SaaS Commercial Transparency Index 2026

Executive Summary

SaaS procurement rarely fails because no price is visible at all. It fails because of the wrong economic unit. A monthly amount does not answer how many Seats are actually needed, how usage and AI add-ons scale, which one-off costs arise, when the contract renews, which notice period applies for an effective termination, and which states are actually released at Exit.

This report therefore combines three checks that have so far often been kept separate:

1. Pricing transparency: Can a buyer follow the direct list price logic without a sales conversation?
2. Renewal and Exit transparency: Are renewal, termination, data access, retention and switching publicly documented?
3. Scaling stress: How does linear per-Seat pricing change the annual cost base at 10, 50 and 200 Seats – and which AI components decouple from it?

The study is not a ranking of product quality. It measures exclusively the public procurement surface as of the data cut-off date. Individual enterprise offers, discounts and contract amendments can differ substantially.

Citable key findings

1. [CALCULATED] The median of the pricing transparency index is 80 out of 100 points; 13 of 20 products reach at least 80 points.
2. [CALCULATED] In the ten-vendor sample, the renewal/Exit module reaches a mean of 79.0 and a median of 80.0 points.
3. [CALCULATED] In the joint eight-vendor sample, the equally weighted Commercial Transparency Score ranges from 72.5 to 88.8 points.
4. [CALCULATED] In the selected plan sample, the publicly normalised annual costs of a 200-Seat scenario range from roughly 23,204 EUR to 249,679 EUR per product and year.
5. [CALCULATED] If all ten selected plans were procured simultaneously for 200 Seats, the pure list price addition would amount to roughly 1,119,286 EUR per year – without implementation, API overages, support, taxes or internal administration.
6. [OBSERVED] The most common public gaps are not in the entry price, but in enterprise plans, overage mechanics, post-termination access, full export scope and historical price changes.
7. [INTERPRETATION] Commercial Transparency is not a synonym for cheap software. It describes how much uncertainty a vendor leaves with the buyer ahead of the sales process.
8. [RECOMMENDATION] A renewal should be treated as a technical Exit project no later than 180 days before renewal, not only as a purchasing exercise.

Joint eight-vendor sample

Rank	Vendor	Pricing transparency	Renewal/Exit transparency	Composite / 100
1	Intercom	90.0	87.5	88.8
2	Airtable	80.0	92.5	86.2
3	Asana	75.0	92.5	83.8
4	HubSpot	80.0	87.5	83.8
5	monday.com	90.0	72.5	81.2
6	Zapier	85.0	70.0	77.5
7	Webflow	90.0	62.5	76.2
8	Notion	80.0	65.0	72.5

[LIMITATION] The composite applies exclusively to the eight vendors that appear in both modules. It must not be equated with the 20-product pricing index or with the SaaS landscape as a whole.

---

1. Research design

1.1 Research question

How completely can a knowledgeable buyer model the direct economic commitment of a SaaS product from public vendor information before signing a contract?

“Economic commitment” in this report covers more than the visible monthly amount. It covers:

plan price
+ billing unit
+ volume and Seat scaling
+ add-ons and AI
+ minimum term and onboarding
+ renewal mechanism
+ termination path
+ data and configuration export
+ post-termination window
+ retention and deletion

1.2 Three separate modules

The modules are not merged into a universal market value. Each answers a question of its own:

Module	Primary question	Sample	Main file
A: Pricing surface	Can the direct list price be modelled?	20 products	data/FDR-2026-01_pricing_transparency_scores.csv
B: Renewal and Exit	Is the path out of the contract publicly traceable?	10 vendors	data/FDR-2026-01_renewal_exit_transparency.csv
C: Seat and AI stress	How does the visible base scale with a larger team?	10 plans	data/FDR-2026-01_per_seat_stress.csv
D: Joint composite	How do vendors that appear in A and B perform?	8 vendors	data/FDR-2026-01_commercial_transparency_composite.csv

1.3 Statement classes

• [OBSERVED]: taken directly from a public primary source.
• [CALCULATED]: reproducibly calculated from published input data.
• [SCENARIO]: hypothetical stress test with disclosed assumptions.
• [INTERPRETATION]: editorial classification.
• [RECOMMENDATION]: operational recommendation for action.
• [LIMITATION]: explicit limitation of the statement.

