Contents
Key takeawaysThe four leversSettling the countCapping the upliftTrading the termBundles and unlimited dealsThe renewal calendarWhat the renewals showedWhat to do nextFAQSaaS vendors price the renewal on three inputs the buyer controls and rarely checks: the seat count, the annual uplift and the term. Every discount percentage is applied after those three are set, which is why the discount is the last thing worth negotiating.
- The count comes first. Between 15 and 30 percent of ServiceNow fulfiller seats had no active work against them; 60 to 75 percent of Salesforce seats quoted on Unlimited had no documented requirement above Enterprise.
- The uplift is the quiet loss. Uncapped annual increases of 7 to 12 percent compounded through the term; a written cap of 3 percent on a 5 million ServiceNow contract was worth about 600,000 dollars over three years.
- Term is currency. A longer term buys a lower rate only once the count is verified and the uplift is capped; before that it extends both problems.
- Bundles are priced on what you do not use. Salesforce SELA quotes ran 18 to 35 percent above the a la carte cost of the licenses in use.
- Preparation beats the calendar. Workday buyers who opened 150 days out cut the quote by a median 11 percent; the vendor's quarter end added 3 to 7 points on top.
What decides a SaaS negotiation?
Four inputs decide the price, in this order: the seat count, the annual uplift, the term and the structure. The discount percentage is applied to the result. A buyer who negotiates the percentage first is negotiating on the vendor's count, with the vendor's uplift, for the vendor's term.
| Lever | Vendor default | Buyer target | What it was worth in our file |
|---|---|---|---|
| Seat count | Last true up quantity, or the vendor's usage report | Active users as defined in writing, reconciled 90 days before the quote | 15 to 30 percent of fulfiller seats reclaimed (ServiceNow); 60 to 75 percent of Unlimited seats moved to Enterprise (Salesforce) |
| Annual uplift | 7 percent compounding (ServiceNow, Salesforce); 8 to 14 percent on the first quote (Workday) | 0 to 3 percent cap on total order value, written on the order form | About 600,000 dollars over three years on a 5 million ServiceNow contract |
| Term | Three years, no exit, quantities fixed | Term traded only against a verified count and a capped uplift; true down rights at each anniversary | Prevents a 25 to 40 percent over commitment being locked for three years |
| Structure | The bundle, enterprise suite or unlimited agreement | A la carte priced on real use first; bundle only if it beats it | SELA quotes 18 to 35 percent above a la carte cost of licenses in use |
Why the order matters
Each lever sets the base for the next. A 30 percent discount on a count that is 25 percent too high, with a 7 percent uncapped uplift, costs more by year three than a 20 percent discount on the right count with a 3 percent cap. The arithmetic favors the boring work.
What the vendor controls
List price, the quarter end and the approval ladder. None of them are yours to move. The count, the cap and the term are, and they are the inputs the account team hopes you will leave alone.
How do you settle the SaaS seat count before the quote?
Define active in writing, measure it yourself 90 days before the quote, and put the number on the table before the vendor does. The vendor's usage report counts provisioned seats; your count should include only people who worked a record, logged in or ran a workflow in the last 90 days.
Define active
- ServiceNow: a fulfiller who created, updated, resolved or approved a record in the last 90 days; approval only users stay outside the fulfiller count.
- Salesforce: a login plus an action in the cloud being licensed; a Service Cloud login does not justify a Sales Cloud seat.
- Workday: the worker count against the band edges, with contractors and inactive workers reconciled to the contract's definition.
- Microsoft 365 and GenAI seats: feature use, not assignment; assistant licenses bought for every knowledge worker settled at 40 to 60 percent active use inside two quarters.
Edition mix is a count problem
The second count is which edition each person needs. In the Salesforce renewals we reviewed, 60 to 75 percent of seats quoted on Unlimited had no documented requirement above Enterprise, which made edition mix the largest single error in most proposals. Microsoft E5 right sizing recovered 5 to 12 percent of the agreement for the same reason.
Build the count by role, with the edition each role needs, and price the vendor's proposal against it line by line. A proposal that cannot be mapped to your role table is quoting seats you do not have.
How do you cap the SaaS renewal uplift?
