.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.
How SaaS Teams Can Segment ICPs to Build Better Pricing Tiers
A SaaS pricing page often looks more precise than the thinking behind it. Three or four columns sit neatly beside one another. The cheapest plan is for “small teams”, the middle plan is for “growing businesses”, and Enterprise is for “large organisations”. Yet those labels frequently describe headcount while saying little about why customers buy, what they need to accomplish, or why one buyer should rationally pay more than another.
That shortcut matters because packaging is not merely a merchandising exercise. Research on industrial markets has long shown that company size can affect buying behaviour, but later empirical work found that observable variables such as size and location alone can be poor predictors of how businesses actually buy. Adding differences in organisational strategy improves segmentation because those differences affect buying behaviour. Pricing research reaches a related conclusion: when willingness to pay cannot be observed directly, firms can design distinct packages that encourage buyers to reveal it through self-selection.
Monetizely’s position is firm: SaaS teams should segment ICPs first by the job the customer is buying, then by workflow complexity and organisational risk, with company size used only as a secondary signal. Pricing tiers should turn those differences into self-selecting packages - a lower tier for simple adoption, a middle tier for deeper workflow and collaboration, and an enterprise tier for control, security and service.
Revenue tiers fail when company size stands in for customer need
The sequence matters. Monetizely’s 5-Step Pricing Framework begins with Goals and Segmentation, where the company defines what the pricing system must achieve and which customer groups it intends to serve. Packaging turns those differences into offers. Pricing Metric decides what customers pay on, such as users, transactions or usage. Rate Setting establishes the actual price levels. Operationalization makes the model work in quoting, entitlements, billing, sales compensation and renewal processes. As developed further in Monetizing Agentic AI, the logic is deliberately sequential: weak segmentation corrupts every pricing decision that follows it.
A common mistake is to start at Step Four. Teams debate whether Pro should cost £49 or £59 before establishing whether Pro represents a meaningfully different customer from Starter. No amount of price optimisation repairs a package whose intended buyer cannot explain why the higher tier exists.
There is good economic theory behind that sequence. Moorthy’s 1984 Marketing Science model showed that a product line can be designed so customers select the version intended for their segment. The model also showed why firms cannot ignore cannibalisation between packages: buyers compare the alternatives and choose among them. Belleflamme’s 2005 review of information-product versioning made the commercial implication explicit: where sellers cannot directly observe willingness to pay, they can target packages at different classes of buyers and design the menu so each class chooses its intended package.
The first test for an ICP, therefore, should not be “How many employees does the company have?” It should be “What changes about the problem, value and buying burden when this customer moves from one group to another?”
Four differences usually provide enough information to answer that question.
| Segmentation variable | What teams should ask | What should change in the package | Weak substitute |
|---|---|---|---|
| Job being bought | What outcome is the buyer hiring the product to achieve? | Workflow depth, product scope, automation | Industry label alone |
| Operational complexity | How many people, workflows, integrations or business units must coordinate? | Collaboration, integrations, reporting, higher limits | Employee count |
| Risk and control needs | What happens if access, data or governance fails? | SSO, permissions, audit controls, data residency, sandbox | “Enterprise features” as an arbitrary bundle |
| Buying and service burden | Does the customer need procurement, onboarding, support commitments or contractual review? | SLA, support, implementation and contracting | Sales-assisted versus self-serve as the only split |
The table makes company size useful but subordinate. A 200-person regulated fintech company may have more demanding identity and audit requirements than a 2,000-person media business using the same application for one departmental workflow. Research into industrial segmentation supports precisely that caution: firm size affects some elements of buyer behaviour, yet size and other observable traits do not reliably explain the whole purchase.
The implication for SaaS leaders is uncomfortable. SMB, mid-market and enterprise are often routes to market, not sufficient ICPs. A segment becomes useful for pricing only when its members value materially different things strongly enough to choose, use and pay for a different offer.
The strongest SaaS tiers separate workflow depth from organisational control
Current SaaS pricing pages make the distinction visible. Across collaboration, developer tooling, customer service and marketing software, the premium between lower and upper tiers is rarely justified by “more features” in the abstract. The upper tiers increasingly concentrate permissions, governance, support, analytics and organisational scale.