1.4 What the report deliberately does not measure

• product quality and feature scope
• negotiated discounts
• individual DPA, MSA or order form conditions
• actual implementation and migration costs
• security, data protection or availability
• customer satisfaction
• complete TCO
• legal effectiveness of specific terminations
• import capability of a target platform

---

2. Synthesis: where procurement uncertainty arises

2.1 The entry price is only the first layer

Most of the products examined make the entry level visible. This creates the impression of complete pricing transparency. The relevant uncertainty, however, begins the moment a company no longer matches the marketing example:

• more Seats than in the base scenario,
• several workspaces or business units,
• higher data usage,
• API or automation volume,
• governance and security functions,
• AI Credits or copilot add-ons,
• premium support,
• annual minimum commitment,
• region- or currency-dependent price lists.

[INTERPRETATION] The visible price is often a qualifying figure: it shows that the product is purchasable in principle. It is not necessarily the figure a growing company can budget with.

2.2 Renewal is an architecture event

A contract renews on a date. A system, however, can only be left in a controlled way if several clocks are taken into account at the same time:

Clock	Critical question
Contract clock	When does the contract renew, and by when must notice be given?
Data clock	How long does a complete, verified export take?
Logic clock	Which workflows, permissions, rules and integrations have to be reconstructed?
Operations clock	How long do the legacy and target systems have to run in parallel?
Deletion clock	When are production data, backups and remaining copies deleted?

The latest economically safe Exit date therefore regularly falls before the formal notice deadline.

2.3 Per-Seat pricing creates linear costs – but not linear usage

A Seat is contractually binary: booked or not booked. Actual usage is continuous:

• daily active power users,
• occasional users,
• seasonal users,
• service accounts,
• external partners,
• managers with read-only access,
• employees in onboarding or offboarding.

A 200-Seat contract can therefore contain 200 paid accounts but considerably fewer productive users. The economically meaningful denominator is not just “booked Seats”, but for example:

cost per weekly active user
cost per completed transaction
cost per revenue event generated
cost per productive hour

2.4 AI makes the cost function multidimensional

AI can be priced as a Seat add-on, a Credit system, usage-dependent consumption, an included allowance or a separate product. This creates a second scale alongside the human user:

Seat costs
+ AI add-on per Seat
+ Credits/tokens/actions
+ retrieval and external APIs
+ review and correction work

[RECOMMENDATION] Procurement should not record AI as a feature checkbox, but as a billing class of its own with a metric, a budget limit, an overage rule and a shutdown path.

---

Part I - Public pricing and meter transparency

I.1 Research question

How completely can companies trace the direct list price costs of a SaaS product from publicly available vendor information?

The question is deliberately narrower than “What does SaaS cost?”. Real total costs additionally depend on discount, contract, implementation, migration, training, administration, integrations, outage risk, data volume, support and Exit. This index measures only the publicly documented pricing surface.

---

I.2 What the index measures – and what it does not

What is measured

• Visibility of a genuinely usable entry price
• Comprehensibility of the billing unit
• Public pricing of higher plans
• Documentation of usage, overage or Credits
• References to minimum quantities, contract commitment and onboarding
• Visibility of paid add-ons, in particular AI
• Availability of a calculator or interactive selector
• Traceability of price changes or versions

What is not measured

• Product quality
• Feature scope
• Price/performance ratio
• Negotiated enterprise discounts
• TCO
• Customer service
• Data protection or security
• Portability
• Vendor Lock-in
• Legality of the pricing model
• Individual negotiation outcomes

A product can be expensive and transparent. It can be cheap and opaque. The score makes no statement about this.

