Speed to Lead in 2026: the 5-Minute Rule Is Half the Answer

In July 2026 a post from Corey Ganim did the rounds on X, arguing that "speed to lead agents are a million dollar startup idea" — an AI agent that answers inbound leads "in under 5 minutes with a useful, personalized response", sold to businesses after you mystery-shop their own form and show them how slow they are (source, 16 July 2026). Nearly 200 likes, dozens of replies, and not one of them arguing the premise.

The premise is sound. The benchmark data agrees. What almost nobody says out loud is the second half of the sentence: a five-minute reply to a lead who was never going to buy is not a win. It is the same waste, delivered faster.

Speed to lead, answered directly

Speed to lead is the elapsed time between a prospect showing interest and your first contact with them. The industry target is under 5 minutes, and the reason it works is competitive rather than psychological: most buyers contact several vendors at once, so the first credible response usually takes the conversation. In B2B the correct implementation is not "reply to everything in 5 minutes" — it is qualify in seconds, then reply in minutes to the leads that fit, and send everything else an instant automated acknowledgement. Speed multiplies whatever it is applied to, including waste.

The benchmark: the bar is on the floor

A benchmark rundown published on 6 July 2026 by GrowthSpree put numbers on how bad the average is (source):

• The median B2B SaaS team takes 42 to 47 hours to respond to an inbound lead.
• In a 1,000-company test, 63.5% never responded at all. Not late — never.
• Leads contacted within five minutes convert at roughly 21-32%, against 2.3-12% for teams that wait a day.

Read the second bullet again. Nearly two thirds of companies fail the test completely. This is why speed to lead has survived as advice for fifteen years: the bar is so low that simply showing up on time is a differentiator.

But notice what the third bullet does not say. It does not say fast replies cause conversion. Fast-responding teams are usually also the teams with tighter targeting, better routing and fewer junk leads in the funnel. Speed is correlated with a functioning revenue operation, and it is the cheapest part of it to copy — which is why so many teams copy only that part and wonder why the number did not move.

Why the standard speed-to-lead playbook does not survive B2B

Search "speed to lead" today and the results are dominated by real estate, home services and insurance: motivated-seller marketplaces, roofers, wholesalers. That is not an accident. Those businesses have a genuinely undifferentiated buy — one roof, one quote, one closing — where the first caller wins by default and qualification barely matters.

B2B breaks three assumptions that playbook depends on:

• The buyer is not one person. A form fill from an intern researching a procurement doc looks identical to one from the VP who owns the budget.
• Fit is not obvious from the form. An email domain and a company name do not tell you headcount, sector, maturity or whether they can even use your product.
• Rep time is the scarce resource. A roofer's five-minute callback costs five minutes. An AE's costs a context switch out of a live deal.

Approach	What it optimises	Typical first-touch latency	Where it breaks
Email alert to a shared inbox	Nothing. It is a notification, not a process.	Hours to days	Nobody owns it, so everybody assumes someone else took it.
Round-robin routing rules	Fair distribution across reps	Seconds to assign, hours to contact	Distributes junk as evenly as gold. Reps learn to distrust the queue and cherry-pick.
Instant scheduler on the form	Removing the back-and-forth for high-intent buyers	Zero, if they book	Only converts the already-convinced. Fills calendars with unqualified demos nobody screened.
AI qualify-then-route	Rep attention, by filtering before assigning	Seconds to score, minutes to contact the ones that fit	Depends entirely on the quality of the scoring model and the data behind it. A hallucinated firmographic is worse than no score.

The 5-minute tax: run the numbers on your own funnel

Here is a calculation we have not seen anyone publish, and it is the one that decides whether a blanket 5-minute SLA is a good idea for you. Call it the 5-minute tax: the rep hours you spend per month buying speed on leads that will never buy.

The inputs are three numbers you already have:

• L — inbound leads per month
• F — your ICP-fit rate (what share of inbound is genuinely targetable)
• C — the true cost of one urgent response, in rep minutes. Not the call length: the interrupt, the scramble for context, the attempt, the log, and the re-entry into whatever they were doing. Twelve minutes is a conservative figure for an AE.

Then:

• Blanket SLA cost = L × C
• Qualify-first cost = (L × F) × C
• The 5-minute tax = L × (1 − F) × C

Put a realistic mid-market funnel through it — 200 inbound leads a month, a 30% ICP-fit rate, 12 minutes per urgent response:

• Blanket SLA: 200 × 12 = 40 hours a month, about a quarter of one rep's selling time.
• Qualify-first: 60 × 12 = 12 hours a month, at the same 5-minute speed.
• The 5-minute tax: 28 rep-hours a month spent being impressively fast at people who were never going to buy.

