Your target file is not a call plan

Every commercial team selling into physician offices runs the same Monday. A vendor extract goes out to the field: NPI, surname, specialty, practice address, a weighted decile, and some prescribing volumes. Attached to it is a routing template and a number — ten to twelve calls a day, five days a week. What comes back on Friday is a spreadsheet of surnames and cities that somebody built by hand, and that nobody, including the person who built it, can meaningfully check.

The instinct is to treat this as a discipline problem. It is not. A target file and a call plan are different kinds of object, and the distance between them is four transformations that a person cannot reliably do in their head with a spreadsheet open. Naming those four is the useful thing, because it tells you exactly where your process is losing money.

What the file actually is

A target file is a ranking. It says: within this territory, these providers matter more than those providers, and here is the evidence. That is genuinely valuable and it is the expensive part — your data vendor did work you should not repeat. The weighted decile in particular is a considered judgement about a prescriber's weight inside this specific geography, and second-guessing it with your own homebrew score is usually a way of making it worse.

But notice what a ranking is missing. It has no order in time: nothing in the file says who gets seen this week and who gets seen in three weeks. It has no geography beyond a postal address: two providers three miles apart and two providers thirty miles apart look identical in adjacent rows. And it has no notion of a day at all — no idea that a working day contains about seven hours in front of doors and that everything else is windshield.

A ranking tells you who matters. A plan tells you where you will be at 10:40 on Tuesday. The gap between those is where the whole week goes.

Transformation one: score, then tier

The first move is to turn a continuous ranking into a small number of buckets that promise something. A score is a sorting key; a tier is a commitment. "This provider is 87 out of 100" tells a rep nothing about what to do on Tuesday. "This provider is High Potential, and High Potential means weekly" does.

Two things go wrong here reliably. The first is linear scaling of prescription volume. If your top prescriber writes 5,562 scripts and your median writes 341, a linear score makes the top one sixteen times more interesting — and once you normalise, everyone below the ninetieth percentile flattens to approximately zero. You have turned a 539-name territory into a five-name one. Log-scale the volume terms. 5,562 is more interesting than 341, but not sixteen times more, and a rep cannot visit anyone sixteen times.

The second is treating tier membership as purely positional. Some signals should promote a record outright regardless of where it ranks: a practice that has already sampled the product, a physician already prescribing the on-label drug. Those are qualitative facts about the conversation, not another column to weight. A practice that has already tried the thing you sell belongs in the weekly tier even if its score puts it three-hundredth, because the cost of losing it is not proportional to its decile.

One more, quieter, failure: how the file marks an account you already have. In every vendor extract I have worked with, an existing account is encoded by putting the practice name in the NPI column and leaving the surname and street blank. If your import does not recognise that shape, the most valuable records in the territory arrive as unnamed providers with no decile — and the rep, reasonably, ignores them.

Transformation two: carve the map, and pin it to a weekday

This is the one that most teams skip, and it is the one that makes the rest possible. You cannot see a weekly-tier provider weekly if their patch comes round at random. You can see them weekly, easily, if their patch is always Tuesday.

So: split the territory into as many zones as you have field days, and pin each zone to a weekday. Monday is one town, Tuesday is another, and scheduling happens inside a zone rather than across the metro. This is not a clever optimisation. It is the structure that a good rep with a paper map arrives at on their own after two years, and the reason it is worth writing down is that it lets everything downstream be automated.

Two rules about how you carve, both of which I got wrong first:

1. Geography decides which zone; capacity only breaks ties. If you balance zones on workload alone you will produce a Friday that runs two towns twenty-four miles apart on opposite sides of the metro. A day the rep cannot drive is not balanced. It is evenly bad.
2. When no zone has room, the overflow goes to the zone that stays most drivable, not to the emptiest one. Sending it to whichever zone happened to be lightest is exactly how you get a twenty-six-mile Tuesday.

And the corollary in the other direction: pure geographic clustering, with no capacity check at all, produces tidy zones and absurd workloads — a Tuesday carrying twenty cadence commitments against a Friday carrying six. Every weekday has the same number of hours in it, so the zones have to hold roughly the same amount of work. You need both rules, in that priority order.

If your district has already agreed a zone-to-weekday assignment — many routing sheets have one — use it. Re-drawing the map underneath a call plan people have already signed up to is not a feature.

Transformation three: place the calls against the cadence

Now the tiers have to become dates. Weekly means weeks one, two, three and four. Twice-monthly means one and three, or two and four — never one and two, because two visits in consecutive weeks followed by three weeks of silence is not "twice a month" to the person being visited. Monthly means one of the four.

The trap here is trying to balance the load by alternating assignments down the list. Give the first twice-monthly record weeks one and three, the second weeks two and four, and so on. It looks balanced on paper. In practice the tiers do not interleave evenly across weekdays, and you end up with nineteen commitments on one Tuesday and eight on another. Assign each record whichever legal week-combination is currently least loaded for its own weekday, and the problem disappears.

Transformation four: size the day with a clock

The last transformation is the one nobody does, and it is the reason the first three do not stick.

Ten to twelve calls a day is a goal. A day is bounded by time. Take an eight-hour field window, take an hour out for lunch, and you have 420 minutes. A detail — sign in, wait, five to eight minutes with the provider — costs about twenty-two of them. That is 264 minutes for twelve calls, leaving 156 minutes of driving. At twenty-six miles an hour, which is roughly what mid-morning metro driving averages, that is about sixty-seven miles.

So: can your Wednesday zone be driven in sixty-seven miles with twelve stops in it? For a compact suburban cluster, comfortably. For a zone that spans two towns, no — and no amount of asking will change that. That zone fits nine, and a plan that prints twelve on it is a plan the rep abandons in week two, along with everything else you asked them to do.

The number of calls in a day is a consequence of the geography, not an input to it. Treating it as an input is how you end up managing effort instead of managing territory.

Redo that arithmetic with your own numbers. If your detail is fifteen minutes rather than twenty-two, you buy back eighty-four minutes and roughly two more calls. If your zones average eight miles between offices instead of two, you lose four. Both of those are bigger levers than anything you can say in a Monday call.

What this changes about the conversation

When all four transformations exist somewhere you can inspect, a territory review changes character. Today, when a manager asks why a territory is behind cadence, neither person has anything to point at: the plan was in a spreadsheet in an inbox, the week happened in a car, and the two were never comparable objects. The conversation defaults to effort, because effort is the only variable either of them can see.

With a plan that states, per weekday, how many calls the zone actually supports, and a record of which of those days were worked, the conversation becomes: this territory supports 48 calls a week, we planned 48, we worked 39, and here are the nine and why. That is a solvable problem. "Are you working hard enough" is not.

None of this requires new data. Everything above runs on the file you already send out on Monday.