Platform · Sales incentives
Every point on the SAGARIS board has a receipt
Sales gamification and incentives on one ledger. Twenty two actions score, from a logged note worth one point to a closed deal worth a hundred, and a lost deal worth minus ten. Each one writes a single durable row keyed to the event behind it, so a rep can trace a score back to the thing they actually did.
Included at $499 per seat per month
Twenty two actions that score
minus 10 to 100
Losing a deal is the only action worth negative points, and the catalog shows it as a penalty in plain sight. Your reps can read the whole board before they play on it, so nobody argues about the rules later.
How a score is built
One row per action. Every board reads from it.
A scored action writes a single row carrying the rep, the action, the points and the moment it happened. Lifetime totals, the weekly and monthly boards, the badge wall and the streak all read that one ledger. The number on the leaderboard and the number on a rep's profile always match, so your Monday standup is about selling and not about the scoreboard.
The streak bonus doubles every fifth day
consecutive active days
Only completed milestones count, so a seven day streak is worth what a five day streak is worth. Below five consecutive active days the schedule returns zero rather than a prorated fraction, which is the part a rep needs to know on day four.
Ten badges, each with the moment it was earned
The ledger stores the moment a badge threshold was first crossed, taken from the event that crossed it. An award keeps its real date through any rebuild, so a rep's wall is a record of their year rather than a snapshot of today. Streaks work the same way, and the longest streak a rep has ever held only ever moves up.
- First Blood
- Deal Closer
- Rainmaker
- Pipeline Builder
- Voice of Reason
- Inbox Hero
- Meeting Machine
- Positive Pro
- Marathoner
- Social Seller
Why this is different
The engine tells you why a deal did not qualify.
An incentive only works if the team believes the number. This one is built so they can. The points ledger only ever appends, its idempotency is a unique index in the database rather than a convention in the code, and the bonus engine names the rule that stopped a deal instead of returning nothing. Your reps get a scoreboard they can check, and you get one you never have to defend.
How a closed deal meets a SPIFF
first failure wins
Is the spiff switched on
no: inactiveDid the deal close inside its window
no: outside_windowIs the product one it covers
no: product_mismatchIs the customer segment one it covers
no: segment_mismatchIs the deal above the minimum amount
no: below_min_amountQualifies, and pays one of three ways
- flata fixed bonus
- percent_of_amounta share of deal value
- per_unita rate times units
A deal that misses on two counts is told about the earlier one, because the engine returns the first failure and stops. The rep gets the rule and the manager gets something to answer with, in one line.
Double counting is rejected, not discouraged
Every ledger row is keyed to the event that produced it by a unique index on the workspace, the source type and the source id. Replaying the same activity writes nothing at all. That is a constraint in the database, so it holds even when a job runs twice or a backfill overlaps a live write.
It explains the no
Five named reasons, evaluated in a fixed order, and the first failure is the one a rep is told about. A deal that missed a bonus by one rule comes back with that rule attached. That is the case where an answer actually changes what a rep does next week.
Rolling windows, not convenient ones
The weekly board is the trailing seven days and the monthly board the trailing thirty, measured from the moment you look rather than from a calendar boundary. There is no closed period to reopen, and a rep's longest streak only ever moves up, so a partial rebuild can shorten nobody's record.
The questions buyers actually ask.
- MOST ASKED
The daily scoreboard that sits above payroll. A durable points ledger, streaks, badges, leaderboards, quota pacing and time boxed SPIFF bonuses, all reading off activity your reps already log. Quota attainment is entered by you, so the board reflects the targets your team actually carries. Your commission and payroll systems carry on doing what they do.
Twenty two actions earn points, running from one point for a logged note up to a hundred for a closed deal, and the whole table is published rather than tuned behind the scenes. A referral is forty, a created opportunity thirty, a meeting held twenty five, a booked meeting fifteen, a positive reply twelve. Routine actions carry daily caps, so a day of emailing tops out at fifty scored sends and a day of dialling at forty scored calls, which stops volume alone from carrying a board. The whole table is printed on this page rather than described on a call.
Deal lost is minus ten, and it is the only entry in the catalog flagged as a penalty. We kept it because a board that counts only wins rewards opening opportunities that were never going to close, and the ledger should record the deal that got away as well as the one that landed. It is a small number on purpose: enough that the score reflects the loss, not so much that a rep starts avoiding the hard deals.
Every score traces back to the event that produced it. Each ledger row is keyed to its source by a unique database index. Re-running an award or rebuilding the ledger over the same activity inserts nothing rather than paying twice. That holds even when two jobs overlap. The writer only ever appends, so a score changes by adding an event and the history behind today's total is still there to read.
Three windows read off the same ledger: the trailing seven days, the trailing thirty, and all time. They roll rather than following the calendar, so the weekly board on a Wednesday covers the previous Wednesday onward and there is no Monday reset to plan around. There is also a blind mode, where a rep sees their own row in full and everyone else anonymised, for teams that want the pacing without the public ranking.
They are measured against a ramped target rather than full quota, so a new rep is not judged against a number they were never expected to hit: twenty five percent of quota in the first quarter, then fifty, then seventy five, then the full hundred. Pacing is reported against that ramped number in four bands, ahead, on track, behind and at risk, so a new hire who is doing well shows as doing well instead of showing as a fraction of a quota nobody expected them to carry yet.
A short lived bonus on top of whatever else a rep earns, defined by three things: a window it is live in, a minimum deal amount that qualifies, and an on switch a manager controls. Outside the window it pays nothing, which is the point of running one. It pays one of three ways: a flat amount, a percentage of deal value, or a rate per unit, with an optional cap on any single deal so one large order cannot consume the budget. When a deal does not qualify the engine returns the reason by name, and it returns the first rule that failed rather than a list, so the answer a rep gets is the one thing that would have had to change.
Bring a real week. Watch it score, row by row.
Bring a week of real activity to a demo and watch it score row by row. You will see the points, the streak, where each rep lands on the board, and every bonus that fired or did not fire with the rule that decided it. You will know exactly how your team's week looks on the board before you roll it out.
In this topic
Conversation intelligence and coaching
- Conversation intelligenceEvery call scored on a published rubric of six weighted dimensions.
- The AI coachCoaching on every call, not a sample of them.
- How accurate AI note takers areAccurate to what, measured how, on whose audio.
- Recording in-person sales meetingsThe questions that decide whether you record a meeting in the room.