The evidence is especially clear when we compare six B2B SaaS businesses using their official pricing pages as accessed on 13 August 2026.
| SaaS | Current package signal, 13 Aug 2026 | Meter | What the tier boundary tells us |
|---|---|---|---|
| Notion | Free $0, Plus $10 per licence/month, Business $20, Enterprise contact sales | Licence | Free limits collaborative use; higher tiers expand history, administration and organisational controls. |
| GitLab | Free $0; Premium $29/user/month billed annually; Ultimate custom | User | Premium adds team workflow and support; Ultimate is explicitly positioned for advanced security and compliance requirements. |
| Jira | Free, Standard, Premium, Enterprise; Free capped at 10 users | User | Premium adds advanced planning, admin controls and 99.9% SLA; Enterprise adds multi-site operation, deeper data capabilities and 99.95% SLA. |
| Zendesk | Support Team, Suite Team, Suite Professional, Suite Enterprise + Copilot | Agent | Progression moves from basic ticketing to multichannel support and automation, then advanced operations, then security and governance. |
| HubSpot Content Hub | Free $0; Starter from $7/seat/month; Professional from $450/month; Enterprise from $1,500/month | Seat plus edition, with credits for some usage | The model shifts from basic access to scaled workflows, then advanced control; Professional and Enterprise also include bundles of Core Seats and credits. |
| monday.com | Multiple paid editions; Enterprise quote-based | Seat plus usage allowances | Higher tiers increase automation, integrations and analytics; Enterprise adds large-scale automation, permissions, security, governance and enterprise support. |
The pattern is stronger than any one vendor example. Functional depth separates basic from serious use; organisational risk separates serious use from enterprise use.
GitLab makes that second boundary unusually explicit. Its current Ultimate tier is for organisations with advanced security and compliance requirements, while its GitLab Dedicated offer addresses data residency, isolation and private networking for highly regulated organisations. In August 2025, GitLab CRO Ian Steward described the underlying buyer problem as a choice between cloud efficiency and compliance requirements, then pointed to “data residency and isolation controls” as part of the answer.
Jira follows the same logic. As of August 2026, Free, Standard and Premium can all provide core project workflows, but Enterprise introduces capabilities such as multiple sites, Atlassian Analytics, the Data Lake, multiple identity providers and centralised subscriptions. Premium carries a 99.9% uptime SLA, while Enterprise carries 99.95%. Those are not merely features for companies with more employees. They answer problems created when software becomes a governed organisational system.
The middle tier should monetise deeper work, not basic product viability
Once the segments are defined, SaaS teams face a harder packaging question: which value belongs in which tier?
Our view is that the lower paid tier should let the core job work properly. Weak entry packages often suppress adoption by withholding something customers need to experience the product’s value. The middle tier should then monetise a step-change in how deeply the customer can run its workflow, while Enterprise monetises the cost and value of operating the product safely across an organisation.
A useful three-tier translation looks like this.
| Buyer situation | Entry paid tier | Main tier | Enterprise tier |
|---|---|---|---|
| Job | Complete one important workflow | Run several important workflows | Standardise a business process across the organisation |
| Collaboration | Small team | Cross-functional team | Multiple departments or business units |
| Automation and integrations | Basic | Substantial | High-volume, controlled and extensible |
| Analytics | Operational visibility | Deeper management insight | Cross-team or centralised reporting |
| Security and permissions | Sensible baseline | Additional administration | SSO, SCIM, granular permissions, audit, residency or comparable controls |
| Service | Standard support | Faster or broader support | SLA, enterprise support and possibly named success coverage |
| Price rationale | Product usefulness | More value from the workflow | More value plus lower organisational risk |
The distinction prevents a familiar pricing error: charging the enterprise because it is large rather than because it has different needs. Proper enterprise packaging gives the buyer a reason to pay more even when only a modest number of people require licences.
Public-company commentary reinforces the point. GitLab CFO Brian Robins said in March 2025 that the company was seeing “significant demand from our enterprise customers” for a platform supporting complex software demands. GitLab’s current Ultimate packaging then connects that demand to application security, software supply-chain protection, vulnerability management, governance and compliance.
Atlassian gives us another signal. In its August 2025 shareholder letter, the company reported Jira Service Management Premium and Enterprise edition sales growing more than 50% year on year and Premium plus Enterprise ARR growing more than 40% after AI capabilities were added to those editions. The evidence does not prove that every SaaS vendor should move more value upward. It does show that buyers will upgrade when higher editions contain capabilities tied to more demanding workflows and organisational use.
A package should therefore answer a sentence the buyer can finish: “We need this tier because we now need to…” If the answer is merely “because we have 500 employees”, segmentation has not done enough work.
Self-selection beats a pricing page that requires sales to explain every tier
The best tier architecture does something commercially valuable before a salesperson enters the process: it lets the buyer recognise themselves.
Academic pricing research describes this as self-selection. When willingness to pay is hidden, versioning can separate buyer types by presenting packages designed around different needs. The practical SaaS version is straightforward. A buyer with a simple departmental problem should see no economic reason to buy Enterprise; a buyer with strict audit and identity requirements should quickly discover that the lower package cannot satisfy the purchase.