---

I.3 Sample

The sample was formed purposively, not randomly. It includes 20 products that represent different cost logics in typical digital operating models:

• per Task, Credit, Execution or flow
• per Seat or agent
• per Monthly Tracked User
• per Monthly Active Row
• per host, data volume or compute
• per site, workspace or user role
• hybrid Seat-plus-usage models

The selection allows a cross-section of relevant billing mechanisms. It is not representative of the SaaS market as a whole.

---

I.4 Methodology

4.1 Assessment dimensions

Code	Criterion	Weight	Full score if …
B1	Public entry price	10	a genuinely bookable entry plan with a price is visible
B2	Definition of the billing unit	15	Seat, Task, Credit, host, MTU, MAR or another unit is clearly explained
B3	Public prices of higher plans	15	growth remains modellable without a mandatory sales contact
B4	Usage/overage	15	excess consumption, limits and relevant tiering are documented comprehensibly
B5	Minimum quantities, contract, onboarding	10	term, minimum scope and one-off costs are identifiable
B6	Add-ons and AI	10	relevant additional costs are visible and assignable
B7	Calculator or selector	10	robust public modelling of several parameters is possible
B8	Price change/version evidence	15	changes or historical versions are documented traceably

Total score:

transparency score = B1 + B2 + B3 + B4 + B5 + B6 + B7 + B8
maximum = 100

4.2 Scoring logic

The criteria were scored ordinally. Full points do not mean that every conceivable contract variable is published. They mean that the respective dimension was sufficiently traceable for a typical public procurement check.

The analysis was carried out as desk research. The following were examined:

• official price lists
• official help centres
• official billing documentation
• official calculators
• official plan and add-on descriptions

Sales quotes, partner prices, individual discounts and non-public contracts were not collected.

4.3 Data cut-off date

All dynamic pages were assessed with a data cut-off of July 2026. Price lists can change at any time. The published version must therefore either appear as a historical snapshot or be re-checked and versioned before deployment.

---

I.5 Overall result

Rank	Product	Category	Score / 100
1	Intercom	Support	90
2	Microsoft Power BI	BI	90
3	Webflow	Web	90
4	monday.com	Work Management	90
5	New Relic	Observability	85
6	Zapier	Automation	85
7	Airtable	Work Management	80
8	HubSpot Sales Hub	CRM	80
9	Make	Automation	80
10	Notion	Work Management	80
11	Pipedrive	CRM	80
12	Zendesk	Support	80
13	n8n	Automation	80
14	Asana	Work Management	75
15	Datadog	Observability	75
16	Salesforce Sales Cloud	CRM	75
17	Tableau	BI	75
18	Activepieces	Automation	70
19	Fivetran	Data & Analytics	65
20	Twilio Segment	Data & Analytics	55

Interpretation of the score bands

Score	Interpretation
90–100	very high public pricing documentation
80–89	high public pricing documentation
70–79	usable, but with relevant gaps
60–69	key cost components can be modelled only to a limited extent
below 60	public information is barely sufficient for a robust growth scenario

The bands are editorial interpretation aids for this report. They are not an external standard.

---

I.6 Categories

Category	Products	Average
Web	1	90.0
Support	2	85.0
Business Intelligence	2	82.5
Work Management	4	81.3
Observability	2	80.0
Automation	4	78.8
CRM	3	78.3
Data & Analytics	2	60.0

[LIMITATION] Categories with one or two products must not be read as a robust market comparison. The value “Web 90”, for example, is identical to the only web product examined.

