Hunter and Farmer Sales Roles: What the Split Decides

Hunter and Farmer Sales Roles: What the Split Decides

An account executive carries thirty named accounts and a target with a new-logo component in it. The quarter closes on plan. Every dollar of it came from expanding three accounts that were already customers, and the new-logo line finished at zero. Nobody was lazy and nobody was hiding. Expansion was faster, more certain and cheaper to work, and a seat that owns both numbers will choose the certain one every time it is behind, which is exactly when the choice gets made.

That is the whole argument for splitting the role, and it is an argument about incentives and coverage rather than about temperament. The hunter and farmer vocabulary has drifted into a personality frame, complete with tests that sort sellers into types, and the drift has made the decision harder rather than easier. The question a sales leader is actually facing is which number each seat owns, and how many accounts fit behind it.

What each label actually names

Strip the metaphor and two jobs are left, distinguished by what they start with.

The hunting job starts with no relationship. Its work is deciding which companies are worth a conversation, finding a reason the conversation is worth having this quarter, and earning a first meeting from someone who did not ask for one. The scarce resource is relevance, and the default state of the job is silence.

The farming job starts with a contract. Its work is renewal, adoption, expansion into new teams or products, and being the person the customer calls before they call anyone else. The scarce resource is attention across a portfolio, and the default state of the job is a queue of things that all look urgent.

The reason those cannot be casually combined is not that different people enjoy them. It is that they are measured on different clocks. Expansion revenue lands inside a quarter with a short cycle and a known buyer. New-logo revenue lands after a cycle that is usually longer than the quarter it is worked in. A seat carrying both, judged quarterly, is being asked to fund a long-cycle number out of the same hours as a short-cycle one, and the arithmetic of the scorecard settles it before anyone's preference does.

HuntingStarts with no relationship

• Decides which accounts are worth contacting
• Supplies its own reason and its own timing
• Rejection and silence are the normal condition
• Revenue lands after the quarter it was worked in

FarmingStarts with a contract

• Renewal, adoption, expansion, advocacy
• Timing is supplied by the customer's own calendar
• Most contact is welcome
• Revenue lands inside the quarter it was worked in

Combined seatOwns both numbers

• Both jobs compete for the same hours
• The certain, faster number wins whenever the seat is behind
• New logo becomes the line that slips
• Looks like a motivation problem and is a clock problem

The three coverage models, and what each one decides

A coverage model is the answer to who is responsible for an account at each point in its life. Three shapes account for most of what companies actually run.

One seat lands and keeps. A single seller wins the account and owns it afterwards. Its advantage is continuity: the person who made the promises is the person who has to live with them, which disciplines what gets promised. Its failure mode is the opening scenario, and it arrives roughly when a seller's book of existing accounts becomes large enough to fill the week on its own.

Split at the signature. A hunter wins the logo and hands it to an account manager who owns expansion and renewal. Its advantage is that the new-logo number has an owner who cannot fund it out of expansion. Its failure mode is the handoff, which loses whatever the hunter learned about why the customer bought and what they were promised, and which the customer experiences as being passed to a stranger in month one.

Everyone does everything. The generalist model, which is what a company runs below a certain size whether or not it has named it. Its advantage is that it needs no handoff and no coordination. Its failure mode is that it is the combined seat, and it works only while the installed base is small enough not to compete for the week.

None of the three is correct in general. What decides it is arithmetic rather than philosophy.

The arithmetic that decides which one you need

Four inputs settle it, and they are all numbers a company already has.

How many accounts are worth covering, how much time an existing account genuinely consumes per month, how long a new-logo cycle runs, and what share of next year's target is supposed to come from logos that do not exist yet. That last one is the input most often left implicit, and it is the one that makes the decision obvious once it is written down.

Every figure in the next paragraph is invented for the worked example. None of it is a benchmark and none of it is a RevenueFlow result.

Suppose a seller has capacity for roughly one hundred hours of account-facing work a quarter. Suppose an existing customer consumes two hours a month and a new-logo pursuit consumes fifteen hours across a cycle that runs two quarters. On those invented figures a seller holding twenty existing accounts has already committed one hundred and twenty hours a quarter before touching a new logo, and the new-logo number is not slipping because of effort. It is slipping because it was never in the budget. The moment the installed base per seat crosses the point where it consumes the available hours, the model has already changed whether or not anybody decided to change it.

