Salesforce Enterprise License Agreement (SELA). What it costs, what to strike.

Sales Cloud Enterprise lists at $175 per user per month. Unlimited lists at $350. A SELA bundles the clouds you run, and some you do not, into one three year envelope priced off those two numbers, then adds a true forward clause that converts every seat and every conversation you add into the floor you renew against. The discount is real. What funds it is almost always the scope.

How does a Salesforce SELA actually work?

A SELA bundles a customer Salesforce clouds into one three year contract anchored to the Salesforce Master Subscription Agreement. It replaces a stack of individual cloud contracts with one commercial envelope, one renewal date, and one negotiated discount.

The trade is a larger discount in exchange for a multi cloud commitment, a fixed term, and a true forward clause. The discount is real. The cost of the commitment usually exceeds it.

Why does the bundled discount feel bigger than it is?

Salesforce account teams quote the SELA discount against the full list price of every included cloud, including the clouds the customer was not buying. The headline percentage is therefore measured against a denominator that the customer would never have paid.

The real comparator is the discount against the clouds the customer would have bought standalone. Run that math first. The bundled number usually shrinks by half.

Run it with numbers. Say the proposed bundle lists at $12,000,000 a year: $7,000,000 of Sales Cloud, $3,000,000 of Service Cloud, and $2,000,000 of Marketing Cloud Engagement you had no plan to deploy. Salesforce offers 35 percent off the bundle. That is $12,000,000 × 0.65 = $7,800,000 a year net. Now price only what you were going to buy: $10,000,000 of Sales and Service at the 22 percent you would have been offered standalone is $10,000,000 × 0.78 = $7,800,000 a year. The same money. The extra thirteen points of headline discount bought a $2,000,000 line you did not want, and that line sits in the baseline for the whole term and escalates with everything else.

Make the account team quote both, side by side, on the same page, before anyone says a percentage out loud. The bundle is only cheaper if the standalone column is higher.

What does a Salesforce SELA cost before the discount?

Salesforce publishes edition pricing, so start there and not at the discount percentage. The figures below are list, pre discount, per user per month, billed annually, as published on the Sales Cloud and Service Cloud pricing pages. Every SELA in the market is built off this column.

These rose about 6 percent on 1 August 2025. Enterprise went from $165 to $175 and Unlimited from $330 to $350. If your current SELA was papered before that date, the renewal is being modeled on the new list, and the discount percentage you held last time will not reproduce the price you held last time. Ask for the renewal quote to show list, discount percentage and net price on every line, and ask for it against both the old list and the new one. A renewal that holds your discount and still costs 6 percent more is not a renewal that held.

Two prices are not on this table on purpose. The Agentforce 1 editions sit above Unlimited and bundle Data Cloud and Agentforce allowances into the seat; that tier has been renamed and repriced since 2024, so read its rate off the pricing page on the day you build the model rather than off an article. Salesforce Shield, which is Platform Encryption, Event Monitoring and Field Audit Trail, has no published rate at all. It is quoted as a percentage of net subscription spend and negotiated deal by deal. Get that percentage written into the order form before Shield is folded into a bundle line, because once it is inside the bundle it scales with every seat you add afterwards.

Worked example: twelve thousand seats, two edition mixes

Same estate, same discount, one difference. Twelve thousand Sales Cloud seats on a three year term.

What Salesforce proposed. Unlimited on all twelve thousand.
12,000 × $350 × 12 = $50,400,000 a year at list.

What the edition audit supported. Four thousand seats with a documented need for the higher full sandbox count and API ceilings. Eight thousand without one.
4,000 × $350 × 12 = $16,800,000
8,000 × $175 × 12 = $16,800,000
Total: $33,600,000 a year at list.

The distance between those two order forms is $16,800,000 a year at list, $50,400,000 across the term, before anyone has argued about a discount.

Now apply the same 30 percent SELA discount to both, because the discount is struck against list and the mix change survives it:

$105,840,000 against $70,560,000. A third of the contract removed by an edition audit and nothing else. Not one point of extra discount was negotiated to get it, which is why the edition audit runs before the discount conversation and not after.

Substitute your own numbers. Seats with a documented Unlimited requirement × $4,200, plus everyone else × $2,100, times three years, times one minus your discount. Then put that number next to the one on the order form.

What scope should a Salesforce SELA cover?

Scope should match the actual edition mix and user population, not the vendor projection. A SELA priced on Unlimited when most users fit Enterprise is overpaid from day one. The right scope is the one a senior buyer would defend in front of an internal finance review.

