.png)
Frameworks, core principles and top case studies for SaaS pricing, learnt and refined over 28+ years of SaaS-monetization experience.
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.
Volume Pricing vs Flat Discounts Which Pricing Strategy Is Better for Your SaaS Business
Discounts look like a sales tactic. In SaaS, they are part of the pricing architecture. A company can start with exactly the same list price and end up with materially different ARR, expansion economics and sales behaviour depending on whether lower prices are earned through greater volume or granted as a flat percentage across the whole contract. The distinction matters more as SaaS companies span self-serve customers, mid-market teams and enterprise deployments on one rate card.
The market already shows both approaches. As of 21 August 2026, Atlassian uses progressive per-user rates on monthly Jira Cloud subscriptions, HubSpot uses blended rates as marketing-contact volumes rise, and Twilio automatically reduces rates at higher usage volumes. Notion, monday.com and Asana, by contrast, advertise percentage savings for annual rather than monthly payment. These are not equivalent ways of discounting. They reward different customer behaviours.
Monetizely's position is clear: graduated volume pricing is the better default for B2B SaaS when customers expand through seats, contacts, messages, transactions or another measurable unit. Flat discounts should be a narrow, separately governed concession for something the customer gives back, such as annual commitment or cash upfront. They should not be the main way a SaaS business rewards scale.
The discount should follow the quantity the customer controls
Volume pricing starts from a simple economic fact: a SaaS buyer often chooses not only whether to buy, but how much to buy. A 30-seat customer, a 300-seat customer and a 3,000-seat customer create three different commercial opportunities. Treating them all as the same list price less an arbitrary percentage throws away information that is already visible in the transaction.
The strongest version is graduated, or incremental, volume pricing. Seats 1 to 100 might carry one unit price, seats 101 to 250 a lower price, and seats 251 onward a lower price again. Only the units inside each band receive that band's rate.
A flat discount works differently. A 20% discount takes 20% off every unit, including the first units that the customer might have bought at full price. Once approved, the reduction has little connection to whether the account buys 100 seats or 1,000, unless the sales team negotiates the percentage all over again.
Recent B2B research supports treating quantity as part of the pricing problem rather than setting one price and negotiating away from it. In a Marketing Science paper published on 12 May 2026, Soheil Ghili and Russ Yoon estimated price sensitivity across customer sizes for a B2B educational-services company. In their application, an optimised nonlinear tariff increased estimated profit by at least 8.2% over linear pricing. The figure is specific to that business, but the mechanism matters for SaaS: customer size can contain useful information about willingness to pay and should influence the price schedule.
Earlier peer-reviewed work points in the same direction. Iyengar and Jedidi's 2012 Marketing Science study modelled willingness to pay as increasing with quantity at a diminishing rate and found sharply declining marginal willingness to pay in its online rental application. In practical terms, the hundredth unit can still create value while being worth less at the margin than the first. A declining unit-price schedule can accommodate that behaviour without repricing everything already purchased.
The structural difference becomes clearer side by side.
| Dimension | Graduated volume pricing | Flat discount | Monetizely's read |
|---|---|---|---|
| Pricing structure | Unit rate falls in defined quantity bands; lower rates apply to incremental units | One percentage or amount reduces the full eligible contract | Volume makes the concession earned rather than arbitrary |
| Target buyer | Customers whose seats, contacts, messages or transactions can expand materially | Customers with fairly stable quantity where the seller mainly wants a longer term or faster cash | Volume fits expansion; flat fits commitment |
| Packaging | Can operate inside Starter, Pro and Enterprise packages without changing feature fences | Often overlays the whole package and can blur the intended price difference between tiers | Volume preserves package logic better |
| Pricing metric | Discount changes with the underlying measurable metric | Underlying metric may remain seats or usage, but the discount itself does not respond to it | Volume keeps price tied to customer scale |
| Seller economics | Protects the price of earlier units while rewarding additional purchase | Gives the same concession to units that may not require it | Volume protects ARR more effectively |
| Buyer experience | Buyer can forecast how expansion changes average unit price | Very easy to understand at quote level | Flat wins on simplicity, not value capture |
| Governance | Can be encoded into a public rate card, CPQ and billing | Frequently becomes an approval and negotiation decision | Volume is easier to make systematic |
The table points to the central distinction: volume pricing is a price architecture; a flat discount is usually a concession mechanism. One can run predictably for thousands of customers. The other becomes dangerous when every large deal is allowed to redefine it.