That threshold, rather than headcount or company stage, is the trigger to watch. It is also why the split so often gets made a year late: nobody is measuring hours consumed by existing accounts, so the first visible symptom is a missed new-logo number, which reads as a performance problem in the seat.

Reading the model off your own inputs

• ~Depends: Hours consumed by existing accounts per seat are measured rather than assumed
• ~Depends: The share of next year's target that must come from new logos is written down
• ~Depends: New-logo cycle length is known and is compared against the quota period
• ✓Yes: If the roles are split, the hunter's quota contains no expansion component
• ✓Yes: The handoff has a named owner, a written trigger and a context document
• ✕No: The hunter is expected to build and enrich their own list as well
• ✕No: Sellers were assigned to a side on the basis of a personality assessment

Where the split starves new logos anyway

Splitting the roles does not by itself protect the new-logo number. Two things undo it quietly.

A hunter quota with expansion in it. If any part of the hunting seat's target can be satisfied by growing an account it already touched, the same clock arithmetic reappears inside the seat that was created to escape it. The split only works if the hunter's number is unmixed.

A hunter who also has to source. A seat that spends half its week building and cleaning a list is not a hunting seat, it is a research seat with a quota attached. This is the most common way a split fails, because sourcing looks like part of hunting until you cost it. The four activities inside outbound work and the order in which they consume a week are set out in outbound SDR and inbound SDR, and the same reasoning applies a level up: the visible part of the job is the smallest part.

There is a third failure that belongs to the farming side. An account manager measured only on retention will not push an expansion conversation that risks the renewal, which is a rational response to the scoreboard and produces a book that never grows. A seam between two functions needs at least one number both sides carry, and the general case is worked through in GTM teams.

Why the personality version does the damage

A large part of what is published under this heading sorts sellers into types, sometimes with an assessment attached. Treat that as an answer to a different question.

Whether a person prefers new-relationship work or portfolio work is real and worth knowing at interview. What it cannot do is scope a seat. A company that assigns people to sides on the basis of a type test has still not decided how many accounts sit behind each seat, what the handoff trigger is, or whether the hunter's quota is clean, and those three decisions are what determine the outcome. The type test then explains the failure in terms of the people, which is the most expensive possible reading.

The practical order is the reverse. Design the seats from the coverage arithmetic, write down what each one owns, and only then ask which of the people you have would rather sit in which. Where the same reasoning shows up along a different axis, outside sales rep versus inside rep is the same argument about travel and coverage per account rather than about desk-versus-field temperament, and territory planning covers the carve itself, including why capacity rather than the map decides who is actually contacted.

What the split does not resolve

Neither model produces conversations. A hunter with a clean quota, protected hours and no sourcing obligation still needs a list of accounts worth contacting and a reason to contact them this quarter, and that is a supply problem sitting upstream of any coverage decision. Companies frequently reorganise the coverage model when the actual constraint was the top of the funnel, and the reorganisation then gets blamed for a result it was never going to change.

The test is the same one that applies to a headcount request. If the accounts worth contacting outnumber the hours available, a coverage change helps. If the hours are there and the conversations are not, the constraint is the list, the message or the infrastructure, and moving people between seats moves the problem without touching it. SDR role definition covers what is left in a prospecting seat once the mechanical part of the work has been automated away, which is the other input into how many hunting hours a company actually has.

The short version

Hunter and farmer name two jobs separated by what they start with: no relationship, or a contract. They are hard to combine because they run on different clocks, and a seat that owns both numbers will fund the long-cycle one out of the short-cycle one whenever it is behind.

Three coverage models exist. One seat lands and keeps, the roles split at the signature, or everybody does everything. Which one fits is decided by four numbers you already have: accounts worth covering, hours an existing account consumes, new-logo cycle length, and the share of next year's target that has to come from logos that do not exist. The trigger for a split is the point where the installed base per seat consumes the available hours, and it is passed well before anyone notices because nobody is measuring it.

A split protects the new-logo number only if the hunting quota carries no expansion component and the hunting seat is not also the sourcing seat. Assign the seats from the arithmetic first and ask about temperament second, because a type test cannot tell you how many accounts fit behind a seat.

If the constraint turns out to be conversations rather than coverage, we run the outbound half and hand back meetings that meet criteria agreed in writing before anything sends.