---

I.7 Full score matrix

Product	B1 /10	B2 /15	B3 /15	B4 /15	B5 /10	B6 /10	B7 /10	B8 /15	Total
Zapier	10	15	10	15	10	10	10	5	85
Make	10	15	10	15	10	5	10	5	80
n8n	10	15	15	15	10	5	5	5	80
Activepieces	10	15	5	15	10	5	5	5	70
HubSpot Sales Hub	10	15	10	10	10	10	10	5	80
Salesforce Sales Cloud	10	15	10	10	10	10	5	5	75
Pipedrive	10	15	10	10	10	5	10	10	80
Zendesk	10	15	10	10	10	10	10	5	80
Intercom	10	15	10	15	10	10	10	10	90
Asana	10	15	10	10	10	10	5	5	75
monday.com	10	15	10	15	10	10	10	10	90
Notion	10	15	10	15	10	10	5	5	80
Airtable	10	15	10	15	10	5	10	5	80
Twilio Segment	10	15	0	10	5	5	5	5	55
Fivetran	10	15	5	15	5	5	5	5	65
Tableau	10	15	10	10	10	5	10	5	75
Microsoft Power BI	10	15	15	15	10	5	10	10	90
Datadog	10	15	5	15	5	10	10	5	75
New Relic	10	15	10	15	10	10	10	5	85
Webflow	10	15	15	15	10	10	10	5	90

The machine-readable version is included in the package at:

data/FDR-2026-01_pricing_transparency_scores.csv

---

I.8 Observations by pricing model

8.1 Automation: the unit decides more than the entry price

Zapier, Make, n8n and Activepieces each publish a different economic logic:

• Zapier counts successful external actions as Tasks; triggers and certain built-in steps are treated differently.
• Make bills predominantly via Credits, with module executions forming the relevant unit.
• n8n prices by workflow Executions, regardless of the number of steps within a run.
• Activepieces uses active flows in its standard model and allows unlimited runs there; the Community Edition can be self-hosted.

[INTERPRETATION] A low monthly amount says little without translation into business events. A ten-step process can produce one, nine or ten billable units depending on the platform. This is precisely why the definition of the meter carries 15 percent of the score.

8.2 CRM: public Seats, private enterprise reality

HubSpot, Salesforce and Pipedrive show entry prices and Seat models. As complexity increases, however, further variables are added:

• onboarding
• API access
• AI functions
• additional hubs or clouds
• minimum commitments
• support and governance packages

[OBSERVED] HubSpot publishes a one-off onboarding fee for certain Professional products. Salesforce lists separate prices for individual packages and add-ons. Pipedrive remains comparatively linear in its standard plans.

[INTERPRETATION] The Seat model is easy to understand, but not automatically easy to forecast. The more functions are distributed across add-ons or package boundaries, the less the visible Seat price carries on its own.

8.3 Support: Seat plus outcome

Zendesk and Intercom show how AI is changing the traditional agent model:

• Zendesk combines agent/Seat prices with copilot- and automation-related components.
• Intercom prices Seats and Fin AI outcomes separately.

These models can be more transparent than blanket enterprise offers, but they create a new forecasting task: costs depend simultaneously on team size and on successfully automated transactions.

8.4 Work Management: Seats plus AI Credits

Asana, monday.com, Notion and Airtable publish the most important standard prices. AI functions, however, introduce additional Credits or usage logics. A procurement model must therefore separate at least three levels:

1. paid members,
2. roles and permissions,
3. AI or automation consumption.

8.5 Data & Analytics: value is measurable, the invoice is not always self-calculable

Segment and Fivetran document their central metrics – Monthly Tracked Users and Monthly Active Rows respectively – relatively clearly. The full cost curve for higher volumes, enterprise services and individual architecture can only be represented publicly to a limited extent.

[INTERPRETATION] For volume-dependent data products, a “from” price is structurally less informative than for a fixed Seat. This is not necessarily a transparency failure; it does, however, raise the requirements for calculators, tiers and worked examples.

8.6 BI, observability and web

Power BI achieves a high transparency value because standard plans and key prices are publicly visible. Tableau likewise publishes role-based prices, but requires a combination of roles for realistic deployments.

Datadog and New Relic explain many billing units, yet the large number of potential products and metrics makes an overall forecast difficult. Webflow provides prices, plan logic and a calculator, although site, workspace, ecommerce and usage-related levels still have to be separated carefully.