Leading SaaS rate cards already separate scale from commitment
The best evidence is visible in current SaaS and cloud pricing. Vendors that have a natural scale metric tend to reduce unit rates as scale rises. Vendors selling relatively simple per-member collaboration plans frequently use a flat annual saving instead.
Atlassian provides one of the cleanest current examples. Its 21 August 2026 cloud licensing page says monthly pricing for products including Jira and Confluence is progressive per user. Its worked Jira Cloud Standard example for 450 seats charges $8.60 per user for seats 1-100, $7.30 for seats 101-250 and $6.10 for seats 251-450, producing a blended average of $7.06 per user. The lower rate does not travel backwards and reprice the first 100 seats.
HubSpot applies the same principle to a different metric. On 21 August 2026, Marketing Hub Enterprise included 10,000 marketing contacts and then listed declining monthly prices per additional 10,000 contacts: $100 from 10,001-50,000, $90 from 50,001-100,000, $80 from 100,001-200,000, $70 from 200,001-500,000 and $60 above 500,000. HubSpot explicitly describes the contact fee as a blended rate.
The comparison across six current B2B products makes the pattern visible.
| Product, pricing checked 21 Aug 2026 | Pricing structure | Target buyer | Packaging | Pricing metric | Discount signal |
|---|---|---|---|---|---|
| Atlassian Jira Cloud | Progressive monthly price bands | Software, IT and business teams scaling user deployment | Standard and higher cloud editions | Billable users | More seats earn lower marginal rates |
| HubSpot Marketing Hub | Blended contact pricing with declining marginal rates | Marketing organisations whose addressable contact base expands | Starter, Professional, Enterprise | Marketing contacts, plus seats | Larger contact volumes receive lower incremental rates |
| Twilio Messaging | Pay as you go with automatic volume discounts | Developers and businesses with variable communications traffic | Messaging APIs and channels | Messages and related usage | Higher monthly volume unlocks discounts; advance volume commitments can save more |
| Notion | Per-member plans plus yearly saving | Plus is positioned for small teams and professionals | Free, Plus, Business, Enterprise | Member per month | Yearly payment advertised at up to 20% saving |
| monday.com | Per-seat plans plus yearly saving | Team-based work-management buyers | Free through Enterprise | Seat per month | Pricing page advertises 18% saving yearly |
| Asana | Team plans with annual versus monthly billing | Teams managing projects and workflows | Personal through Enterprise | User per month | Pricing page advertises up to 18% annual saving |
Twilio makes the distinction unusually explicit. Its current Messaging page says higher monthly message volumes unlock discounts, while choosing monthly volume in advance can produce further savings. Its US SMS page also separates automatic volume tiers from committed-use discounts available for customers prepared to commit to annual message volume.
That distinction is excellent pricing hygiene. Scale earns one kind of economic benefit. Commitment earns another. Mixing both into a generic "20% enterprise discount" makes it impossible to know what the SaaS company received in exchange for the lost revenue.
Graduated tiers preserve revenue before the customer earns the next discount
Consider a SaaS product with a $100-per-seat monthly list price. Suppose management wants to reward larger deployments with four graduated bands: $100 for seats 1-100, $92 for 101-250, $84 for 251-500 and $76 thereafter.
Now compare that architecture with a sales policy that simply allows a 20% flat discount, taking every seat to $80.
The purpose of the model is not to prescribe those particular price points. It is to isolate what happens when the same customer receives lower prices through marginal volume bands rather than a blanket percentage.
| Seats | Graduated volume monthly bill | Effective volume discount vs $100 list | Flat 20% monthly bill | Annual revenue preserved by graduated pricing |
|---|---|---|---|---|
| 100 | $10,000 | 0.0% | $8,000 | $24,000 |
| 250 | $23,800 | 4.8% | $20,000 | $45,600 |
| 500 | $44,800 | 10.4% | $40,000 | $57,600 |
| 1,000 | $82,800 | 17.2% | $80,000 | $33,600 |
At 500 seats, the buyer receives a meaningful 10.4% average reduction from list price under the graduated schedule. The SaaS vendor still retains $57,600 more annual revenue than it would under a blanket 20% cut.
