On this page
The direct answer
What speed to lead actually means
Speed to lead is the elapsed time between the moment a prospective customer contacts you and the moment they receive a meaningful reply. It applies across inbound channels such as website forms, Google Local Services Ads, Yelp, Thumbtack, SMS and WhatsApp, and phone or missed calls.
Two words in that definition do most of the work. Meaningful means the reply advances the conversation, not just an automated acknowledgement that reads like a receipt. Reply means an actual response, not a system generated bounce. If your only response for the first hour is a note that says you got the message and will be in touch, your effective speed to lead is still one hour, regardless of what any dashboard shows.
What the original studies measured
Two data sets get cited constantly. Both are older than they sound, and both were designed for a different world than a plumbing shop in 2026. They are still useful, but only if you read them for what they actually say.
The Harvard Business Review audit
In 2011, Harvard Business Review published an audit of 2,241 US companies (see The Short Life of Online Sales Leads). Among those that did reply within 30 days of a web form submission, the average first response time was 42 hours. Twenty three percent never replied at all.
A separate dataset in the same article looked at 1.25 million sales leads across 29 business to consumer and 13 business to business companies. It reported that companies that attempted contact within an hour of receiving a lead were nearly seven times as likely to qualify that lead as companies that waited even an hour longer, and more than sixty times as likely as companies that waited a full day. Note the word qualify. That is not the same as close, and it is not the same as book the job. It means the seller managed a real conversation with a decision maker.
The Lead Response Management study
The often quoted five minute rule traces to a 2007 study by InsideSales.com in collaboration with Professor James Oldroyd, then at MIT (see the original PDF). It analyzed three years of data across six companies, more than 15,000 web generated leads, and over 100,000 outbound call attempts.
The study's most cited finding: the odds of making contact with a lead were roughly 100 times higher when the first call went out within five minutes of the form submission versus after 30 minutes, and the odds of qualifying that lead were roughly 21 times higher. Again, qualifying was defined per company and did not equal winning the job.
The study measured contact and qualification odds, not revenue or close rates. It also predates smartphones being universal, predates modern messaging channels, and predates AI first response tools that can hold a real short conversation without a human sitting at a keyboard.
Both studies point in the same direction: replying quickly noticeably increases the odds of a conversation. That is a real, defensible pattern. It is not a guarantee that a faster reply wins the job, and cutting response time in half does not double revenue.
A few honest caveats worth keeping in mind.
- Both data sets are from B2B and B2C sales pipelines that look different from a plumber getting a leaking dishwasher message on a Saturday night. Some of the dynamics carry over. Some do not.
- Odds ratios are relative, not absolute. One hundred times more likely to make contact was measured against a specific comparison window and a specific type of lead. It does not mean one hundred times more revenue.
- Customer expectations have shifted since these studies were published. Some platforms now expose response information to customers; Google Local Services Ads, for example, may display an estimated messaging response time on the profile.
- Faster is not a substitute for better. A quick, wrong, or robotic answer can lose a job that a slower, thoughtful answer would have won.
Treat the research as directional evidence that responding within minutes is meaningfully better than responding in hours.
A meaningful first response usually does three small things at once. It acknowledges the specific request. We can help with a leaking kitchen sink reads differently than thanks for reaching out. It asks a real qualifying question that the next step will depend on: zip code, urgency, or whether the customer has already shut off the water. It offers a concrete next step: a callback window, a preferred time question, or a live handoff to someone on the team.
That is enough. It is not a full quote, and it should not try to be. Chasing a one shot pricing reply usually costs more time than it saves.
