Free guide
Lead routing and SLA benchmarks: what the data actually shows
The five-minute rule is real, older than most people assume, and routinely misquoted. Here is what the original research found, what it doesn't cover, and what to do about it.
Section 1What the research found
Two pieces of work sit under almost every speed-to-lead claim you will read.
The Lead Response Management Study, 2007. Dr James Oldroyd, then at MIT Sloan, analysed roughly 15,000 leads and over 100,000 call attempts across six companies, in partnership with InsideSales.com. Two findings are quoted constantly: the odds of contacting a lead were about 100 times higher when the first call went out at 5 minutes rather than 30, and the odds of qualifying a lead were about 21 times higher over the same interval. Between 5 and 10 minutes, qualification odds had already fallen roughly fourfold.
"The Short Life of Online Sales Leads," Harvard Business Review, 2011. Oldroyd with Kristina McElheran and David Elkington audited 2,241 US companies by submitting a test enquiry to each. Average response time among firms that responded at all was 42 hours. Twenty-three percent never responded. Firms responding within an hour were nearly seven times more likely to have a meaningful conversation with a decision maker than those waiting one more hour.
Section 2How the numbers drift
Three things to know before you put any of this in a board deck.
- The primary data is old. The 2007 study predates live chat, calendar-booking links and most modern marketing automation. It describes web-form leads calling a B2B inside sales team in the mid-2000s. It says nothing directly about product-led signups, partner-sourced leads, or enterprise ABM.
- It measured call attempts. If your first touch is an email, the clock still matters, but the mechanism (catching a buyer while they are still at their desk thinking about the problem) is weaker. Where you can call, call.
- Speed helps contact and qualification, not closing. Nothing here shows that responding faster wins deals with buyers who were never a fit. Speed removes a reason to lose. It does not create a reason to win.
Claims like "78% of buyers choose the first vendor to respond" circulate widely and trace back to vendor-published material without full methodology. Directional at best.
Section 3Vendor data vs independent research
| Source | Type | Use it for | Caveat |
|---|---|---|---|
| Oldroyd, Lead Response Management Study (2007) | Academic, MIT Sloan, with a vendor partner | The shape of the decay curve | Nearly two decades old; inside sales, web-form context; the vendor partner sells response software |
| Oldroyd, McElheran, Elkington, HBR (2011) | Peer-reviewed publication, audit design | How slow the average company actually is | US firms, 2011; response norms have shifted since |
| InsideSales.com / XANT whitepapers | Vendor | Talking points | Methodology not public; company sells the remedy |
| Routing vendor blogs (LeanData, Chili Piper and similar) | Vendor | Implementation patterns and tool comparisons | Selection bias toward their own customers |
| Conversational marketing surveys | Vendor survey | Chat response norms | Self-reported; chat-specific, not phone or email SLA data |
A workable rule: if a number does not come with an author, a year and a described method, it does not go in the deck.
Section 4Where SLAs break in practice
These patterns are reported consistently across implementation write-ups and practitioner accounts. They are not statistically proven; they are the failure modes you should expect.
- Assignment and response are measured in different systems. The CRM enforces "assign within a minute" while the rep's obligation to actually call is tracked in a spreadsheet or nowhere. Assignment speed gets optimised; response speed doesn't move.
- Round robin ignores capacity and availability. Leads go to reps on leave or already sitting on forty open records. The clock starts on a lead that will realistically wait hours.
- Breach alerts without reassignment. Most setups notify on breach; few reassign automatically. Within a week the notification is noise.
- Working hours are not modelled. A five-minute SLA quietly becomes five minutes during headquarters hours, which is a fourteen-hour wait for a buyer in another region. Decide whether you are measuring elapsed or covered minutes, and report both.
- Definition drift. "Response" means attempted call to one team and connected call to another. Compliance numbers get reported before anyone agrees which it is.
- No holdout for existing customers. An expansion enquiry from a current customer lands in the new-business pool because the form said "1 to 50 employees".
Section 5A starting SLA
Rather than borrowing a vendor number, start here and tune it against your own data.
| Step | Target | Notes |
|---|---|---|
| Assignment | Under 2 minutes, system-enforced | Event-triggered, never batched. Move slow enrichment after assignment. |
| First touch, high intent | Under 5 to 15 minutes in the queue's working hours | Demo and contact-sales requests, named accounts, enterprise. |
| First touch, volume queues | Under 30 minutes | Set a target you can staff. An unmet 5-minute SLA is worse than a met 15-minute one. |
| Out of hours | Automated acknowledgement, clock resumes at open | Report covered and elapsed minutes separately so weekend waiting stays visible. |
| Escalation | Alert at the SLA, reassign at twice it | Reassign, don't just notify. Add a cooldown so leads don't ping-pong. |
| Reporting | Attempt and connect tracked separately | Bucket leads by time to first touch and compare qualification rates. That is your own version of the curve. |
After four to six weeks you will have enough data to see where your own contact rate falls off, which is a better basis for the target than anything in a 2007 study.
Section 6References
- Oldroyd, J. (2007). Lead Response Management Study. MIT Sloan with InsideSales.com. Approximately 15,000 leads and 100,000+ call attempts across six companies.
- Oldroyd, J., McElheran, K., Elkington, D. (2011). "The Short Life of Online Sales Leads." Harvard Business Review, March 2011. Audit of 2,241 US firms.
- Vendor whitepapers and implementation blogs, reviewed for failure patterns and tooling options rather than benchmark figures.