Self-selection requires more discipline than putting feature counts into columns. Each potential fence should be evaluated against whether it represents genuine segment variation.
| Possible tier fence | Segment signal | Monetizely assessment |
|---|---|---|
| Core workflow capability | Needed by nearly everyone who gets value | Keep low |
| Number of advanced workflows | Value rises as adoption deepens | Strong middle-tier fence |
| Automation volume | Often rises with operational use | Strong middle or usage fence |
| Integrations | Depends on whether integrations signal workflow depth | Selective fence |
| Advanced analytics | More useful as management complexity rises | Strong middle-to-upper fence |
| SSO, SCIM and audit tooling | Driven by governance rather than ordinary product use | Strong enterprise fence |
| Data residency or single tenancy | Driven by regulatory and architectural requirements | Strong enterprise fence |
| Premium support and SLA | Driven by business criticality | Strong enterprise fence |
| Arbitrary feature removal | Does not correspond to a customer difference | Weak fence |
Zendesk provides a clean live example. On 13 August 2026, Support Team is positioned for organisations that have outgrown a shared inbox; Suite Team introduces unified channels and AI automation; Suite Professional addresses more advanced support operations; Suite Enterprise + Copilot is aimed at teams requiring advanced security and governance. The buyer can see an operating journey, not merely a sequence of bigger feature bundles.
Slack’s June 2025 pricing reset followed similar logic. Slack kept Pro pricing unchanged, increased Business+ from $12.50 to $15 per active user per month on annual billing after adding more advanced AI and Salesforce capabilities, and introduced Enterprise+ with enterprise search, stronger administration, governance and compliance. By August 2026, Business+ was still listed at $15 per active user per month annually and included SSO, provisioning, data export capabilities and 24/7 support.
A customer can dispute whether any individual vendor draws every boundary perfectly. The broader lesson is harder to dispute: good fences reflect a change in the customer's operating problem.
ICP segmentation should combine need, scale and risk in that order
Company size should not disappear from the analysis. It should move to its proper place.
A larger organisation often has more users, stakeholders and buying-centre participants. Peer-reviewed research has found that organisation size affects aspects of industrial buying behaviour, while purchase complexity increases the size of the buying centre. The error is treating size as the underlying cause when it is frequently correlated with more important variables such as governance, workflow breadth or purchase complexity.
We recommend scoring prospective ICPs in a sequence that mirrors how value and buying difficulty actually emerge.
| Question | Low score | Medium score | High score | Pricing implication |
|---|---|---|---|---|
| How deep is the job? | Single task | Multi-step workflow | Mission-critical process | Drives product depth |
| How broad is adoption? | Individual or small team | Department | Cross-functional or organisation-wide | Drives collaboration and scale |
| How much control is required? | Standard | Admin-heavy | Regulated or tightly governed | Drives enterprise package |
| How complex is buying? | Card/self-serve | Sales-assisted | Procurement, security, legal | Drives GTM and service model |
| How high is willingness to pay? | Convenience value | Material productivity value | Operational or risk value | Informs rate setting after packages are defined |
The sequence matters. A startup running a mission-critical cybersecurity workflow may score high on job depth and risk even with few employees. A large enterprise using a design tool for five occasional users may score far lower.
The public SaaS market increasingly demonstrates why this richer segmentation matters. HubSpot reported 288,706 customers at the end of 2025 and an average subscription revenue per customer of $11,683 in Q4 2025. CEO Yamini Rangan described the year as one of “clear acceleration upmarket”, saying larger companies were adopting HubSpot to consolidate technology and reduce TCO. Its August 2026 Content Hub pricing simultaneously spans Free, Starter, Professional and Enterprise, with Enterprise starting at $1,500 per month and offering a materially different level of control and flexibility.
monday.com tells a comparable story. In November 2025, co-CEOs Roy Mann and Eran Zinman described their “strategy to move upmarket while expanding our product suite” and said larger customers were adopting multiple products for critical workflows. By February 2026 they again reported progress upmarket as larger customers standardised on monday.com for mission-critical work. Its current Enterprise packaging places advanced analytics, 250,000 monthly automation and integration actions, multi-level permissions, enterprise security and governance, enterprise support and a 99.9% SLA above Pro.
Those companies are not simply chasing bigger logos. Their packaging increasingly recognises that larger, more strategic deployments create different product and operating requirements.
Better tiers start with fewer, sharper ICP decisions
The practical danger is over-segmentation. Once a team realises “SMB versus enterprise” is crude, it can respond by creating twelve personas, seven verticals and dozens of bespoke packages.
Economic research offers a useful warning. Moorthy’s self-selection model found that segmentation does not imply serving every theoretically distinct segment with a separate product; some segments can rationally be aggregated. Information-goods research likewise finds that whether multiple versions are attractive depends on the pattern of customer heterogeneity, rather than on an assumption that more packages are always superior.
