How to set SLAs for lead response time

Lead-response SLAs are a discipline before they are a number. The most common mistake is to set a target ("respond within 1 hour") before measuring current behavior. The second most common is to set one global target instead of segmenting by source and market.

Both mistakes lead to SLAs that the team either ignores or games. This guide is about setting SLAs that hold.

Measure first, target second

Before you set a target, measure what your current response time actually is. Two weeks of data is enough for a baseline.

For each new lead in the period, record:

• Source
• Market
• Time of arrival
• Time of first contact (call, email, or other initial outreach)
• The owner who responded

Compute the median and 90th percentile time-to-first-contact per (market, source) combination. The 90th percentile is more useful than the median because it tells you where the long tail is.

A typical baseline looks like:

• Meta Lead Ads, India: median 47 minutes, p90 4 hours.
• HubSpot inbound, US: median 2 hours, p90 18 hours.
• LinkedIn ads, EMEA: median 6 hours, p90 36 hours.

The baseline tells you what the team can do today. The target tells you what they should do tomorrow. Set the target a meaningful but achievable step beyond the baseline.

Set targets per (market, source, state)

A single global SLA ("respond within 1 hour for everything") is wrong. It is too aggressive for low-intent sources and too lenient for high-intent ones.

The right structure is a matrix. For each combination of market and source and lead state, set a target:

Adjust per market based on business hours and team coverage.

Layer escalations on the window

The SLA target is one number. The escalation thresholds are a percentage of it.

A clean escalation policy:

• At 80% of the window. Notify the owner. They have a chance to act before the breach.
• At 100%. Breach is recorded. Notify the owner's manager.
• At 150%. Auto-reassign to another available rep. The lead does not sit unassigned forever.

The threshold percentages are stable across SLAs. The window changes per (market, source, state). The platform applies the same percentage to whichever window applies.

Pause outside business hours

Business hours pausing is essential and easy to get wrong.

A lead that arrives at 3 AM local time should not breach a 1-hour SLA by 4 AM. The SLA timer should pause when the market's business hours are closed and resume when they open.

Configure business hours per market: typical office hours, observed weekends (Friday-Saturday in some Gulf countries, Saturday-Sunday in most others), and configured public holidays.

Without business-hour pausing, you have two equally bad options: SLAs that always breach overnight, or SLAs that ignore overnight breaches and miss real ones. With pausing, the SLA is honest.

Decide what counts as "responded"

Surprisingly contentious in practice. Three common definitions:

Strict: first outbound activity from the owner. A logged call, sent email, or completed task. Anything else does not count.

Lenient: any owner activity on the lead. Including viewing the lead, adding a note, changing the state.

Practical: configurable per organization. The platform supports both modes; the customer picks.

The strict definition is honest but harsh. A rep who reviews a lead but does not actually contact the prospect is not really responding. The lenient definition is gameable; a rep can satisfy the SLA by clicking the lead once.

Most well-run organizations use a hybrid: the SLA is satisfied by either a logged outbound activity or a documented decision to disqualify. Both are explicit acts of work.

Report on it weekly

An SLA policy that no one reviews is a policy that decays. Operations should review SLA performance weekly:

• Breach count per market, per source, this week vs last.
• p90 response time per (market, source) trend.
• Top offending owners (not for shaming; for coaching).
• Top offending sources (some sources are genuinely lower-intent and the SLA should be loosened).

Leadership should see the same numbers monthly, rolled up. If operations and leadership see different SLA numbers, the policy is at risk.

Calibrate quarterly

Sales motion changes. New sources come online. Teams grow. The SLA policy is not set once; it is calibrated.

Quarterly:

• Review breach trends. If a (market, source) is consistently breaching at 20%+, the target is too aggressive for current capacity. Either invest in capacity or loosen the target.
• Review p90 vs target gaps. If p90 is far below the target, the target is too lenient; tighten it.
• Retire SLAs for sources you no longer use. Reduce policy complexity.
• Add SLAs for new sources you have started using.

This quarterly review is 30 minutes of operations work. It prevents the slow drift that turns a sharp SLA policy into a vague one.

Common pitfalls

Three patterns to avoid:

Setting a single global SLA. Either too aggressive or too lenient for most of your sources. Always segment.

Setting SLAs without measuring. You will pick a number out of a benchmark report and the team will treat it as theater. Always baseline first.

Treating SLA as a vanity metric. The point is to drive response time and detect breach in real time, not to produce a dashboard. If the SLA does not fire workflows that change owner behavior, it is decoration.

For how MegatronLead expresses SLA policy and escalation rules, see SLA and escalation. For the workflow engine that fires the escalations, see workflow automation.