The bands below are an illustrative example of an internal service level agreement a small team might set for itself. They are not findings from the cited studies and they are not universal benchmarks. Use them as a starting frame for your own SLA, then replace them with numbers your own audit says are realistic.
| Scenario | Illustrative SLA | Guidance |
|---|---|---|
| Emergency inbound | 2 min · 24/7 | Emergency inbound, where a homeowner is describing possible active water, gas, or no heat in winter, warrants immediate action. The workflow should follow pre-approved escalation and safety guidance and involve a human dispatcher or the appropriate emergency service without delay. Automation must not improvise safety advice. As an illustrative acknowledgement target, aim for a first meaningful reply within two minutes, 24 hours a day, alongside the human escalation. |
| Standard, business hours | 5 min | Standard inbound during business hours, meaning a routine repair, quote request, or appointment request, might target a first meaningful reply within five minutes. |
| Standard, after hours | 5 min | Standard inbound after business hours might target an honest, personalized acknowledgement within five minutes, plus a clear next available window and confirmation that a human will follow up in the morning. |
| Missed inbound calls | 5 min | Missed inbound calls might target either a return call or a text back with a preferred time question or a clear callback expectation, within five minutes. |
- Emergency inbound2 min · 24/7
Emergency inbound, where a homeowner is describing possible active water, gas, or no heat in winter, warrants immediate action. The workflow should follow pre-approved escalation and safety guidance and involve a human dispatcher or the appropriate emergency service without delay. Automation must not improvise safety advice. As an illustrative acknowledgement target, aim for a first meaningful reply within two minutes, 24 hours a day, alongside the human escalation.
- Standard, business hours5 min
Standard inbound during business hours, meaning a routine repair, quote request, or appointment request, might target a first meaningful reply within five minutes.
- Standard, after hours5 min
Standard inbound after business hours might target an honest, personalized acknowledgement within five minutes, plus a clear next available window and confirmation that a human will follow up in the morning.
- Missed inbound calls5 min
Missed inbound calls might target either a return call or a text back with a preferred time question or a clear callback expectation, within five minutes.
These are not targets to write on a wall. They are numbers to measure yourself against, honestly, in a self audit.
A 30 day lead response self audit
Before buying tools or restructuring your team, run a 30 day audit. You need surprisingly little to do it well.
For every inbound lead across every channel, capture the same fields on each row:
- 01Arrival timestamp, meaning when the message actually hit your business.
- 02First meaningful response timestamp, not an auto acknowledgement.
- 03Channel.
- 04Whether the lead arrived inside or outside business hours.
- 05Initial owner, meaning the person or automation that took the first reply.
- 06Whether the lead qualified into a scheduled appointment or job.
- 07Whether a preferred day and time was confirmed with the customer.
- 08Handoff outcome, meaning what happened when the lead was passed on.
- 09Closed revenue attributed to that lead, where available.
Once the month is collected, calculate the median and 75th percentile response time overall, and again broken down by channel and by inside or outside business hours. Median and percentile numbers are audit outputs, not per lead fields.
Read the results carefully rather than reaching for a story. A few patterns worth investigating:
- A large gap between median and 75th percentile response time often points at a small number of very slow outliers rather than a broad slowdown.
- Response times that jump on evenings, weekends, or specific channels can indicate an ownership or coverage gap.
- Leads that qualify but never confirm a preferred time may indicate a handoff problem, not a first reply problem.
Treat these as hypotheses to test, not conclusions.
If you already use InstantResponse.AI, use the response and outcome fields available in your centralized dashboard to seed this audit, and export or supplement them with a spreadsheet where needed. Not every audit field may exist as a first class report today. If you do not use the product, a shared spreadsheet for 30 days is enough to reach a real answer.
Once you have real numbers, slow response is often a workflow problem, not simply a motivation problem. Teams are often slow not because they do not care, but because ownership is unclear.
- Assign one clear owner per channel per shift. Whoever sees it first is not an owner.
- Define office hours per channel, and define what happens outside them. A rule that says Yelp gets a human response from 8 AM to 6 PM and an AI first response the rest of the time is workable. A rule that says someone will check Yelp eventually is not.
- Define an explicit fallback. If the primary owner does not respond within a set number of minutes, who takes ownership next, and what happens if they are on a job site.
- Write down what a good first reply looks like for each channel. Not a script to read verbatim, a template so no one has to think from scratch at 9:47 PM.
- Review the second reply and handoff too. Delays can reappear after the first message, especially when a customer has responded and no one is watching the inbox.
Automation is genuinely good at the boring, high volume part of speed to lead. Being awake at 11:14 PM when a lead lands. Sending a first meaningful reply that acknowledges the specific request and asks a useful qualifying question. Live confirming a preferred day and time from your configured availability using AI Settings. Sharing a configured rate card range so the customer has real information, not a placeholder. Making sure a lead does not sit unread in one channel while everyone is busy in another.
