An in-person SPICED session for Affirm sellers, built around one taught move and one standing post-it exercise. Three exercises are written up as drop-in swaps for the same block, so the spine never changes.
This room is not here to learn a framework. They are here because merchants agree with them and then do not sign. Sixty minutes with a live team buys one selling behavior, not five, and the one worth buying is the move from a merchant nodding along about cart abandonment to a merchant saying a number out loud.
Affirm sellers already know the merchant's world. They know mobile checkout, they know what a high-ticket cart looks like, they know the incumbent. What most of them have not been taught is how to convert that knowledge into the merchant's own arithmetic, which is the only version of the value that survives a finance review.
The exercise is not a break from the teach. The wall the room builds is the teach. In every one of the three options, the room's own post-its cluster in the wrong place, and that visible skew is what Amy debriefs. She never has to assert the lesson, because it is on the wall in the participants' own handwriting.
The one thing that makes standing exercises work: Sharpies, not pens. A post-it written in ballpoint is unreadable from six feet, the gallery walk collapses, and the room sits back down. This is the single most common failure in a physical facilitation block and it is entirely preventable at the materials table.
| Clock | Min | Block | Room is | Compressible |
|---|---|---|---|---|
| 0:00 | 5 | Open. Why you are here, and the one thing we are doing today | Seated | Yes, to 3 |
| 0:05 | 8 | SPICED end to end, one fast pass | Seated | Yes, to 6 |
| 0:13 | 5 | Teach the move. The one element the exercise drills | Seated | No |
| 0:18 | 22 | The exercise. Everybody stands | Standing | No |
| 0:40 | 10 | Debrief at the wall | Standing | No |
| 0:50 | 6 | What good sounds like. Amy models the corrected version | Seated | Yes, to 4 |
| 0:56 | 4 | Commit and close | Seated | Yes, to 2 |
| 1:00 | 60 | End |
Amy talks for 23 of the 60 minutes. The rest is the room working. Clock times are elapsed from session start.
A 60-minute slot is functionally 50. Late arrivals, the laptop that will not project, and one long question at the end are the base rate, not the exception. If Amy has not started the open by minute five, she runs the compressed column throughout and still lands the exercise and the debrief whole.
Where the float lives. This hour has no built-in slack. The designed shock absorbers are the opening and the closing model, which together give back seven minutes on cue. The exercise and the debrief are protected and never get cut, because they are the only two blocks that produce evidence the room can take away. If Amy is running long at 0:40, she cuts the closing model to four minutes and the commit to two. She does not shorten the debrief.
All three fit the same 22-minute block and use the same materials table. Pick one on the diagnosis below, not on preference. Running two is not an option in an hour.
| If the room looks like this | Run this | Because |
|---|---|---|
| They know SPICED already. Their notes are full of merchant context and thin on numbers. Deals stall after a good first call and die in the merchant's finance review. | Pain to Impact ladder wall | The wall makes the thinness visible. Every seller can write cart abandonment; almost nobody can climb it to a number the merchant said out loud. |
| SPICED is new, or half the room learned it and half did not. People use the words loosely and mean different things. | SPICED sort race | Fastest way to install shared vocabulary in a room. The arguments about where a card goes are the learning, and they happen in the first ten minutes. |
| Pipeline is full and slipping. The forecast keeps moving right. Leadership is asking why merchants who said yes in June are still not live. | Critical Event date hunt | The no-date column fills up in front of the people who own those accounts. Nothing Amy says lands harder than that column, and code freeze is the date most of them never asked for. |
If Amy cannot diagnose the room in advance, run the sort race. It is the only one of the three that works with zero prior SPICED exposure, it needs no live deal data from participants, and it is the most forgiving of a room that arrives cold or skeptical.
Every one of the three exercises dies on a room problem that a single email would have caught. Ask the host these four things and get answers in writing.
If the answer on wall space is no, decide before the day, not in the room. The fallback is four flipchart sheets taped into a portable wall, which works and takes ten minutes to build. The other fallback is the sort race on tabletops, which needs no wall at all. Both are fine. Discovering the problem at 9:02 is not.
Set the wall before anyone walks in. Bands and columns taped up, labels written, materials on a side table rather than at seats. A room that walks in past a prepared wall knows within three seconds that it is going to have to get up, and the standing transition at 0:18 stops being a negotiation.
Three things, in this order, and then stop talking.
Do not do a round of introductions. In a 60-minute session, a twelve-person go-around costs eight minutes and buys nothing. If the room genuinely does not know each other, they will meet at the wall in twenty minutes, which is a better introduction anyway.