Twenty-eight hours is most of a working week, every month, per team. And the tax scales with the thing most companies are actively trying to increase: inbound volume. Double your lead flow without touching your fit rate and you double the tax, not the pipeline.

The fix is not to slow down. It is to move qualification in front of the SLA, so the five-minute promise only ever applies to the 30%.

Tiered SLAs: what "fast" should actually mean

One SLA for all inbound is the root error. Three tiers, assigned automatically at capture, cover almost every B2B funnel:

• Tier A — strong ICP fit, decision-maker seniority. Under 5 minutes, by phone, by a named owner. This is where the 21-32% conversion band lives, and it is worth interrupting a rep for.
• Tier B — company fits, person does not (or vice versa). Under an hour, by email, with a booking link. Fast enough to stay warm, cheap enough not to burn an AE.
• Tier C — off-target, competitor, student, wrong geography. Instant automated acknowledgement, nurture list, no human. Rejecting cleanly is a feature; leaving them unanswered is how you end up in the 63.5%.

Tiering only works if the tier is decided by something that actually reads the company and the person. That is a scoring problem, not a routing problem — which is why teams who buy a routing tool first usually end up back at the drawing board. If you have not defined what qualified means yet, start with our B2B lead qualification framework and the difference between MQL and SQL before you buy anything.

How Lead Scorer handles the qualification half

Lead Scorer is an AI SDR cofounder, so its centre of gravity is outbound rather than inbound forms. But the engine that makes tiered SLAs possible is exactly the piece it was built around: two-level scoring. It scores the company on ICP fit and the person on decision-making relevance, then combines the two and rejects off-target leads with a written reason instead of silently ranking them low.

Two things about that matter for speed. First, the score arrives with the reason attached, so the rep who picks up a Tier A lead is not starting from a name and a domain. Second, discovery runs against the French State company registry (recherche-entreprises.api.gouv.fr, feeding SIRENE and the INPI RNE) rather than a scraped database, so the SIREN, the legal form and the actual dirigeant are verified facts. A fast response built on an invented firmographic is worse than a slow one, and this is the failure mode nobody markets around.

On the outbound side the same clock applies with a different trigger. The SDR agent finds fresh companies, scores them, drafts the LinkedIn and email sequence anchored on real profile facts, and has a second model (Mistral) review every message before you see it. Your time-to-first-touch on a new signal is a queue you approve, not a week of list-building. Plans start at €49/month — see pricing — and if you are weighing it against a classic sequencer, Lead Scorer vs Apollo covers the trade-off directly.

The counter-argument, taken seriously

Not everyone is convinced. Harsh Makadia put it flatly on X in late July 2026: "Speed-to-Lead Is Not About Replying Faster" (source, 27 July 2026). And from the automotive side, Brett Sutherlin's widely-shared LinkedIn take argues the industry "has become obsessed with speed-to-lead while completely ignoring persistence-to-close".

Both are right about the failure mode. A team that hits a 3-minute median and then never follows up again has optimised the opening move of a game it still loses. The 2026 benchmarks that make speed look decisive are measuring first contact, not the eight touches after it. Pair your speed SLA with a real cadence — our sales cadence benchmarks lay out what that looks like — or you will simply arrive first and leave first.

A build checklist you can run this quarter

1. Mystery-shop yourself. Fill in your own form from a personal address on a Friday afternoon. Time the response. Most teams discover they are in the 63.5%.
2. Measure F, not just speed. Take last month's inbound and mark each lead genuinely targetable or not. That percentage is the number that decides your SLA design.
3. Compute your 5-minute tax with the formula above. Put the hours in front of whoever owns the sales budget.
4. Score before you route. Company fit and person fit, separately, with a written rejection reason. No score, no tier.
5. Set three SLAs, not one. 5 minutes, 1 hour, automated. Publish them so reps know what they owe.
6. Instrument the whole path. Capture timestamp, score timestamp, assignment timestamp, first-touch timestamp. Almost every team that "has a speed problem" actually has a routing gap between two of those four.
7. Review the rejects monthly. If good leads are landing in Tier C, your ICP definition is wrong, not your speed.

What to take away

Speed to lead is still one of the few unfair advantages available to a small B2B team, precisely because most competitors are measured in days and a majority never reply at all. But the version of the advice that circulates — reply to everything in five minutes — was written for single-transaction local services, and importing it into B2B buys you a 28-hour monthly tax and a burnt-out AE.

Qualify in seconds, then be fast for the ones that fit. That is the whole play. If you want the scoring layer that makes it possible, see how Lead Scorer scores companies and decision-makers, or read how the same logic applies to B2B buying signals on the outbound side.