Spiff and Sales Cloud: compensation is not a guess
Salesforce completed the Spiff acquisition on 1 February 2024. Spiff is incentive compensation. The team joined Sales Cloud.
Platform & ArchitectureA sales manager asks for a Spiff data migration quote before the quarter ends. The only artefact is a spreadsheet with names and a column of rates someone typed in the workshop so the demo would show a payout. Nobody signed those rates. The opportunity amount the sheet uses is a formula the forecast does not use. On Monday a seller sees a number in a dashboard and treats it as the commission they will be paid. On Friday payroll runs a different number from the plan document the lawyer holds. The seller has already spent against Monday's figure. The workshop invented a rate. The product was never asked to read the plan. That is the failure, and it is an expensive one, because pay is not a slide.
Salesforce completed the Spiff acquisition on 1 February 2024. Spiff is incentive compensation. The team joined Sales Cloud. The close puts the work next to opportunities, not in a side system you paste into a custom object over a weekend. It does not supply the rates. Your plan document supplies the rates. If the document is missing, the implementation has not started. A guessed percentage is not a placeholder. It is a false promise to a seller.
Incentive compensation, next to Sales Cloud
The close is one sentence: "Salesforce completed its acquisition of Spiff on February 1, 2024." Read it as a date. 1 February 2024. The sentence does not describe your plans, your accelerators, or your exceptions. It does not say a spreadsheet is the plan. It does not authorise anyone to type a rate so a chart looks finished. The team joined Sales Cloud. Use that as a placement. The facts the plan needs, such as amount, date, and owner, are facts Sales Cloud may already store. The plan is the rule that reads those facts. The rule is not included in the acquisition. You still have to bring it.
Write the plan in words a seller and a payroll owner can both apply. Which opportunity counts. When it counts: when it closes, when the invoice is paid, or when a clawback expires. Who is credited when two people worked the deal. What happens when the amount changes after close. If you cannot answer those, you cannot configure a tool, and you should not demo one. A demo with an invented rate teaches the room the wrong number. They will remember the number and forget that it was fake.
Do not migrate "the spreadsheet" as if the columns were a schema. Columns drift. One tab is this year's plan. Another tab is a manager's sketch. A third is a payout someone already received, which is history, not a rule. Label them. Only the signed plan is a candidate for the tool. History might be loaded later, as history, with a mark that it is not recalculated. A sketch is not loaded at all.
A custom object that stores a payout can still be the wrong first build. Sales Cloud may already hold the opportunity the plan reads. Building a shadow object because the workshop needed a place to type a rate duplicates the amount and invites the Monday disagreement. Start from the field the forecast already trusts. If that field is wrong, fix the field. Do not paper over it with a compensation object full of typed numbers.
What this page will not invent
This page will not give you a commission percentage, a quota, an accelerator, or a sample payout. Those figures are not in the acquisition sentence, and a helpful example would be a lie you might paste into a plan. If a consultant offers sample rates "so you can see the screen", refuse the sample. Use a plan you are allowed to show, or use no number and stop the demo. A blank payout is a better demo than a fictional one. The blank tells you the rule is missing. The fiction tells a seller they will be paid.
Do not treat a sandbox copy of opportunities as a copy of the plan. A sandbox may contain deals. It does not contain the agreement you have with your sellers unless you put that agreement there on purpose, and you should be careful what you put in a sandbox that many people can open. Pay data is sensitive. A full sandbox refreshed from production may already hold more of it than the test requires. Say who can see a payout in the test. A shared login for the workshop is the wrong audience.
Do not describe the acquisition as a migration of payroll into Salesforce. Payroll may remain the system that pays. The compensation tool may calculate what should be paid. Those are different duties. A quote that says Salesforce will pay the commission has jumped a boundary the close notice does not cross. Name the system that calculates and the system that pays. If you do not know the second, write unknown. Do not guess.
What the bill is made of
Implementation cost means the parts of the bill. No dollar amount, and no rate card disguised as a plan.
Discovery is the list of plans you will configure, each with the document that authorises it and the person who owns the document. A plan without a document is out of scope until the document exists. The owner is not the partner. The owner is the person who can change the plan when a seller disputes a deal. Put their name in the design.
Fields are the opportunity data the plan reads, and the fields it must not read. Credit, split, and close date are decisions. Write them down. If two fields both claim to be the amount, pick one before anyone calculates. The Monday failure was two amounts. You already know how that ends. Testing uses deals you understand, with the payout checked by the plan owner against the document, by hand, for a handful of deals. If the hand check disagrees with the tool, the tool is wrong or the document was misread. Do not average them.
