Speed to Lead: The Conversion Funnel for a Home-Services Business That Runs on Phone Calls

A funnel that ends in a ringing phone

Most conversion literature is written for a checkout: a cart, a form, a card. A home-services funnel is different in kind. The transaction that matters is a phone call to a plumber, an electrician, a roofer, or a dental or med-spa front desk, and the digital surfaces around it exist only to produce and protect that call. The homeowner with a burst pipe does not comparison-shop a landing page. They reach for a phone, ask an engine or an assistant who to call, and dial the first credible name that answers.

That structure relocates the conversion problem. On a service-area business the decisive events are not on the page at all. They are the milliseconds before the call, when a slow site can lose the visitor before the number is ever seen; the minutes after a lead arrives, when a delayed reply lets a competitor answer first; and the moment of the call itself, when a ring that goes unanswered is a job handed to whoever picks up. The evidence base for each of those three points is strong enough to build an operating model on, provided each source is read at the confidence it actually warrants.

The through-line across the whole body of research is that the largest, most reliable conversion gains for this kind of business are structural, not persuasive. Fewer form fields, faster paint, faster human response, answered phones. The clever headline and the color of the button sit far down the list.

Speed to lead: what an audit of 2,241 companies found

The foundational study on response time is not a vendor blog. In 2011 the Harvard Business Review published "The Short Life of Online Sales Leads" by James Oldroyd, Kristina McElheran, and David Elkington, who audited 2,241 United States companies using mystery-shopper web inquiries and measured how fast each firm actually responded.

The operational findings were stark. Only 37 percent of firms attempted contact within one hour. Twenty-three percent never responded at all. Among the firms that did respond, the average time to first contact was 42 hours, long past the point at which a buyer with an urgent problem has already hired someone else.

The causal finding is the one that should reorganize a phone-driven business. Firms that attempted contact within one hour were nearly seven times as likely to qualify the lead as firms that waited even one hour longer, and more than sixty times as likely as firms that waited a full day or more. The odds of ever having a meaningful conversation with the buyer collapse within the first hour and keep collapsing. For a service-area trade, where the buyer is often mid-emergency, the practical window is far tighter than the study's own one-hour benchmark.

The unanswered call as a measured revenue leak

Response time assumes a lead was captured in the first place. For a business that runs on phone calls, a large share of demand never becomes a lead at all, because the phone rings while a clinician is mid-treatment, a crew is on a roof, or the office has closed for the weekend, and no one picks up.

The call-intelligence platform Invoca has published first-party analysis putting roughly 27 percent of calls to home-services businesses in the unanswered column, and estimating that a single missed call can cost a home-services firm an average of about 1,200 dollars in lost business. Invoca's cross-industry benchmarking separately reports phone leads in home services converting at up to ten to fifteen times the rate of web-form leads once they are answered, which is consistent with the intuition that a person who dials is closer to buying than a person who fills in a form.

These figures should be read at their real evidentiary weight. Invoca is a commercial vendor with an interest in the finding, and the numbers are first-party rather than peer-reviewed or governmental, so this is emerging evidence, not academic fact. What earns them a place in the model is that they point in the same direction as the independently peer-reviewed response-time literature: the cheapest available conversion gain for this business is not persuading a new stranger, it is not losing the buyer who already tried to reach you. A missed-call text-back, an instant automated reply that turns an unanswered ring into a live conversation, is the direct operational response, provided it sends from a properly registered business number and honors opt-outs.

Milliseconds before the call: what a thirty-million-session study found

Before any lead or call exists, the visitor has to survive the page. The largest public study tying speed to commercial outcome is "Milliseconds Make Millions," commissioned by Google and conducted by the agency 55 with Deloitte Digital in 2020, spanning more than thirty million real user sessions across thirty-seven brand sites in Europe and the United States.

The study found that a mobile speed improvement of one-tenth of a second was associated with an 8.4 percent increase in retail conversions and a 9.2 percent increase in average order value. In the travel category the same tenth-of-a-second gain was associated with a 10.1 percent rise in conversions, and in lead generation, the category closest to a home-services quote request, it was associated with an 8.3 percent improvement in bounce rate. A tenth of a second is below the threshold most people can consciously perceive, and it still moved money at population scale.