Write the cap on the order form as a not to exceed percentage of total order value, before the first proposal. Salesforce and ServiceNow default to 7 percent compounding; Workday first quotes carried 8 to 14 percent. In the Salesforce renewals we reviewed, a price hold moved a 7 to 12 percent uplift into 0 to 4 percent.
| Uplift regime on a 5 million ServiceNow contract | Year 2 | Year 3 | Three year cost against flat |
|---|---|---|---|
| Order form default, 7 percent compounding | plus 350,000 | plus 375,000 | plus 1.05 million |
| Negotiated cap, 3 percent | plus 150,000 | plus 155,000 | plus 455,000 |
| CPI tied cap at about 2.5 percent | plus 125,000 | plus 128,000 | plus 378,000 |
| Flat across the term | 0 | 0 | 0 |
Where the uplift hides
Rarely on a line called uplift. It arrives inside the renewal quote as one annual figure, in a repriced bundle, or as co termed add ons that reset to the master end date and inflate the next true up by 10 to 20 percent. Set the quote beside the expiring order form, line by line, before you read the discount.
Ask for the cap in writing, early
Workday renewals where the buyer challenged the uplift early recovered 60 to 70 percent of it. The same request made in the last two weeks of the vendor's quarter competes with the discount for the one concession the deal desk has left.
Should you sign a longer SaaS term?
Only after the count is verified and the uplift is capped. A longer term buys a lower rate and protection from list increases, and it also locks the quantity: Salesforce's agreement does not allow quantities to be decreased during the term, so a three year signature on an inflated count is a three year loss.
- Verify first. The count and the edition mix are settled and written.
- Cap second. The uplift cap is on the order form.
- Add the rights. True down at each anniversary within a stated band, swap rights for unused entitlements, co termination of every line onto one date.
- Then trade the term. Each extra year is currency the vendor wants; price it as a discount step, not as goodwill.
The true down window
Most SaaS paper allows reductions only at renewal and only inside a notice window, 30 days before expiry at Salesforce, longer at Workday and ServiceNow as written on the order form. Miss it and the term renews on the old count. Put the date in the calendar before anything else.
When does a SaaS bundle beat a la carte pricing?
When the a la carte price of what you actually use is higher, which is rarer than the account team suggests. Across the SELA renewals we benchmarked, the renewal quote ran 18 to 35 percent above the a la carte cost of the licenses in use. The bundle is a financing structure, not a discount guarantee.
| Structure | What the vendor sells it on | What to price it against | Where it did beat a la carte |
|---|---|---|---|
| Salesforce SELA | Unlimited use across clouds at one discount | The a la carte exit on real utilization by cloud | Multi cloud estates growing fast in every cloud; rare at renewal |
| ServiceNow ELA | Platform wide rate above 5 million ACV | The right sized fulfiller base plus named modules | Above 5 million, when it carried an uplift cap, reserved capacity and a 20 percent reduction right |
| Microsoft 365 E5 standard | Suite price beats buying the parts | An E3 and E5 mix by persona | Security led estates with the advanced features enabled; not for users on mail and Teams |
| Adobe ETLA | Flexibility across the creative estate | The corrected mix priced through the VIP program | About 4 of 12 to 16 negotiations; the rest were cheaper as a corrected mix |
Price both, then decide
Build the a la carte model on the verified count first and let the bundle earn its place against it. The buyers who did this made the vendor negotiate against a number it could check; the buyers who started from the bundle negotiated a percentage off an inflated list.
What does the SaaS renewal calendar look like?
Start 150 days out at minimum, twelve months for an estate with several vendors, and run the calendar backward from the notice deadline rather than forward from the vendor's quarter end. The vendor's close is where the deal lands, not where it starts.
| Days before expiry | What is finished | Why |
|---|---|---|
| 365 to 180 | Notice deadline in the calendar; usage definition agreed; role and edition table built. | Every later number depends on the count. |
| 180 to 120 | Count reconciled; a la carte model built; one alternative costed with a budget owner. | The vendor must be able to verify the alternative. |
| 120 to 60 | Clause set sent: uplift cap, true down, swap, co term, price hold; benchmark run on the first quote. | Terms are traded in rounds and need time. |
| 60 to 0 | Signature date held; order form checked line by line; vendor quarter end used to close. | The last two weeks have room for one item. |
Vendor quarter ends that matter
Salesforce and Workday close their year on January 31, ServiceNow on December 31, Microsoft on June 30.
A fourth quarter close with signature authority was worth 3 to 5 points at ServiceNow and about 7 points at Workday in our file, on top of what preparation earned. The full table of dates is in the vendor fiscal year end guide.
