nearly every version. Here is the map and the lines we would use.From the Valet team · 8 min · 1 October 2026 Trust & AI handoffCustomers don't hate AI, they hate being trapped: the 2025–26 numbers and what a small business should do differentlyEvery recent survey on AI service carries the same condition: the customer can reach a person. Here are the dated numbers, and a step-by-step way to make sure your own setup never traps anyone.Playbooks · 10 min · 1 October 2026 Trust & AI handoff"I don't trust AI with my customers": the objection taken seriously, and the six controls that answer itThe owners who refuse to let AI near their customers are reading the incident record correctly. The fix is not a more human-sounding bot. It is six controls that limit what the AI may say and leave a record of what it did.Technical desk · 10 min · 1 October 2026 Trust & AI handoffDoes my business insurance cover an AI mistake? CG 40 47 and the generative-AI exclusion, for a 3-person businessInsurers now have standard wording to exclude generative-AI claims from a general liability policy. Whether your policy carries it is a fact you can check in ten minutes. What to look for, what it could leave uncovered, and what to ask.Policy desk · 8 min · 1 October 2026 New channelsWhat happens when a customer asks ChatGPT or Yelp Assistant to contact my business?AI assistants now ask businesses for quotes on a customer's behalf. The request does not arrive labelled as AI. Here is where each one lands, what to check first, and a three-message reply sequence.Playbooks · 9 min · 1 October 2026 New channelsWhy marketplace leads die in your inbox (Thumbtack, Bark, Angi, The Knot, Yelp): the lead-email problem, and what you can actually controlA paid lead arrives as a notification email, shared with other pros, and the clock starts before you open it. You cannot fix lead quality from your side. You can fix what happens in the first hour and the next five days.From the Valet team · 10 min · 1 October 2026 New channelsHow fast do you have to answer a marketplace lead to win it? Platform benchmarks and the "responds in about X" badgeEvery marketplace now tells pros to reply faster, and most of them sell a tool that sends the first reply for you. When everyone answers in two minutes, the clock that decides the job is a different one.From the Valet team · 8 min · 1 October 2026 New channelsCost per answered lead: Thumbtack, Angi, Bark, Yelp, The Knot and WeddingWire in 2026, and whether the listing is worth itLead prices are the number everyone argues about. The number that decides whether a listing pays is cost per booking, and half of it depends on what happens after the lead arrives.Data desk · 9 min · 1 October 2026 New channelsMeta's Business Agent: what it answers (DMs), what it doesn't (email, forms, marketplace leads), and what happens after it hands offMeta's agent can now answer the DMs on a small business's phone. It does not see the email inbox, the website form or the marketplace lead, and when it passes a chat to a person, it goes quiet.People at work · 9 min · 1 October 2026 New channelsAI agent on your Google listing, Meta's agent in your DMs, nobody in your inbox: the orphan-channel map, 2026 editionEvery platform now answers the first message in its own app. This runbook tests each of your channels, marks the ones nobody answers, and gives each orphan an owner and a follow-up rule.Playbooks · 10 min · 1 October 2026 New channels45% of consumers now ask an AI for local recommendations and ChatGPT names 1.2% of businesses: what happens after an AI names youEveryone sells the part where an AI names you. This is the part after it: the website check, the question, the reply that comes the next morning, and the follow-up that never comes.Field reports · 8 min · 1 October 2026 New channelsGoogle's AI called my business asking for prices: what it is, why it matters, and what to have readyThe automated call is not a scam, and opting out is not free. It is a price comparison run on the customer's behalf, and you win or lose it on whether your price was written down before the phone rang.Strategy desk · 9 min · 1 October 2026 New channelsShould I publish my prices? The AI will ask anyway, and 48% of businesses can't answerCustom-work businesses hide prices for good reasons. Buyers, and now the AI assistants asking for them, treat a missing number as a no. The middle path is a written range and a rule against improvising.People at work · 9 min · 1 October 2026 New channelsGoogle AI Mode now books tables and appointments, through partner platforms. What if you're not on one?AI Mode's booking feature runs through the reservation and scheduling platforms Google partners with. A wedding venue, a photographer or a consultant on none of them is reached another way: by questions. Here is the map, three scenarios and what to watch.Trends desk · 10 min · 1 October 2026 New channelsWhat does "agent-ready" mean for a small business? When ChatGPT, Gemini or Yelp Assistant asks about