Field notesField routing

How to plan a sales route: zone, order, then clock

Pick the day's zone, put the stops in an order that never crosses itself, then fit lunches and appointments to the clock with a buffer for bad traffic.

Adam DeLeon · Founder, Adelo CRM
10 min read

Here is how to plan a sales route that holds up in the car: pick the day's zone first, then put the stops in an order that never doubles back, then check the whole thing against the clock, with lunches and appointments placed before anything else. Most advice stops at the second step and hands you the shortest loop. For a rep with a catered lunch at noon and a provider who will only see you at 2:30, the shortest loop is often the wrong day.

Below are the three passes, a worked Tuesday with every number stated so you can redo it with your own, and how much buffer to leave before anything you cannot be late for.

Pass one: pick the zone before you pick the stops

The biggest decision about tomorrow's route was made when your territory was split into days. If Tuesday's offices are spread across two towns, no amount of clever ordering will make Tuesday short. If they sit in a few tight clusters, almost any sensible order works.

So start by asking which part of the map the day belongs to, not which names you want to see. A good weekday zone has three things going for it.

  • It is tight. The stops sit a few minutes apart, so most of the day is calls rather than driving.
  • It holds a day. It has roughly as many offices due this week as one day can carry, not twice that and not half.
  • It is the same zone every week. Offices learn that you are in on Tuesdays, and appointments and lunches start landing on the right day by themselves.

If you have never worked out how many calls a zone can hold, the arithmetic is in ten to twelve calls a day is a consequence, not a target. The short version: the average distance between stops decides your call count more than effort does, and that distance is set by how the zone was drawn.

Then, inside the zone, list who is due: the people your cadence says you owe a visit this week, any lunch or appointment already booked, and a couple of standby names close by in case the day runs early.

Pass two: put the stops in an order that never crosses itself

With the stops chosen, the order is where most wasted miles come from. People building a day from a list tend to write names down in the order they remember the roads, and the result crosses itself two or three times.

With 12 stops and a fixed starting point at home, there are 12 × 11 × 10 and so on down to 1 possible orders, which is 479,001,600. Nobody checks those by eye, and neither does software. Researchers who study routing note that the basic version of this problem, finding the shortest tour through a set of stops, is already one of the hard ones, so in practice it is solved with fast approximations rather than by checking every option. The good news is that a few simple rules get you close.

  1. Work one cluster at a time. Finish everything in one cluster before driving to the next one. Never visit a cluster twice in a day unless a fixed time forces it.
  2. Never cross your own path. If you sketch the day and two legs form an X, swapping the order of the stops between them makes it shorter.
  3. Cross between clusters once. Each crossing is usually the longest leg of the day. Plan the day so you make as few as possible.
  4. Decide where the day ends. Ending near home means the last leg is short when you are tired and the last office runs long. Starting near home gets you to your first call sooner. Which matters more depends on where the early and late time windows sit, which is the next pass.

That gives you the shortest sensible loop. But the loop does not know what time anything happens.

Pass three: fit the route to the clock

A field day has things in it that only happen at one time. A lunch you are providing for an office. An appointment with a provider who sees reps for ten minutes at 2:30. An office that only lets reps in before 10:30, or one that closes for lunch. These are the fixed points, and the route has to bend around them, not the other way round.

The method is to place the fixed points first and then fill the time between them.

  1. Put every fixed time on the day. Lunch, appointments, any office with a narrow window.
  2. Fill each gap from the cluster nearest the next fixed point. The stops before a noon lunch should be the ones near the lunch office. The stops right before a 2:30 appointment should be near that office.
  3. Plan the drive into each fixed point at its bad day length. More on that below.
  4. Keep one standby stop near where the day ends. If an office runs short or someone cancels, you have a call ready that costs one short drive.

A worked Tuesday

Here is a made up day with every input stated, so you can swap in your own.

The zone. Two clusters. North holds seven offices due this week (A to G) plus L, where you are providing lunch for the staff at noon. South holds three offices due (H, I and J) plus P, where a provider has agreed to see you at 2:30.

The inputs. Each call takes 22 minutes, door to door. Stops inside a cluster are 7 minutes apart. The drive between clusters averages 18 minutes. Home is 12 minutes from the south cluster and 25 minutes from the north one. Lunch runs 45 minutes, and you allow 10 more to pack up.

The shortest loop. Sorted for the least driving from home, the day starts at the closer south cluster and finishes in the north. Started at 8:00, it looks like this:

StopShortest loop arrivesFixed time
H, I8:00, 8:29
P8:58appointment at 2:30
J9:27
A to G10:07 to 1:01
L1:30lunch at 12:00

The mileage is fine. The day is not. You reach P five and a half hours before the provider will see you, and you reach the lunch office an hour and a half after the food was supposed to be on the table. Fixing it by hand, by driving back north for lunch and south again for the appointment, adds crossings, which are the longest legs you have.

The route built around the clock. Lunch is in the north and the appointment is in the south, so the morning belongs to the north and the afternoon to the south. You cross once, after lunch.