By 1,000 seats, the effective volume discount has risen naturally to 17.2%. The buyer is receiving close to the flat-discount economics, but only after bringing ten times the quantity of the 100-seat account. The concession grows alongside the behaviour the vendor actually wants.
Atlassian's current Jira approach follows this progressive structure, and HubSpot explicitly calculates marketing contacts with blended rates. Both avoid the more problematic "all-units" volume model in which crossing one threshold can suddenly reprice every earlier unit.
A 500-seat prospect may still need 20% off to clear its buying threshold. We would not solve that by declaring flat discounts superior. Repeated evidence that 500-seat customers will only buy at $80 tells us something more important: the list price, segment definition, package or tier boundaries may be wrong.
Monetizely's five steps make volume the default before the rate is set
Monetizely's 5-Step Pricing Framework, developed in Monetizing Agentic AI, treats pricing as a sequence rather than a negotiation exercise. Business goals and customer segments come first because the company must know what it is trying to achieve and which buyers value the product differently. Packaging converts those segments into offers. The pricing metric establishes what customers actually pay on. Rate setting then determines the price points and discount schedule. Operationalization makes those decisions work in quoting, billing, reporting and the customer experience. For the volume-versus-flat question, the order matters: a discount schedule should follow the segment, package and metric rather than compensate for flaws in them.
Applied to this decision, the five steps become concrete:
Goals and Segmentation: decide whether the commercial goal is expansion within accounts, faster acquisition, cash collection or longer commitments, and separate buyer groups whose purchasing behaviour is genuinely different.
Packaging: build offers that give each segment an appropriate product and service level before using price cuts to manufacture differentiation.
Pricing Metric: choose the unit that best tracks how the customer receives value, such as seats, contacts, transactions or messages.
Rate Setting: put the actual unit prices and graduated discount bands around that metric, then reserve other concessions for a defined exchange.
Operationalization: ensure quoting and billing systems calculate the bands correctly and show buyers how expansion changes their bill.
The framework also exposes why broad flat discounting often feels easier than it really is.
The implication is larger than discount design. A well-built volume schedule turns pricing policy into product logic. A flat discount leaves more of the realised price in the hands of individual deals.
Flat discounts are not inherently bad. They become bad when the reason for giving them is vague.
Current self-serve SaaS pricing shows a defensible use. On 21 August 2026, Notion advertised savings of up to 20% for yearly payment, monday.com advertised an 18% yearly saving, and Asana advertised annual savings of up to 18%. Each discount is tied to a visible customer choice between monthly and yearly billing.
The seller is not saying that every larger customer deserves 18% off. It is making a different trade: accept a longer billing commitment and receive a lower effective rate. Twilio goes one step further by separating automatic volume pricing from additional committed-use discounts for customers willing to commit to annual message volumes.
The dangerous version is discretionary flat discounting inside a sales process. Ian Larkin's peer-reviewed study of a leading enterprise software vendor, published in 2014, found that salespeople offered significantly lower prices in quarters when their commission incentives made closing the deal more valuable. The estimated mispricing cost the vendor 6% to 8% of revenue. The paper does not show that every flat discount causes an equivalent loss; it shows why unmanaged deal discretion can move prices according to salesperson incentives rather than customer economics.
Our buyer-fit view therefore has a strong bias toward volume pricing.