Exceptions are deals the plan does not cover: a one-off split, a partner deal, a clawback. Each exception needs a person who may approve it. An exception typed by an administrator to make a seller stop asking is how the plan quietly changes. Log the exception. Review it. Crediting is part of the cost, because someone has to confirm the name on the deal is the name who should be paid.
Cutover is the first period you will calculate for real, with the old spreadsheet still available for comparison and not for payment. Hypercare is that period's disputes, handled by the plan owner. Training is the next person in the organisation who can explain a payout without calling the partner. Leave the comparison period out and you will pay from a screen you have not checked.
| Line on the quote | What you are buying | Send it back if |
|---|---|---|
| Plans | Named documents, each with an owner | The source is a workshop spreadsheet |
| Fields | Which amount, date, and credit the plan reads | Two amounts are both in the calculation |
| Rates | Taken from the signed plan only | A sample percentage was typed for the demo |
| Exceptions | Who may approve a one-off | An admin can type a payout to end a dispute |
| Pay versus calculate | Which system pays, if it is not this one | Salesforce is said to run payroll |
| Comparison | A period where the old sheet does not pay | The new number is paid on day one |
The week a seller spends a number you typed
Monday the dashboard shows a payout. The seller screenshots it. Tuesday they ask their manager to confirm. The manager was not in the workshop and does not recognise the rate. Wednesday payroll says the plan document has a different treatment of the same deal. Thursday the seller, who has already committed the money, asks which screen is lying. Friday you discover the rate existed only in the demo spreadsheet. There is nothing to migrate. There is an apology, and there is a rule: no rate enters the tool unless the plan owner can point to the sentence in the document.
Put that rule in the quote. A partner who needs a fictional rate to show progress is showing the wrong progress. Progress is a hand check on a real deal. It is slower. It is the implementation.
The test a buyer should use
Searches ask for the best Spiff Salesforce integration agency, for a consultant to hire, and for an implementation partner. The test is whether they will refuse a made-up rate.
Ask which plan document is in scope. Ask which opportunity fields it reads. Ask who approves an exception. Ask them to run a hand check on one closed deal and to show that the tool matches the document. A partner who offers a sample commission to fill the gap has failed. A partner who says the 1 February 2024 close moved your plan into Sales Cloud automatically has failed. The team joined Sales Cloud. Your plan did not write itself.
Ask where calculation stops and payment starts. If they do not know who pays, they should say so. Do not rank firms, including this one. Compare them on the hand check and on the refusal. A firm that will not refuse a fictional rate will not refuse a fictional payout later.
Website conversations with inbound buyers are a different acquisition. They do not set compensation. Keep them off this quote. The neighbouring page is Qualified, for agentic marketing. A chat on a site is not a plan, and a plan is not a journey.
What you settle before the first calculation
Collect the signed plan. Name the owner. Mark the field that is the amount, the field that is the date, and the rule for credit when two sellers appear. List exceptions you already know. Hide payouts from anyone who does not need them, in the sandbox as well as production. Run the hand check on five deals. If you cannot explain one of them from the document, leave it out of the demo until you can.
Then tell sellers, in writing, that workshop numbers are not payouts. Do it before anyone shares a screenshot. The acquisition date explains when the company joined. It does not explain what a person will be paid. Only the plan does that.
Walk one disputed deal with the plan owner before you widen the audience. Sit with the opportunity open and the document open. Say which field supplied the amount. Say whether a split applied. Say what the tool calculated and what the document says. If you need a sentence that is not in the document, stop. That sentence is a plan change, and a plan change is not an implementation task. It goes back to the owner. Record the five deals you could explain, and the one you could not. The one you could not is the scope gap. A partner who wants to average past it is hiding the gap inside a number. Refuse the average. Pay is not a place for a rounded story.
The quote is ready when a seller can ask why a number appeared and a named owner can answer from the document. If the page still says data migration and names no plan, send it back. The close on 1 February 2024 placed a compensation product with Sales Cloud. It did not invent your rates, and it did not make a spreadsheet into payroll.
The other searches
The project quote is on spiff implementation cost. That page does not print a figure. Who bought the company is on what happened to spiff. The connector is on integrating spiff with Salesforce. Whether to move is on should I migrate from spiff. The customer fee is on spiff licence cost. Who to hire is on hiring for spiff. This page does not rank an agency.