price and availabilityThe phrase gets sold as markup, feeds and AI text files. For a small business the measurable version is simpler and harder: when an assistant asks what you charge and when you are free, does it get an answer?Trends desk · 8 min · 1 October 2026 New channelsAgentic commerce in plain English (Oct 2026): what shipped, who is eligible, and what it means off-platformCheckout protocols, agent payments and AI booking all shipped within a year. Almost all of it runs through platforms. For a small business on none of them, agentic commerce arrives as a question about price and availability, and that is where the money is lost.From the Valet team · 10 min · 1 October 2026 New channelsGoogle now charges for missed LSA calls (from 1 Oct 2026): what every unanswered lead costs in money and rankA missed call on Local Services Ads used to cost you a customer. From 1 October 2026 it can also cost you the lead fee. Here is how the rule works, a worked example with real numbers, and a calculator for calls and messages.Explainers · 8 min · 1 October 2026 The owner's inboxYou've answered that question a hundred times: FAQ page vs answering emails vs answering from a facts sheetOwners treat the FAQ page as the asset and the replies as the chore. Both are copies of something most small businesses never write down: one maintained list of facts that a person, a page or an AI can quote.Strategy desk · 10 min · 1 October 2026 The owner's inboxWhat does it actually cost to answer one customer inquiry? Receptionist vs VA vs AI (2026 numbers)Every price list in this market counts contacts, minutes or seats. Owners who wrote reviews about the bill were counting something else: what each booking ended up costing. Here are 12 dated prices and the calculator that converts them.Field reports · 8 min · 1 October 2026 The owner's inboxHire a VA, a receptionist, or software? VA vs AI receptionist vs AI inquiry handler: what each actually does with an inquiryMost comparisons line up hourly rates against monthly plans. The data that predicts lost customers is elsewhere: who follows up, who hands off, and whether a human is reachable at all.Data desk · 8 min · 1 October 2026 The owner's inboxWhat an inquiry costs you in your own time: the "36% of the week on admin" number, traced and applied to a solo professionalThe figure travels as "Xero says owners spend 36% of the week on admin". Followed hop by hop, it leads to a 2023 survey of 251 entrepreneurs that left out freelancers and sole traders. Here is the chain, what it does and does not support, and how to measure your own number instead.Technical desk · 8 min · 1 October 2026 The owner's inboxAI receptionist complaints in owners' and customers' own words: cost per call, and why written inquiries are the safer place to startThe recurring complaint about AI receptionists is not the voice. It is the call that cannot be transferred, the caller asked twice for the same details, and the high-value request answered with a factory setting.Policy desk · 11 min · 1 October 2026 Tools & pricingFirst-reply tools compared (2026): Meta Business Agent, Podium, Perfect Venue, GoHighLevel, HoneyBook AI, Tripleseat, Calendly Callie, Google's GBP agent, and what happens after the first replyEvery tool on this list can send the first reply. We read their pricing pages and help docs to see which ones send the second, what they charge for it, and what you sign.From the Valet team · 10 min · 1 October 2026 Tools & pricingPerfect Venue First Reply and AI Reply (and Perfect Venue vs Tripleseat): "everything after that is you"Perfect Venue's help centre describes First Reply in one honest sentence. This is a teardown of the inquiry path around that sentence: which inquiries trigger it, what happens at minute six, and where the second message has to come from.Technical desk · 9 min · 1 October 2026 Tools & pricingPodium pricing and the AI Employee ("Jerry"): what small businesses actually pay, and the contract fine printPodium's pricing page has no prices on it. The figures that circulate come from competitors and aggregators, they disagree, and the number that matters most is not monthly at all. It is twelve.Data desk · 7 min · 1 October 2026 Tools & pricingAI email drafters and schedulers (Fyxer, Lindy, Superhuman AI, Calendly Callie, Agentic Calendars) vs an inquiry handler: why "drafts right 50% of the time" isn't follow-throughDrafters save the owner's keystrokes and schedulers coordinate the owner's calendar. The customer who wrote in still needs an answer from your facts, a held time and a second message. Here is where each tool stops.Trends desk · 10 min · 1 October 2026 Tools & pricingHoneyBook AI for photographers, planners and venues: drafts, price hikes, and what still lands on youHoneyBook now puts AI on every plan. It writes the reply; you still press Send, chase the couple who went quiet and decide what to do about price. Here is what the AI does, what it costs, and how to cover the gap without leaving HoneyBook.Trends