Buyer profile Pricing architecture we would choose Named market analogue Why Team expected to grow from 80 to 500 users Graduated volume pricing Atlassian Jira Price falls as deployment expands without discounting the initial seats retroactively Marketing organisation growing from 20,000 to hundreds of thousands of contacts Graduated volume pricing HubSpot Marketing Hub Blended rates let the cost curve follow database scale Communications product with highly variable message traffic Graduated usage-volume pricing Twilio Messaging Usage naturally provides a measurable quantity on which to reward scale Stable 15-person team prepared to pay for a year Flat annual discount can be appropriate Notion, monday.com, Asana Quantity offers limited segmentation value; annual billing provides a clear exchange Large enterprise demanding 25% off because procurement asks every supplier Keep volume pricing primary Atlassian/Twilio pattern A lower rate should be linked to purchased scale; any extra term discount should require a separate commitment High-volume customer willing to guarantee annual consumption Volume pricing as the primary meter, with a separately priced commitment concession Twilio Scale and commitment are economically distinct and should remain visible The exceptions cluster around time and commitment, not customer size. That is the boundary we would preserve.
For SaaS leadership teams, the choice is not really between a complicated pricing model and a simple one. It is between deciding the economics in advance and deciding them deal by deal.
A progressive schedule such as Atlassian's tells a 450-seat buyer what each portion of its deployment costs before an account executive enters the conversation. HubSpot similarly lets the contact count change the marginal rate through published bands. Twilio can apply volume reductions automatically as monthly traffic crosses thresholds.
Flat discretionary discounts invert that governance. The salesperson starts from list price, the customer asks for a concession, management approves a percentage, and the realised price becomes partly a function of timing and negotiating skill. Larkin's enterprise-software evidence shows how expensive that gap between corporate pricing intent and salesperson incentives can become.
We would make five decisions at leadership level:
Adopt graduated volume pricing as the company-wide default wherever account growth has an observable unit. The relevant unit may be users, contacts, transactions, API calls or another established value metric, but larger customers should earn better marginal economics through the rate card rather than through repeated negotiation.
Track every dollar of discount by what the business received in return. Reporting should distinguish price reductions earned through quantity from those exchanged for annual payment, multi-year commitment or another contractual benefit. A generic "discount percentage" hides the commercial reason behind the revenue given up.
Make realised price quality part of sales management, not just pricing-team analysis. A compensation system that rewards bookings while ignoring unnecessary price cuts can create precisely the incentive problem documented in enterprise software sales.
Re-estimate tier boundaries from observed buying behaviour rather than preserving round numbers forever. The 2026 B2B evidence from Ghili and Yoon shows that purchase-size data, including lost opportunities, can reveal how price sensitivity changes with customer size. A SaaS company should use its own won, lost and expansion data to test where the rate curve should bend.
Treat persistent demands for large flat concessions as information about the pricing model. When one segment repeatedly closes only after a 25% reduction, leadership should revisit its intended price level and offer design rather than institutionalise 25% as the unofficial real price.
A SaaS company will still grant flat discounts. Monetizely's position is that they should be rare enough to explain. Volume pricing should carry the job of rewarding scale; flat discounts should have to buy something else. That separation protects revenue, gives customers a transparent path to better unit economics and makes the realised price the product of policy rather than whoever negotiated the final call.
Assumptions
The financial scenario is a Monetizely model, not a quoted vendor rate card. It assumes a $100-per-seat monthly list price, graduated rates of $100 for seats 1-100, $92 for 101-250, $84 for 251-500 and $76 thereafter, versus a 20% flat discount; it excludes tax, foreign exchange, customer ramp, churn, COGS and contract-specific terms. Vendor pricing was checked on 21 August 2026 and can vary by geography, package and billing arrangement. The 8.2% research result is specific to the B2B educational-services application studied by Ghili and Yoon and is not a forecast of the profit uplift an individual SaaS company will achieve.
Footnotes
https://www.amazon.com/Monetizing-Agentic-AI-Handbook-Transformation/dp/B0H7Z13VKJ/
https://www.atlassian.com/licensing/cloud
https://legal.hubspot.com/hubspot-product-and-services-catalog
https://www.twilio.com/en-us/pricing/messaging
https://help.twilio.com/articles/223183328
https://www.twilio.com/en-us/sms/pricing/us
https://www.notion.com/en-gb/pricing
https://monday.com/pricing
https://asana.com/pricing
https://pubsonline.informs.org/doi/10.1287/mksc.2023.0487
https://pubsonline.informs.org/doi/10.1287/mksc.1110.0702
https://www.journals.uchicago.edu/doi/10.1086/673371