TimeWhereWhat
7:35HomeLeave, 25 minute drive north
8:00 to 11:16A to GSeven calls, 7 minutes apart
11:23LArrive and set up for lunch
12:00 to 12:45LLunch with the office
12:55LLeave, cross to the south (planned at 25 minutes, see below)
1:20 to 2:11H, ITwo calls
2:18PArrive, 12 minutes early
2:30 to 2:52PThe appointment
2:59 to 3:21JLast planned call
3:33HomeOr one standby call near J first, if the day ran early

That is twelve visits, every fixed time met, and one crossing instead of two. The north cluster is worked before lunch because the lunch is there, and the south cluster is worked around the appointment because the appointment is there. The total planned driving comes to about 132 minutes, close to the shortest loop's, and every minute of it lands where the day needs it.

Notice the order inside each cluster still follows pass two. The clock decided which cluster goes when. The map decided the order within it.

How much buffer to leave before a fixed time

Look at the 12:55 crossing. It averages 18 minutes, but the plan gives it 25. That gap is the difference between arriving on time most days and arriving on time on your bad days, and the Federal Highway Administration has a clean way to think about it.

The FHWA describes travel time reliability as how consistent a trip is from day to day and across times of day. It measures the extra time you need with a buffer index: the gap between a trip's average time and its 95th percentile time (roughly the worst day of the month), as a share of the average. Its own example is a buffer index of 40 percent on a trip that usually takes 20 minutes, which means budgeting 8 extra minutes, 28 in all, to be on time 95 percent of the time. Its report on the subject is blunt about the alternative: if you plan on the average, you will be late about half the time.

Applied to the worked Tuesday:

LegAverageBuffer usedPlanned
Lunch office to the south cluster18 minutes40 percent25 minutes
Any leg inside a cluster7 minutesnone7 minutes

The 40 percent here is the FHWA's example, not a measured number for your roads. Your own figure comes from driving the leg at that hour a few times, or from the traffic estimate for that time of day. Three rules keep it simple.

  • Buffer only the legs that end at a fixed time. If you are a few minutes late to a drop in call, nothing happens. Being late to a provider who gave you a 2:30 slot costs you the slot.
  • Buffer the long legs, not the short ones. A 7 minute hop does not vary much. A 20 minute crossing that touches a main road can vary a lot.
  • Spend the early minutes, do not waste them. On an ordinary day you will arrive at P with time to spare. That is the moment to write up the last two calls, not to sit in the lot waiting.

When the day changes

Something will move. An office cancels, a provider runs late, a lunch gets pushed to tomorrow. A route built in three passes is easy to repair because you know why every stop is where it is.

  • A call drops out. Pull the standby stop nearest to where you are. Do not change the order of the rest.
  • You are running late before a fixed time. Drop the call right before it, not the fixed time. The dropped name moves to the same zone next week.
  • A new appointment appears. Ask for a time that matches the day you are already in that zone. An appointment in the north on south day costs two crossings.

If you would rather not do the three passes by hand each week, Adelo CRM carves your territory into weekday zones, places lunches and appointments on the day so they shape the route, routes the stops on real roads, and counts drive time against the clock so you can see before you leave whether the day fits. For how often each office should come up in the first place, see call frequency by decile.

Questions people ask

How many stops should a sales rep make in a day?

It depends mostly on how far apart the stops are, not on how hard the rep works. In our own worked arithmetic, a tight suburban zone holds about 12 or 13 calls, and a zone where stops average 9 miles apart holds about 9. Work out what your zones hold and plan to that, rather than to a number someone set for the whole team.

Should I start my route at the closest stop or the farthest one?

Start wherever the earliest fixed time is. If nothing is fixed, starting far and working back toward home means the last drive is short at the end of a long day, while starting close gets you to the first call sooner. Either beats a route that crosses back and forth.

How much time should I leave between sales calls?

Inside a tight cluster, plan the actual drive with no extra. Before anything with a fixed time, plan the drive at its bad day length. The FHWA's example adds 40 percent to a 20 minute trip, about 8 minutes, to arrive on time 95 percent of the time.

Is the shortest route always the best sales route?

Only when nothing in the day has a fixed time. Once there is a lunch, an appointment or an office that only sees reps in the morning, the best route is the one that meets those times with the fewest long drives, even if a mileage sort would order it differently.

How do I plan a route around a lunch appointment?

Put the lunch on the day first. Work the cluster around the lunch office in the morning, arrive early enough to set up, and plan the afternoon in the cluster where your next fixed time is. If the lunch is in a different zone from the rest of the day, ask the office for a date when you are already nearby.

Sources

  1. Travel Time Reliability: Making It There On Time, All The Time (brochure), Federal Highway Administration, FHWA HOP 06 070, on the buffer index and the 20 minute, 40 percent example.
  2. Travel Time Reliability: Making It There On Time, All The Time (report), Federal Highway Administration, on why planning on the average travel time makes you late about half the time.
  3. Vehicle routing with time windows: optimization and approximation, M. Desrochers, J.K. Lenstra, M.W.P. Savelsbergh and F. Soumis, CWI report OS R8715, on why routing problems are solved by fast approximation.
  4. Time Window Constrained Routing and Scheduling Problems, Marius M. Solomon and Jacques Desrosiers, Transportation Science, 1988, a survey of routing and scheduling with time windows.

Written by Adam DeLeon, Founder, Adelo CRM. Where a figure appears it is either arithmetic you can redo from the inputs in the text or it is labeled as an assumption. None of it is a customer outcome.

The product these were written alongside.

Every stop planned. Every mile counted. Send us one territory and find out whether that is true.

Map © OpenFreeMap © OpenMapTiles © OpenStreetMap contributors