desk · 7 min · 1 October 2026 Tools & pricingGoHighLevel for a small business (not an agency): is it worth it, and what's the simpler alternative?GoHighLevel is a whole marketing platform that agencies rent out to their clients. If all you wanted was every inquiry answered and followed up, here is what you are actually buying, what it takes to set up, and when a smaller tool fits better.Explainers · 8 min · 1 October 2026 Tools & pricingSmith.ai pricing explained: what a call really costs once you add booking, follow-up and recordingSmith.ai's receptionist plans are priced per call, and the add-ons are priced per call too. Here is the arithmetic from the live pricing page, where third-party estimates diverge, and what the bill measures.Technical desk · 8 min · 1 October 2026 Tools & pricingWhich small-business AI agents take actions, ask for approval and keep a log? Six agents comparedGoogle's GBP agent, Meta's Business Agent, Shopify Sidekick, HubSpot Breeze, Fin and Yelp Assistant, read against their own documentation on 1 October 2026. Six columns, one "verified from" column, and a list of what nobody publishes.Policy desk · 10 min · 1 October 2026 Tools & pricingAI receptionist free trials and annual plans: the $299-on-day-8 problemCard-up-front trials and annual plans are sold as low-risk. For a front-office tool they move the risk of a tool that does not work onto the owner, before the evidence arrives. Here is the argument, the terms, and the checklist.Strategy desk · 9 min · 1 October 2026 Agencies"My client doesn't reply to the leads we send": the agency's lead follow-through problemYour campaign delivers the lead. Then it lands in an inbox you don't control. Here is where those leads go, what it costs the client, and how an agency can prove it without taking over anyone's email.Explainers · 8 min · 1 October 2026 AgenciesContact form "works" but leads disappear: the enquiry-system test for web designersThe form submits, the plugin says sent, and the client still says nobody ever got in touch. This is a forensic walk down the path from submit button to booked customer, with a 10-point test you can run on every site you hand over.Technical desk · 8 min · 1 October 2026 AgenciesWedding and event marketing agencies: the leads you generate die in the venue's inboxVenue lead volume is rising and the lead-quality fight is loud. The signal agencies misread is quieter: what happens to an inquiry after the first reply. A map of the forces, three scenarios, and the indicators to watch.Trends desk · 8 min · 1 October 2026 AgenciesHow to prove your website generates leads when the client never replies to themA monthly lead-handling report separates the leads your site delivered from the leads your client answered. Here is the runbook, the template, and what to do when the client pushes back.Playbooks · 8 min · 1 October 2026 AgenciesGoogle Ads leads not converting? Read the reply log before you touch the campaignConversion tracking stops at the form submit. The failure is often one hop later, in an inbox nobody is measuring. Here is the diagnostic order the Technical desk uses, and the log it starts from.Technical desk · 9 min · 1 October 2026 Definitions & dataInbound inquiry response and follow-up statistics (2026), by verticalEvery number here names its sample, its channel and its year. The table shows which verticals have measured follow-up at all, and the one statistic nobody has published yet: inquiries answered once and then dropped.Playbooks · 8 min · 1 October 2026 Definitions & dataThe inquiry-state glossary: 11 states from "new" to "booked" or "lost with a reason", with exit rulesMost lead pipelines have entry rules and no exit rules, so inquiries pile up in "contacted". This glossary defines 11 states, what moves an inquiry out of each one, and what the customer is told at every step.Technical desk · 10 min · 1 October 2026 Definitions & dataWhat is follow-through rate? (and why first-reply rate is the wrong number to track)A fast first reply is now easy to buy, and it says little about whether a customer was looked after. Follow-through rate counts what happened next: how many inquiries reached a real ending.People at work · 9 min · 1 October 2026 Definitions & dataWhat counts as a booking made by Valet: the public outcome-attribution ruleValet charges a fee only for bookings or quotes it wins. This page states exactly what that means, what does not count, how an owner checks a fee against their own calendar, and how the rule compares with Intercom, HubSpot, Clio and Tidio.From the Valet team · 9 min · 1 October 2026 Definitions & dataWhat is an action receipt for an AI agent? "What did the AI say, and why?" The eight fields, an example and the undo windowOwners ask one question about an AI that talks to customers: what did it say, and why? An action receipt answers it for each message. This is the definition, the eight fields, a worked example, what can be undone, and a spec anyone can copy.From the Valet team · 8 min · 1 October 2026