What have recent SaaS renewals shown?
The buyers who did well had settled the count, the date and the uplift before the quote arrived, so the discount was applied to a base they had built.
Workday buyers who opened 150 days out cut the opening quote by a median 11 percent. ServiceNow buyers who reconciled fulfiller seats negotiated on a base 15 to 30 percent smaller than the quote.
The discount is applied after the count, the uplift and the term are set. Negotiate those three first and the percentage takes care of itself.
Where the common advice on SaaS discounts is wrong
The common advice is to benchmark the discount percentage and push the vendor to the top of the band. We disagree, as a starting point. Across the Salesforce, ServiceNow and Workday renewals Morten Andersen advised between 2024 and 2026, the percentage rarely moved more than a few points, while the base it was applied to moved by 15 to 40 percent once seats were reconciled and editions corrected, and an uncapped 7 percent uplift returned most of the discount by year three. A top of band percentage on an inflated count with no cap is a worse deal than a median percentage on the right count with a 3 percent cap. Benchmark the rate last, on the base you have proved.
Where the decoder fits
Once the count and the clause set are settled, run the quote through the free VendorBenchmark decoder. It normalizes the proposal against closed agreements at your spend tier and flags the uplift, the term and the structure lines that sit outside the band, which is the checklist for the last round.
Primary sources for this guide
- Salesforce agreements: the Main Services Agreement, including the rule that quantities cannot be decreased during the term.
- ServiceNow pricing: how ServiceNow describes its subscription packaging and quote process.
- Workday legal and agreements: the subscription terms the notice window and uplift live in.
- Microsoft Product Terms: the use rights and true up rules behind Microsoft 365 seat counts.
Source: Redress Compliance advisory engagement file, 2024 to 2026.
What to do next
- Find the notice deadline on the order form and put it in the calendar with a named owner.
- Define active in writing and count it yourself 90 days before the quote.
- Build the role and edition table so the proposal can be mapped seat by seat.
- Send the clause set early: uplift cap, true down, swap, co term and price hold, before the first proposal.
- Price the a la carte exit on real use before you consider the bundle.
- Benchmark the quote last on VendorBenchmark, free and without signup, then bring the result to our negotiation team, fixed fee or 25 percent of what we save.
Frequently asked questions
What should you negotiate first in a SaaS renewal?
The seat count, then the annual uplift, then the term. The discount percentage is applied after those three are set, so negotiating it first means negotiating on the vendor's count with the vendor's uplift.
How much SaaS seat drift should you expect?
A lot. Between 15 and 30 percent of ServiceNow fulfiller seats carried no active work in the renewals we advised, and 60 to 75 percent of Salesforce seats quoted on Unlimited had no documented requirement above Enterprise. GenAI assistant seats settled at 40 to 60 percent active use inside two quarters.
What is a reasonable SaaS renewal uplift cap?
Zero to 3 percent, written as a not to exceed percentage of total order value. Vendors default to 7 percent compounding; a 3 percent cap on a 5 million ServiceNow contract was worth about 600,000 dollars over three years against the default.
Is a longer SaaS term a good idea?
Only after the count is verified and the uplift is capped. Before that it locks an inflated quantity and an uncapped increase for more years. Salesforce's agreement does not allow quantities to be decreased during the term.
Does a SaaS bundle or unlimited agreement save money?
Rarely at renewal. Salesforce SELA quotes ran 18 to 35 percent above the a la carte cost of the licenses in use in the renewals we benchmarked. Price the a la carte exit on real utilization first and let the bundle earn its place against it.
When should SaaS renewal preparation start?
At least 150 days before expiry, and twelve months for an estate with several vendors. Workday buyers who opened 150 days out saved 4 to 9 percent against buyers who reacted inside 60 days and cut the opening quote by a median 11 percent.
Does the vendor's quarter end matter for a SaaS deal?
It closes the deal rather than shapes it. A fourth quarter close with signature authority was worth 3 to 5 points at ServiceNow and about 7 at Workday, on top of what preparation earned. Arrive with an undefended count and the window delivers the standard concession.
How do you benchmark a SaaS quote?
After the count and the clause set are settled, run it through the free VendorBenchmark decoder, which normalizes the proposal against closed agreements at your spend tier, then compare the rate against the vendor benchmark page. Benchmark the right sized base, not the quote as sent.