Monetizely’s position is to segment finely in the analysis and package coarsely in the market. Five internal customer clusters may still collapse into three commercial tiers if two clusters require essentially the same product, controls and buying motion.
Before creating another tier, require evidence of a genuine break.
Buyers in the proposed segment must value a meaningfully different set or depth of capabilities.
Their willingness to pay must differ enough to justify a separate offer rather than merely a different discount.
They should be able to identify themselves from the package without an extensive sales explanation.
The package must be sufficiently distinct that movement between tiers corresponds to rising value, scale or risk.
Sales, product and billing systems must be able to enforce the boundary without a permanent stream of manual exceptions.
A new tier that fails those tests usually creates complexity rather than revenue.
The final design should be simple enough for the market but rigorous underneath. The Free or entry offer establishes the core job. The main paid tier monetises serious workflow adoption. Enterprise prices the added value and cost of running that workflow across a complex organisation. Other architectures, including modular packaging, are warranted when genuinely different use cases require separate products rather than progressively deeper versions of the same one.
For most horizontal SaaS teams, however, the higher-order decisions are clear:
Choose the market you are prepared not to optimise for. A pricing system becomes coherent only when leadership accepts that the product cannot maximise conversion, ARPU and feature access for every possible customer at once.
Give one executive owner responsibility for the segment-to-package logic. Product, finance, sales and marketing should contribute evidence, but fragmented ownership produces packages built from internal compromises rather than customer differences.
Manage tier mix as a strategic metric. Track which ICPs buy which editions, expand into which packages and churn from which tiers. An unexpected concentration of enterprise-like customers in a low plan is a packaging signal, not simply a sales result.
Revisit segmentation when the product’s role inside the customer changes. A tool that moves from individual productivity to a company-wide system has changed its value, risk and buyer even if its core feature set looks familiar.
Treat the next pricing reset as a segmentation reset first. Do not begin with “How much can we raise prices?” Begin with “Which customer groups are materially different now from the groups we designed these tiers for?”
The central lesson is simple. Pricing tiers work when customers see their own operating reality reflected in them. They fail when internal product history, arbitrary feature fences or employee bands determine the columns.
Segment the problem before segmenting the price. Once SaaS teams know who is buying, what job becomes more valuable, and where organisational risk changes, better tiers become far easier to build.
Assumptions
Current pricing and packaging observations reflect official vendor pages available on 13 August 2026; displayed currencies can vary by vendor geolocation, so cited headline prices use the currency visible in the referenced official source. The three-tier examples and scoring tables above are design models rather than claims about a specific SaaS company’s economics. Rate setting still requires company-specific willingness-to-pay, win-loss, expansion, margin and competitive data after segmentation and packaging have been established.
Footnotes
https://www.amazon.com/Monetizing-Agentic-AI-Handbook-Transformation/dp/B0H7Z13VKJ/
https://pubsonline.informs.org/doi/10.1287/mksc.3.4.288
https://academic.oup.com/cesifo/article-abstract/51/2-3/329/306478
https://www.sciencedirect.com/science/article/pii/0019850181900237
https://www.sciencedirect.com/science/article/pii/S0019850197000643
https://www.sciencedirect.com/science/article/pii/0019850185900318
https://pubsonline.informs.org/doi/abs/10.1287/isre.1070.0119
https://www.atlassian.com/software/jira/pricing
https://www.atlassian.com/blog/company-news/shareholder-letter-q4fy25
https://slack.com/pricing/pro
https://slack.com/pricing/businessplus
https://slack.com/blog/news/june-2025-pricing-and-packaging-announcement
https://www.notion.com/it/pricing
https://www.zendesk.com/pricing/
https://www.hubspot.com/pricing/content
https://ir.hubspot.com/news-releases/news-release-details/hubspot-reports-strong-q4-and-full-year-2025-results
https://about.gitlab.com/de-de/pricing/
https://ir.gitlab.com/news/news-details/2025/GitLab-Reports-Fourth-Quarter-and-Full-Fiscal-Year-2025-Financial-Results/default.aspx
https://ir.gitlab.com/news/news-details/2025/GitLab-Signs-Strategic-Collaboration-Agreement-with-AWS-to-Deliver-Secure-DevSecOps-to-GitLab-Dedicated-Customers/default.aspx
https://monday.com/pricing
https://ir.monday.com/news-and-events/news-releases/news-details/2025/monday-com-Announces-Third-Quarter-2025-Results/default.aspx
https://ir.monday.com/news-and-events/news-releases/news-details/2026/monday-com-Announces-Fourth-Quarter-and-Fiscal-Year-2025-Results/default.aspx