• Edition mix. Map every user segment to the lowest edition that meets the documented requirement.
• Cloud coverage. Carve out clouds with low deployment certainty. Add them back on the next renewal cycle if needed.
• User population. Use the actual licensed and active count, not the headcount projection.
• Data Cloud credit pool. Size to current consumption, not the publisher growth model.
• Agentforce flex consumption. Pilot before committing to a full year envelope.

How wide is the gap between Enterprise and Unlimited on real estates?

Enterprise covers the bulk of standard sales and service work. Unlimited adds the higher full sandbox count, the higher API call ceilings and Premier Success. In the estates we have audited, Unlimited earns its $2,100 a year premium on 25 to 40 percent of the base, and the population that earns it is usually the platform, integration and release teams rather than a business unit.

Do not argue this from opinion. Two queries settle it, and both run in the developer console against your own org.

Entitled against consumed, by license type:
SELECT Name, TotalLicenses, UsedLicenses FROM UserLicense

Active seats that have not logged in for a quarter:
SELECT Id, Name, Profile.Name, LastLoginDate FROM User WHERE IsActive = true AND LastLoginDate < LAST_N_DAYS:90

The same two numbers are visible without writing SOQL at Setup → Company Settings → Company Information, in the User Licenses and Permission Set Licenses related lists. Setup → Identity → Login History holds six months, which is enough to build a quarter on quarter picture and not enough to build a three year one. Start pulling it twelve months before the renewal, not one month before, or the evidence window closes on you.

Take that output into the room. An account team concedes an edition line against a login report far faster than against an argument about functionality, because the login report is their own product telling the story.

How do Agentforce and Data Cloud consumption behave inside a SELA?

Seats are not what breaks a 2026 SELA. Consumption is. Agentforce does not bill per user, it bills per conversation, at close to $2.00 on the standard published meter. A conversation is a bounded session, not a message, so one session that runs twenty turns bills once. That sounds cheap until it is multiplied by a contact center: 50,000 to 200,000 deflected tier one conversations a year is $100,000 to $400,000 at that rate.

Data Cloud runs on credits drawn from an annual pool. The draw is visible in Setup → Data Cloud → Digital Wallet, which shows the credit balance and consumption by usage type. Agentforce Flex Credits draw from the same wallet. That page is the only evidence that counts when you size next year's pool. The growth curve on the account team's slide is not evidence, and it is not something you can be held to in reverse.

Neither of these is governed by the Master Subscription Agreement alone. The consumption rules sit in the Product Terms Directory, the per product terms Salesforce publishes separately and points at from the order form. Open the Data Cloud and Agentforce entries in that directory before signature. Most buyers never do, and the definition of a billable conversation lives there rather than in the contract they read.

Why true forward hurts more on consumption than on seats

Seat counts move slowly. Conversation volume does not, which is exactly why a clause that captures peaks is more dangerous here than anywhere else in the contract.

Take a committed 500,000 conversations a year at $2.00, so $1,000,000. Year one actual lands at 900,000 because two agents went live in the third quarter. The 400,000 overage bills at $2.00, which is $800,000, and it bills that way under either clause. Year one costs $1,800,000 whatever the paper says.

Year two is where the clause decides. Assume a 7 percent cap on the unit rate, so $2.00 becomes $2.14, and assume you retire one of those agents so volume falls back to 600,000.

Six hundred and forty two thousand dollars, one year, same usage, same rate. And the $1,926,000 is what the renewal is modeled on, not the $1,284,000 you would have consumed. That is the compounding the clause is for.

Two moves follow from this. Keep consumption lines outside the true forward capture entirely, which is a narrower ask than deleting the clause and lands far more often. And negotiate the rate inside the renewal rather than on a standalone order: at committed volume we have seen the per conversation rate come down 18 to 35 percent that way. That band is observed across engagements, not a published rate, and Salesforce does not publish a discount schedule for it.

Where do Salesforce SELAs typically overpay?

Five overpay patterns recur across the SELA portfolio we benchmark. They are easy to spot once you know what to look for. Address them before signature, not at the next renewal.

1. Unlimited everywhere. The most common overpay. Right size to segments that need it.
2. Data Cloud oversize. Credit pools at 1.5 to 2 times the actual burn pattern. Trim with quarterly true up rights.
3. Service Cloud add ons. Field Service Dispatcher and Technician seats and Service Cloud Voice ride along at full service population when a fraction of the team touches them. Count the dispatchers, not the agents.
4. Marketing Cloud cross subsidy. The Marketing Cloud line item is often used to absorb discount given on Sales Cloud, then escalates harder at renewal.
5. True forward without a cap. Growth captures into the renewal baseline and compounds every year.

The year two trim that never happens

Every SELA gets sold on the same promise, and it sounds reasonable in the room: take the bundle now, take the discount now, and trim the scope in year two once deployment settles. It does not happen. The reason is structural rather than lazy. Salesforce does not credit capacity back at the contract level mid term, so there is nothing to trim into. The order form quantity is a floor and not a ceiling, and the only mechanism that ever moves it downward is the renewal.