One caveat travels with this source. The study was commissioned by Google, which has a direct interest in site speed, so it is best treated as strong corroborating evidence rather than fully independent measurement, and the reported figures are associations across a large observational sample, not a controlled experiment isolating speed as the sole cause. Read that way, it remains one of the most useful large-N signals available for the claim that speed is a revenue variable, not a cosmetic one.

Core Web Vitals conversion evidence, read carefully

The speed story has a second, more granular layer in the Core Web Vitals, Google's field-data metrics for loading (Largest Contentful Paint), responsiveness (Interaction to Next Paint), and visual stability (Cumulative Layout Shift), graded at the seventy-fifth percentile of real users.

Google's own web.dev case-study series documents named before-and-after results. Rakuten 24 in Japan reported a 53.37 percent increase in revenue per visitor and a 33.13 percent increase in conversion rate after optimizing Largest Contentful Paint. Vodafone Italy reported an 8 percent increase in sales after improving LCP by 31 percent. The travel marketplace redBus reported a 7 percent sales increase after improving Interaction to Next Paint.

These are genuinely useful, but they are the weakest tier of evidence in this playbook and must be labeled as such. Each is a single-company before-and-after case study, not a controlled experiment across a population, so it functions as an existence proof (speed work can produce outcomes of this magnitude) rather than as a generalizable elasticity (your site will gain 33 percent). The correct use of these numbers is to justify measuring your own Core Web Vitals against the published thresholds, not to promise a specific lift.

The web form is a second phone: friction at the point of contact

Many home-services funnels also carry a quote-request or booking form beside the phone number, and the same friction principle governs it. The Baymard Institute, drawing on more than 300,000 hours of large-scale checkout-usability testing since 2011, finds that the average large site can gain roughly a 35 percent conversion increase from usability fixes alone, and that a well-built checkout needs as few as twelve to fourteen form elements where unaudited sites routinely field twenty or more.

That research was conducted on ecommerce checkouts rather than service-request forms, so applying it to a "book a visit" flow is a reasonable extrapolation rather than a directly measured result, and it belongs in the model as a directional principle: every field you demand before a person can reach you is a place the lead can quit. On a phone-driven business the form exists to produce a call or a booking, so the right design goal is the minimum viable ask, not the maximum data grab.

Assembling the operating model

Read together, the three established pillars (response time, speed, and the friction principle) plus the emerging missed-call and Core Web Vitals layers describe a single funnel with four instrumented points. The operating model is to measure each point, fix the structural fault, and only then reach for persuasion.

• Point one, the page before the call. Hold the site to the Core Web Vitals thresholds Google reads from real users: Largest Contentful Paint at or under 2.5 seconds, Interaction to Next Paint at or under 200 milliseconds, Cumulative Layout Shift at or under 0.1. Speed here is measured revenue protection, not vanity.
• Point two, the form that stands beside the phone. Strip the quote or booking form to the minimum viable ask. Every removed field is a place a lead stops quitting.
• Point three, the reply once a lead lands. Treat the first hour as the whole game, and in an emergency trade, the first minutes. Route new inquiries to a human fast, because the qualification odds decay from the moment the lead arrives.
• Point four, the call that cannot be answered. Instrument missed calls and close them with an instant, compliant text-back that turns a lost ring into a live thread and writes the caller in as a lead with their number and time.
• The measurement wrapper. Because none of these leaks appear in ordinary reporting, the funnel has to be read directly and repeatedly, so a fix can be proven rather than assumed.

How to read this evidence: tiers and caveats

A playbook like this stands or falls on what each number can actually bear. The response-time audit and the large-session speed study are established, peer-institution or large-N sources, and they anchor the model. The missed-call economics are vendor first-party data, directionally consistent with the peer-reviewed literature but not a substitute for it. The single-company Core Web Vitals results are real and publicly documented but are existence proofs, not elasticities. And the checkout research is established within ecommerce but extrapolated when applied to a service form.

Held to that standard, the safe and defensible conclusion is narrow and strong: for a business that runs on phone calls, structural fixes to speed, response, and answer rate are the evidence-backed place to start, ahead of any copy or design tactic. The exact lift any one business will see is not something the literature can promise, which is precisely why the model begins and ends with measuring your own funnel rather than importing someone else's multiplier.