Walk the framework once. Do not lecture it. Eight minutes, five elements, roughly ninety seconds each.
One framing correction worth making early, because it changes what Situation questions the room asks. At any size above SMB this is usually not a single-vendor replacement. Larger merchants run several providers at once in a waterfall to maximize approval rates, so the deal is often a fight for placement and volume share inside a stack that already exists. "Who do you use today" is the wrong question if it assumes one answer. "Who is in your waterfall today, and how is volume split" is the question that finds the actual opening.
The one anchor to land here. SPICED is a photograph, not a form. Each conversation captures a moment, and it evolves. You do not restart discovery every call, you revisit: here is the situation you described six weeks ago, talk to me, what has changed. This framing is what stops a team from treating SPICED as fields to fill in the CRM.
This block is different for each exercise, and it is the only teaching the exercise depends on. The exact content is written into each exercise option below. Whichever one Amy runs, the five minutes end the same way: she says what the room is about to do, and then the room stands up.
Pick one of the three below. The projectable slide for each is on the exercise page, with the wall diagram and a countdown timer Amy can run from the room screen.
Everyone stays standing. Amy does not return to slides and does not return to the front of the room. She stands at the wall with the group and points at their post-its.
Do not accept a summary. When someone answers the debrief question by describing what their group did, that is a report, not an answer. Ask it again pointing at one specific post-it in their handwriting. Specificity is the whole debrief.
Amy models the corrected version out loud. Not a slide. She takes one weak post-it off the wall, in front of everyone, and talks through the questions that would have turned it into a strong one.
Every person writes one post-it: the name of a real deal, and the one question they will ask on their next call with it. They stick it on the way out the door. Amy photographs the commitment wall.
Everyone writes pain easily. Almost nobody can climb it to a number a customer said out loud. The wall makes that visible in about six minutes.
Impact is what the current state costs, in the merchant's own arithmetic, over a period the merchant chose. It is not a benefit statement and it is not a case study. It has to be a number the merchant says out loud, because a number the seller calculates is a number the merchant's finance team can dismiss.
Affirm sellers have an advantage here that most sellers do not. The merchant already owns every operand. They know their traffic, their conversion rate, their average ticket, and their abandonment rate, and they look at those numbers weekly. The seller's job is asking for them in an order that ends in a total.
Each one produces a single number the merchant already knows, which is what makes this runnable on a first call rather than in a follow-up deck.
Amy should work one example live on the board before the room stands, using round placeholder numbers and saying they are placeholders. Do not put a real Affirm merchant's figures on a whiteboard in a room, and do not present invented figures as though they came from an account. The teaching point is the sequence, not the total.
Say this before the room stands: most sellers stop at rung one, hear the merchant agree that abandonment is a problem, and mistake it for progress. A merchant can agree with every word about their leaking checkout and still do nothing for a year, because agreeing that something is annoying is not the same as knowing what it costs.
"Which number on that top band did a customer actually say out loud, and which one did you calculate for them?"
Expect most greens to be rep-calculated. That is the finding, and it is not a failure. Say so plainly: a rep-calculated number is a starting point, and the work is turning it into a customer-stated one by asking rather than telling. Take three answers. On the third, ask what question they would have had to ask to get the customer to say it.
Fastest way to install shared vocabulary in a room. The cards that land in two different columns are the entire lesson, and they surface in under ten minutes.
Reps already hear all five elements on every call. What they do not do is recognize which is which in the moment, which is why discovery notes end up as a wall of situation with everything else missing. The skill being drilled is recognition speed, not recall.
Say this before the room stands: some of these cards belong in two columns and the team has to argue it out and commit to one. The argument is the exercise. A team that places all sixteen without disagreeing once has not been listening to each other.
"Which card did your team argue about longest, and what were you actually arguing about?"
The answer is almost never the card. It is that pain and impact blur when a cost is implied but not stated, or that a deadline the customer mentioned is not a critical event unless missing it costs something. Those two confusions are worth the whole hour. Take three answers and let one disagreement stay unresolved out loud; a room that watches its facilitator sit with ambiguity learns more than one handed a ruling.
The no-date column fills up in front of the people who own those deals. Nothing the facilitator says lands harder than that column does.
A critical event is a date that already exists in the merchant's world and forces a decision. The seller does not create it. The seller finds it, and helps the merchant see what it costs to miss it.
Pain without a date is a complaint. A merchant can agree with everything a seller says about abandoned carts and still do nothing for a year, because nothing forces the choice.