In our 2024 to 2025 file the year two trim happened in roughly one deal in five. In the other four, the procurement lead who signed had moved to another category, the deployment evidence nobody pulled at signature was still not pulled, and the first person to look at the numbers again was reading a renewal quote already built on the peak.

So the trim happens at signature or it does not happen. Right size the mix before countersignature, cap true forward at five to eight percent, and buy substitution rights for the editions and clouds you are least sure about. Anything you plan to fix later, price as though you never will, because on the evidence you will not.

“Salesforce quoted us Unlimited across twelve thousand seats. We signed four thousand Unlimited and eight thousand Enterprise. At list, the distance between those two order forms was sixteen point eight million dollars a year.

How should buyers redline a Salesforce SELA?

Six clauses move more value than the headline discount. Treat them as a checklist on every redline cycle.

• True forward to true up. Replace the growth capture with a quarterly true up, capped at five to eight percent annually.
• Cloud substitution rights. Allow movement between clouds at the SKU level within the term.
• Edition substitution rights. Allow downgrade from Unlimited to Enterprise without repapering.
• Data Cloud credit rollover. Carry unused credits within the term, not use or lose.
• Non use relief. Pro rated credit on clouds the customer can show non use against, evidenced from Login History rather than asserted.
• Benchmark clause. Annual right to compare list and discount against the Salesforce reference customer pool.

Which redlines does Salesforce resist hardest?

True forward to true up and cloud substitution rights are the two Salesforce account teams resist most. Trade visibility on quarterly utilization for the substitution rights. Both sides gain predictability; the customer keeps optionality.

The true forward redline, in the words to send back

Ask for the clause by name. On a Salesforce order form it sits in the renewal pricing section, and it usually reads close to this:

Customer's subscription quantities for any renewal term shall be no less than the highest quantity of each Service subscribed to or used during the then current term.

That one sentence is the whole mechanism. A single quarter of growth in month thirty becomes the floor for the next three years, and nothing you stop using ever comes back out. Strike it, and put this in its place:

Customer may add subscription quantities at any time during the Term at the unit prices set out in the Order Form, co terminous with the Term. Quantities added during the Term shall not establish a minimum quantity for any renewal term. The renewal baseline shall be the quantities set out in the Order Form as of its effective date, and Customer may reduce quantities at renewal on sixty (60) days' prior written notice. Unit prices at renewal shall not increase by more than seven percent (7%) over the immediately preceding annual unit price for the same Service.

Three things in there are doing the work. Adds are priced at order form rates, so growth stops being a repricing event. Adds do not set a floor, which is the entire fight. And the seven percent attaches to the unit price rather than the total, so it cannot be inflated by volume you added in good faith. Salesforce will counter on the notice period first because it is the cheapest thing to give; do not trade the second sentence for it.

What to do when Salesforce refuses

They will refuse, at least at first. The account team's number is the renewal baseline, so this is the clause they are least free to give away. Five fallbacks, in the order we work them. Stop at the first one that closes.

1. Cap the capture instead of deleting it. Keep true forward, but cap the quantity it can pull into the baseline at ten percent of the year one quantity per year. The floor can then drift. It cannot double.
2. Hold it to seats. Exclude Data Cloud credits and Agentforce conversations from the capture entirely. This gets signed more often than a full deletion, because seat count is what the account executive is measured on and consumption is not, and consumption is where the arithmetic above shows the damage.
3. Sunset it before the final year. True forward applies in years one and two only. The final year spike is the one that sets the renewal quote, so keeping it out of the baseline is worth more than the first two years combined.
4. Buy the cap with term. They sell term length and they sell close dates. Offer thirty six months instead of twenty four, or an end date that lands inside their fourth quarter, in exchange for the cap. This is the trade that closes most often in our experience.
5. Price the refusal and shorten the term. If none of it moves, model the renewal at peak quantity plus the escalator, put that number in the business case as the real three year cost, and buy the shortest term you can. A twelve month order form at a worse discount is sometimes cheaper than a thirty six month one with an uncapped floor. The arithmetic on this page is how you demonstrate that to a CFO.

Whatever lands, get it onto the order form itself, not into an email from the account executive. The order form and the Product Terms Directory are what survive the next account team reorganization. Nothing else does.

What does the Salesforce SELA renewal cycle look like?

A SELA renewal is a nine to twelve month exercise, not a quarter end activity. The renewal compounds every miss from the original contract. Run the cycle on the customer schedule, not the publisher schedule.