Automation is not a replacement for a human on the last mile. Any conversation that changes scope significantly. Emergencies with real safety implications. Any customer who signals frustration, confusion, or a complicated situation the AI has not seen before.
For readers considering InstantResponse.AI, the accurate picture as of this writing is as follows.
Customer specific configured rate card quoting is live.
LiveAI Settings can live confirm a preferred day and time from your configured availability today.
LiveFull end to end appointment and calendar booking is currently in beta.
BetaNamed CRM lead and job creation for ServiceTitan, Housecall Pro, Jobber, Workiz, and GoHighLevel is available today via Zapier and or webhooks, not as native or direct integrations.
Zapier / webhooksVoice AI answering on your existing phone number is in advanced beta.
Advanced BetaGoogle Local Services Ads message leads flow into a unified inbox; see the LSA integration page for current scope.
Live
On Google Local Services Ads specifically, Google's own documentation (managing leads and jobs) recommends responding as soon as leads arrive, notes that estimated messaging response times may be surfaced to customers, and says that a consistently fast response could improve ad ranking. It does not disclose a formula, and it does not promise a specific outcome tied to speed alone. Treat faster replies as one factor Google identifies among several, not as a lever with a guaranteed payoff. We do not extend the same claim to other platforms.
Sources and research notes
- Harvard Business Review, The Short Life of Online Sales Leads (March 2011): hbr.org/2011/03/the-short-life-of-online-sales-leads
- InsideSales.com / MIT Lead Response Management Study (2007): InsideSales.com / MIT study (PDF)
- Google Local Services Ads, Manage leads and jobs: support.google.com/localservices/answer/6224859
- Google Local Services Ads, Reports: support.google.com/localservices/answer/6226265
FAQ
What is a good average lead response time for a local service business?
Under five minutes for standard inbound during business hours, and under two minutes for emergencies at any hour, is one illustrative starting frame a small team might adopt for its own SLA. Your own numbers matter more than any benchmark. Measure your median and 75th percentile for 30 days before you commit to a number publicly, and revise the frame with what you learn.
Does responding in five minutes actually win jobs, or is that outdated?
The five minute figure comes from a 2007 study on B2B web leads and outbound calls, so treat it as directional rather than a law. The pattern it points to, much better contact odds when the first reply is fast, still holds up as guidance for local service messaging, but it does not guarantee revenue, and it does not mean four minutes is dramatically better than six.
How do we handle nights, weekends, and holidays without burning out staff?
Combine an on call rotation for true emergencies with an automated first response for everything else. The automated reply should acknowledge the specific request, offer the earliest realistic window, and set a clear expectation for when a human will follow up.
Do faster replies improve Google Local Services Ads ranking?
For Google Local Services Ads, Google's own documentation (see managing leads and jobs) recommends responding as soon as leads arrive, notes that estimated response times may be surfaced to customers, and says that a consistently fast response could improve ad ranking, without disclosing a formula or promising a specific outcome. Faster replies also help you win the conversation that Google sent you, which is a strong operational reason on its own. This applies to Google Local Services Ads only.
How should we measure speed to lead if we do not have a CRM?
For 30 days, a shared spreadsheet with arrival timestamp, first meaningful response timestamp, channel, whether the lead qualified, and any closed revenue is enough to see the shape of your response times. Calculate median and 75th percentile from those rows. You do not need advanced attribution to know that your 75th percentile response is four hours.
Speed only matters if the first reply is actually useful. For a vendor neutral framework on choosing among voicemail, a human answering service, an AI front desk, or a hybrid, see AI front desk vs answering service vs voicemail: a buyer's framework.
Summary and next step
Speed to lead is a lever, not a magic number. The research supports responding in minutes rather than hours. It does not support any specific promise about revenue. What matters is a meaningful first reply, clear ownership across shifts, an honest picture of your own 30 day numbers, and an automation layer that handles the parts humans should not have to.
If you want to see how an AI first response would sound on your own leads, request a short demo and start the 14 day free trial. Billing begins after the trial, month to month, cancel anytime. Pricing is on the pricing page.