Sales contest: definition, formats and prize sizing

A sales contest is a time-boxed competition in which reps or teams compete on a defined metric for prizes awarded by rank. It has four required parts: a window with a stated start and end, a single scored metric, published rules including tie-breaks, and a prize structure with a limited number of winners. The limited-winner element is what makes it a contest. A SPIFF pays everyone who clears its trigger; a leaderboard ranks continuously with nothing at stake. A contest concentrates effort into a short window by making position, not just output, worth something.

Contest, SPIFF and leaderboard

These three get used as synonyms and they behave differently.

The cost line matters more than teams expect. A SPIFF paying $200 a unit has no ceiling — a good month costs more than a bad one, which is fine if you budgeted from margin and awkward if you did not. A contest's cost is the prize pot, fixed before it starts. That predictability is one reason contests are easier to get signed off. Full detail on SPIFF structures sits in its own entry.

The main formats

• Sprint. One metric, one to two weeks, ranked prizes. The default. Best for a specific push — pipeline coverage before a quarter, or reactivating dormant accounts.
• Bracket. Head-to-head knockout, reps paired each round, winner advances. Produces genuine drama and keeps every match live because each pairing is close by construction. Needs even numbers or a bye system.
• Team battle. Two or more groups compete on a combined total. Good for cross-functional pairing — AE with SDR, or sales with customer success. Watch for one strong member carrying the group.
• Ladder. Reps hold a position and challenge upward; a win swaps places. Runs for weeks without going stale because the ranking keeps rearranging near the top and the middle.
• Raffle. Every qualifying action earns a ticket, one draw at the end. The odds-based structure keeps a rep in the middle of the field engaged on day nine, because tickets accumulate rather than a rank being locked. Weakest for driving peak output, strongest for participation.

Format choice should follow the problem. If the issue is engagement across the whole team, use a raffle or a ladder. If it is a hard push on one number, use a sprint. If it is a flat team where nothing feels at stake, use brackets. All five run as configured competition formats, and there are worked examples in our sales contest ideas library.

Duration

Short. Most contests are run too long. The engagement curve on a single-metric competition is front-loaded: attention peaks in the first days, then decays as the standings settle and reps who have fallen behind stop checking. A useful rule of thumb is one to two weeks for individual sprints, three to four weeks maximum for team formats where the score moves more slowly, and no more than one contest running per rep at a time. Two simultaneous contests on different metrics do not double effort; they split it and neither gets the attention.

Quarter-long contests are the standard mistake. They collapse into a two-week race at the end, which is what you would have run in the first place at a fraction of the prize cost.

Sizing the prize

Size the pot from incremental margin, not revenue, and show the arithmetic.

Take a team of twelve reps with an average deal of $8,000 at 70% gross margin — $5,600 of margin per deal. The team closes 20 deals in a normal fortnight. You expect a two-week sprint to add three deals. Incremental margin is 3 × $5,600 = $16,800. A working range of 10–20% of that gives a prize pot of roughly $1,700 to $3,400.

Then split it so more than one person can win. A single winner-takes-all prize of $2,500 is worth nothing to the eight reps who know by day four that they are out. The same $2,500 split 50/30/20 — $1,250, $750, $500 — keeps third place live to the end, which is where most of the incremental effort actually comes from. Add a participation threshold rather than a fourth prize: everyone who clears a defined activity floor goes into a draw for something small. That gives the bottom third a reason to keep working the window.

Non-cash prizes generally buy more memory per dollar than cash, because cash blends into pay and a weekend away does not. Blueprint sizes the first competition's pot from your own incremental margin and writes the announcement copy, tie-breaks and rules with it — that pass is described under AI onboarding.

Announcing and closing

Two things decide whether a contest lands. Publish the rules in full before it starts, including how ties break and what disqualifies a scored action, because a rule invented mid-contest destroys trust in every future one. And close it publicly — announce the result in the channel where the team works, with the final standings visible. Slack and Teams celebrations handle the closing beat automatically, and standings stay visible throughout on live leaderboards. You can set the first one up on a 14-day trial with no card required at /start.

FAQ

What is a sales contest?

A sales contest is a time-boxed competition where reps or teams compete on a defined metric for prizes awarded by rank. It requires a stated start and end date, one scored metric, published rules including tie-breaks, and a limited number of winners. The limited-winner structure distinguishes it from a SPIFF, which pays every rep who clears a fixed trigger regardless of position.

What is the difference between a sales contest and a SPIFF?

A contest rewards by rank — only the top finishers are paid, and the cost is capped at the prize pot decided in advance. A SPIFF rewards by threshold — everyone who clears the trigger is paid, so the cost scales with volume and has no ceiling. Contests concentrate short bursts of effort; SPIFFs redirect attention to one specific action, such as selling a named product.

How long should a sales contest run?

One to two weeks for individual sprints, three to four at most for team formats where scores move slowly. Engagement on a single-metric contest is front-loaded and decays as the standings settle, so a quarter-long contest generally collapses into a two-week race at the end anyway. Run one contest per rep at a time — concurrent contests split attention rather than doubling effort.

How much should the prize be worth?

Base it on incremental margin, not revenue. Estimate the extra deals the contest should produce, multiply by gross margin per deal, and budget 10–20% of that figure as a rule of thumb. Then split the pot across at least three places — a rough 50/30/20 split keeps third place competitive to the final day, which is where much of the additional effort comes from.

What stops a sales contest from demotivating the middle of the team?

Three mechanisms. Split the prize across several places so more reps have a reachable target. Add a participation draw with an activity floor, so reps out of the running still have a reason to work the window. And rotate the scored metric between contests, so the same profile of rep does not win every time — a call-volume sprint and a conversion-rate sprint reward different people.