1. Months minus twelve to minus nine. Build the utilization baseline. Document active users, license consumption, and Data Cloud burn.
2. Months minus nine to minus six. Run the edition mix audit. Identify the population that should drop to Enterprise.
3. Months minus six to minus three. Open the alternative supplier conversation. The credible substitute is the strongest renewal lever.
4. Months minus three to signature. Redline and signature. Compressed cycles concede true forward and substitution rights.

What to do next

1. Pull the current SELA scope and the actual deployment data. Compare scope to deployment by cloud, edition, and user segment.
2. Run the edition mix audit. Run the two SOQL queries above, then price the result: Unlimited seats × $4,200 plus Enterprise seats × $2,100. Put that number beside the order form.
3. Map Data Cloud credit consumption. Size the next pool against the burn pattern, not the publisher projection.
4. Document the cloud carve out list. Decide which clouds drop before the next signature.
5. Negotiate true forward to true up. Cap renewal growth at five to eight percent annually.
6. Earn substitution rights. SKU and cloud level substitution protects the term against drift.
7. Engage independent buyer side support. Contact our Salesforce advisory practice for renewal scoping.

Frequently asked questions

What is a Salesforce Enterprise License Agreement?

A Salesforce SELA is a three year contract that bundles a customer Salesforce clouds into one commercial envelope anchored to the Master Subscription Agreement. It replaces a stack of individual cloud contracts with one negotiated discount in exchange for a multi cloud commitment.

How much does a Salesforce SELA typically save against list?

A SELA typically lands between twenty and thirty five percent off list across a three year term. The headline discount tracks the size of the commitment, the user population, and the edition mix. The actual saving is real only if the scope matches the deployment you run.

What scope should a Salesforce SELA cover?

Scope should match the actual edition mix and active user population, not the vendor projection. A SELA priced on Unlimited when most users fit Enterprise is overpaid from day one. Carve out the clouds you do not run and right size Data Cloud credits to real consumption.

What is SELA true forward and why does it matter?

True forward captures annual usage growth into the renewal baseline and applies the escalator on top of it. It prevents downsizing at renewal. It is the most expensive clause in a SELA, which is why a cap or a rewrite to true up is the redline worth spending the most political capital on.

Should we replace true forward with a true up clause?

Yes. True up rights let the customer add usage during the term without capturing it permanently into the renewal baseline. Pair the true up with a five to eight percent annual cap. Salesforce resists this clause hardest; treat winning it as a primary contract objective.

How do Data Cloud credits work inside a SELA?

Data Cloud credits are an annual pool sized against a projected consumption pattern. Salesforce prefers use or lose. Ask for quarterly true up with rollover inside the term. Size the next pool against the Digital Wallet page in Setup, which shows the actual credit draw by usage type, and not against the growth model in the deck.

What is the right time to renegotiate a Salesforce SELA?

Begin the renewal exercise twelve months before contract end. Build the utilization baseline at minus twelve to minus nine months, run the edition mix audit at minus nine to minus six, open the alternative supplier conversation at minus six to minus three, and redline in the final quarter.

How much can a buyer side SELA review save on a three year term?

It depends on the mix, not on the discount. The worked example on this page moves a twelve thousand seat estate from $105,840,000 to $70,560,000 across three years at the same 30 percent discount, a one third cut, from edition mix alone. Across the SELAs in our 2024 to 2025 engagement file the three year saving ran 12 to 25 percent where the estate was already close to right sized, and past 30 percent where Unlimited had been bought across the whole population.

What is the list price of Salesforce Enterprise and Unlimited editions?

Enterprise lists at $175 per user per month and Unlimited at $350, billed annually, pre discount, as published on the Sales Cloud and Service Cloud pricing pages. That is $2,100 and $4,200 per user per year. Both rose about 6 percent on 1 August 2025, from $165 and $330. A SELA discount is struck against those numbers, which is why the edition mix moves more money than the discount does.

How does Agentforce consumption interact with SELA true forward?

Agentforce bills per conversation, near $2.00 on the standard published meter, and conversation volume moves far faster than seat counts. Under true forward the peak volume you touch in any year becomes the floor you commit to in the next one, even after you retire the agent that caused the peak. Exclude consumption lines from the true forward capture, or cap the quantity it can pull in.

What if Salesforce refuses to change the true forward clause?

Work down five fallbacks in order: cap the captured quantity at ten percent of the year one baseline, hold the capture to seats and exclude Data Cloud and Agentforce consumption, sunset the clause before the final year, buy the cap with term length, or price the refusal into the business case and shorten the term. Whatever lands has to sit on the order form, not in an email from the account executive.

This work is run out of the Salesforce advisory practice. Read the related Salesforce pillar hub, the Salesforce renewal pillar, and the Salesforce knowledge hub.