How to set up lead routing that actually wins the deal

Routing is the first five minutes of every deal

Most teams treat lead routing as plumbing. Something you set up once, bury in an admin panel, and never look at again. Its the machine that decides whether a lead reaches a human while they still care, or after they have already booked a call with someone else. And the data on that is pretty blunt.

In 2011, a group of researchers published a study in Harvard Business Review with a title that still stings: 'The Short Life of Online Sales Leads.' They audited 2,241 US companies to see how fast each one answered a fresh web lead. The average response time, among the firms that bothered to respond at all, was 42 hours. Almost a quarter of them never replied. Not late. Never.

The same line of research (the MIT and InsideSales 'Lead Response Management' study, run by the same lead author) found that replying within five minutes made you roughly a hundred times likelier to actually reach the person than replying at the thirty minute mark, and twenty-one times likelier to qualify them. Five minutes versus thirty. Not five minutes versus next week.

So before we get into round robin rules and fallback logic, hold onto the why. Routing is not an admin chore. Roughly 78% of buyers end up going with whoever answers first. The routing is the race, and most teams are losing it in the first hour without ever seeing the finish line.

Round robin versus direct assignment

There are two honest ways to hand a lead to a rep, and most arguments about routing are really just people defending one of them.

• Round robin. Spreads new leads evenly across a pool, in turn. Fair, fast, and deliberately blind to who the lead is. It shines when nobody owns the relationship yet and you just need a human on it quickly.
• Direct assignment. Sends the lead to one predetermined rep: by territory, by segment, by named-account list, or simply because they already own the account. Right whenever there is a relationship worth protecting.

In practice you almost never pick one. You layer them: direct assignment for anyone you already know, round robin for everyone you do not. Simple to say. The hard part is teaching the system to tell those two apart at the moment a lead lands, which is the next problem.

What good routing actually looks like

Here is the shape of a routing system that does not drop leads. Read it top to bottom: the lead arrives, the router checks who it is against your CRM, and only then decides between the round robin pool and a specific owner. The dashed rail underneath is the safety net.

That fallback rail at the bottom is the piece most teams forget to build, and the piece that quietly saves the most revenue. We will come back to it.

New lead, returning lead, or somebody else's lead?

This is where naive round robin falls apart. A lead comes in. Is this a brand new company you have never spoken to? Is it someone who downloaded a guide six months ago and went quiet? Or is it already an open deal that belongs to one of your reps, who will be (rightly) furious if it gets reshuffled to a colleague?

Treat all three the same and you get the classic failures: two reps emailing the same buyer, or a warm returning account dumped at the bottom of a queue behind cold traffic. So the system needs to recognise who it is looking at before it assigns anyone. The good news is the answer is almost always already sitting in your CRM. You are just not checking it at the one moment it matters.

• Match on the company, not just the inbox. Tie the lead to an account by email domain, not only the exact address. Otherwise the same buyer using a personal address reads as a stranger.
• Check for an existing owner first. If the account or an open opportunity already has an owner, that person gets the lead. Returning to the owner beats fairness, every time.
• Read lifecycle, not just identity. A known contact raising their hand again is a returning lead. Route them back to their last rep with the prior context attached, not into the cold pool as if you have never met.
• Decide what counts as new, on purpose. Pick a window. No activity in, say, six months and the system can treat them as fresh again. Write that rule down so it is a decision, not an accident.

The tell for a 'personal' lead, one that already belongs to a rep, is simple: there is a name in the owner field, or an open opportunity attached. The tell for a genuinely new one is the absence of both. Your CRM knows this. Make the router ask before it assigns.

How fast is fast enough

Fast. The answer is fast. But 'fast' is exactly how teams talk themselves into 'tomorrow,' so let's put numbers on it.

5 minpeak

30 min21x lower

1 hourthe wall

24 hourscold

100xmore likely to connect at all when you reply in 5 minutes versus 30 (MIT / InsideSales).

7xmore likely to qualify the lead when you reach out within the hour rather than an hour later (HBR, 2011).

Directional, drawn from the MIT / InsideSales Lead Response Management study and "The Short Life of Online Sales Leads," Harvard Business Review, March 2011. The exact multipliers vary by source and industry. The shape of the curve does not.

Speed is not a virtue you can coach into a rep. It is a property of your routing. If a lead sits in a queue for forty minutes because the assigned rep is in a meeting, no amount of hustle fixes that. The system has to move it. Which brings us, finally, to holidays.

What happens when the rep is on holiday

Every routing rule you write quietly assumes the rep on the other end is awake, online, and has room for one more. Reality disagrees. People take holidays, sleep through timezones, and hit their limit on a busy Tuesday. A rule that ignores this does not fail loudly. It fails by handing your best lead of the week to an out-of-office auto-reply.

• Real-time availability. Skip reps who are out, asleep in the wrong timezone, or flagged unavailable. Decide it at the moment of routing, not from a calendar someone forgot to update.
• Capacity caps. Stop assigning to a rep who is already buried. A lead on an overloaded rep is a lead that waits, which is the same as a lead you did not get.
• Named backups. Every rep has a designated backup. Out-of-office reroutes there automatically, with the context, so the buyer never feels the handoff.
• A watched queue, not a black hole. If genuinely nobody is available, the lead lands somewhere a human is actively watching, with an alert. Never in a field nobody reads.

The real test of a routing system is not the happy path. It is the Friday afternoon in August when half the team is on a beach and a six-figure lead fills in the form. If you cannot say exactly where that lead goes, the lead is the one who finds out.

What's out there

You do not have to build this from scratch. The options roughly sort by how much machinery you are willing to run.

• Your CRM's native assignment rules. Free and already there. Fine for plain round robin. They tend to fall down on availability, returning-lead logic, and fallback. Native Salesforce rules will cheerfully hand a lead to someone who left the company in March.
• Calendar-first schedulers. Tools like Chili Piper and RevenueHero are strong when your main motion is 'book a meeting straight off the form.' The routing is built around the calendar, which is great until your motion is not.
• Enterprise orchestration. LeanData is the obvious name. Deep, capable, and scoped and priced for large, complex orgs. If you route millions of records across a dozen business units, this is your tier and it earns its keep.
• Purpose-built routers for growing teams. The middle ground: real round robin, territory, availability, and fallback, running inside the CRM, with no implementation project. This is where our own tool sits, and I will be upfront about that in a moment.

There is no universally correct pick here. There is only the one that matches how your team actually sells and how much overhead you want to carry.

Where LeadRouter fits

Full disclosure: we make one of these. We built LeadRouter because we kept setting up the same routing for clients and watching them either outgrow their CRM's built-in rules or get handed a quote that needed a board meeting to approve. Neither felt right for a team that just wanted leads on the correct rep, fast.

So here is the honest version. LeadRouter is probably a good fit if you recognise yourself in this list:

• You have outgrown your CRM's built-in assignment rules, but you are not an enterprise running a dozen business units.
• You want round robin, territory, real-time availability, and fallback, all inside Salesforce, without a multi-week rollout.
• You would rather start a trial this afternoon than sit through a demo just to find out the price.
• You have already lost a deal because a lead landed on a rep who was on holiday, and you do not want that happen twice.

And it is probably not the right fit if you need full revenue orchestration and AI lead scoring across many teams. That is genuinely what LeanData is for, and we would rather tell you so on a call than sell you the wrong thing. Routing is one of those systems you only notice when it breaks. Set it up properly once, with the returning-lead logic and the fallback rail in place, and it quietly wins you the deals your competitors are still emailing about on Monday.