Make this point to this room specifically. Most sellers in most industries have to hunt for a critical event. Selling into merchants, the dates are lying around, and they are harder dates than almost anyone else gets. Peak does not move. A code freeze does not move. A seller who is not using them is leaving the easiest qualification in the business on the table.
Then teach the subtraction, because this is where sellers lose the deal while feeling urgent. The freeze is a downstream deadline, not the decision date. Scope, contract, integrate, and test runs eight to twelve weeks in front of it. A merchant with an October 15 freeze has a real decision date in July, and a seller still qualifying in September has already lost the season whether or not anyone in the room has said so. The question is not when do you freeze. The question is what has to be signed by when, for you to be live before you freeze.
Say this before the room stands: take-rate objections and "integration is too heavy" objections are usually a missing critical event wearing a disguise. Both resolve to nothing is forcing me to decide right now. Find the date and the objection changes shape.
"For the deals still in the no-date column, what would have to be true in the customer's world for a date to exist at all?"
Two answers come back and both are useful. Sometimes the date exists and the rep has not asked for it, which is a call to make this week. Sometimes there genuinely is no forcing date, which means the deal is not qualified and the forecast is wrong. Naming which of the two applies, deal by deal, is the highest-value thing this room can do in ten minutes.
Sixteen merchant statements, the kind an Affirm seller hears on a first call. Print this page, cut on the boxes, and shuffle. The answer key under each card is for Amy only, so fold it under or print the participant version from the exercise page, which hides the key.
These are composite statements written for training. None of them is a quote from a real Affirm merchant, and no real account figures appear anywhere in this pack.
Have one Affirm seller read these before the day. The cards were written from outside the motion, and a single inauthentic line costs the room's trust in the whole exercise. Send the sixteen to one AE or SE and ask a single question: which of these would you never hear a merchant say. Cut or rewrite whatever they flag. Fifteen minutes of their time is the cheapest insurance on this session, and the ones they rewrite will be better than the ones written for them.
Cards marked in amber below are the deliberately contested ones. They are the debrief. Do not remove them to make the exercise cleaner.
Swap any of these in for a card the room found obvious. Each one exists because it splits a table rather than sorting cleanly, and each reflects a live dynamic in this category rather than generic B2B.
Why these four. The store card seeds the cannibalization objection, which is the most reliable pushback from any retailer running a private-label credit program. The compliance card is realistic rather than paranoid in this regulatory climate. The treasury card is the one that surprises rooms, because it casts finance as an ally rather than the blocker sellers assume, on the grounds that the merchant is paid upfront and does not carry the credit risk. The support card moves the objection out of finance entirely and into operations, where sellers rarely go looking.
Take it further. Once the room can sort these sixteen, have each team write four new cards from their own last week of merchant calls, unlabeled, and hand them to the team on their right to sort. Teams argue about their own accounts rather than a facilitator's examples, and Amy leaves with a deck of real Affirm-language cards that makes the next session sharper than this one. Strip any merchant name before the deck leaves the room.
| What happens | What Amy does |
|---|---|
| The room will not stand up | Amy stands, walks to the wall, and starts writing her own post-it with her back to the room. Silence and a moving facilitator beats a second request. Do not ask twice. |
| No usable wall | Three sheets of flip chart paper on a long table. The exercise works horizontally. What does not survive is a room seated around that table, so the chairs go against the far wall first. |
| Fewer than eight people | Drop the pairs and run the whole thing as one group at the wall. Cut the pair step by three minutes and give that time to the debrief, which gets much richer at small numbers. |
| More than twenty-four people | Two identical walls on opposite sides of the room, and the gallery walk becomes wall to wall. Amy debriefs from the middle. Do not try to run one wall past about twenty-four people. |
| Nobody will name a real merchant | Usually a manager-in-the-room problem, or genuine caution about naming accounts out loud. Switch to the sort race, which needs no live account data at all. This is the reason the sort race is the safe default. |
| Running long at 0:40 | Cut the closing model to four minutes and the commit to two. Never shorten the debrief. The debrief is the only block where the room hears itself think. |
| Nobody is writing | The prompt was too abstract or too exposing. Amy writes the first note herself, out loud, on her own example, and sticks it up. A wall with one note on it is far easier to add to than an empty one. |
| One person is dominating | Move physically next to them and address the group over their shoulder, then ask a direct question to a named quiet person. Do not correct the dominator in front of the room; it costs more than it fixes. |
| Mixed seniority in the sort race | Drop the group-versus-group framing and run it as the whole room against the clock. A VP on the losing team is a dynamic that outlives the session. |
| Running short | Second pass on the wall. Everyone moves one more post-it up a band, or dates one more deal. Do not fill time with more framework. |