---

I.9 Why public pricing transparency is economically relevant

9.1 Budgeting

Missing price tiers shift cost clarification into the sales process. That is not automatically problematic, but it makes the following harder:

• independent scenario comparisons
• early architecture decisions
• robust business cases
• benchmarking against alternatives
• Exit planning

9.2 Architecture

Billing units influence system design. Examples:

• Task-based costs encourage the reduction of external actions.
• MAR-based costs encourage data filtering ahead of the pipeline.
• Host-based costs influence deployment density.
• Seat-based costs influence roles and access models.
• Outcome-based costs shift the invoice towards process performance.

[INTERPRETATION] Pricing is therefore not a purely commercial layer. It feeds back into data models, workflows and organisational design.

9.3 Procurement

A robust RFP should ask not only for “price per month”, but for:

• definition of every billing unit
• measurement point and reset
• deduplication
• overage
• minimum quantities
• plan changes
• API/export costs
• sandbox and staging costs
• support
• AI consumption
• termination and Exit costs

---

I.10 Procurement checklist

[RECOMMENDATION] Before signing a contract, a company should translate the planned usage into a billable-unit map:

business event
→ technical actions
→ billable units
→ plan tier
→ add-ons
→ monthly costs
→ stress case

Mandatory questions

1. Which technical action creates a billable unit?
2. Which actions are free of charge?
3. When does the same entity count again?
4. Which role is chargeable?
5. Which AI function requires Credits or outcomes?
6. What happens when limits are exceeded?
7. Are there automatic upgrades or pay-per-use?
8. Which functions require enterprise?
9. How long does the price stay fixed?
10. Which costs arise for export, API and Exit?

---

---

Part II - Renewal, termination and Exit transparency

II.1 Research question

How transparently do ten SaaS vendors document the economic and operational path from a running subscription to a controlled Exit?

The report considers the public information surface available to a buyer without confidential contract documents. It does not answer:

• whether a specific contract was effectively terminated,
• which individually negotiated deadlines apply,
• whether an export is complete or legally sufficient,
• how high actual migration costs turn out to be,
• whether a vendor fulfils EU Data Act obligations in an individual case,
• whether a target platform can import the exported data without loss.

---

II.2 Why renewal and Exit belong together

A renewal is not a pure billing event. It is the point at which a company either renews or loses its negotiating power. The closer the renewal comes, the less time remains for four tasks:

1. the substantive replacement decision,
2. the data and configuration inventory,
3. technical migration with parallel operation,
4. contractually effective termination.

A technically possible export helps little if the notice period has already expired. Conversely, timely termination is risky if the export has not been tested or if access ends immediately when the contract ends.

[INTERPRETATION] The economically relevant Exit point is therefore not the end of the contract, but the latest date at which contract, data, logic and operations are all still controllable at the same time.

Four separate clocks

Clock	Question
Contract clock	When does the contract renew and which notice period applies?
Data clock	When can data and configurations be extracted completely?
Operations clock	How long do the legacy and target systems have to run in parallel?
Deletion clock	When are production data, backups and remaining copies deleted?

A professional Exit runbook brings these clocks together in a single schedule.

---

II.3 Methodology

3.1 Sample

The ten platforms were chosen because they cover different SaaS categories:

• CRM and marketing automation,
• collaboration and work management,
• workflow automation,
• knowledge management,
• website platforms,
• messaging and support.

The sample is purposive, not representative of the SaaS market as a whole. It is suitable for examining recurring documentation patterns, not for extrapolating a market share.

3.2 Document types

Only public primary sources of the respective vendor were considered:

• Terms of Service or Customer Agreements,
• Product-Specific Terms,
• Data Processing Addenda,
• billing and cancellation help,
• export and backup documentation,
• retention and deletion help,
• versioned legal archives,
• EU Data Act or switching addenda, where available.

Third-party blog posts, affiliate pages and unverifiable price summaries were not used as score evidence.

3.3 Criteria and weighting

Criterion	Weight	Public evidence signal
Renewal mechanism	15	renewal, term and renewal logic publicly findable
Termination process	10	notice period, channel and operational procedure documented
Price change notice	10	rule or deadline for price changes publicly described
Self-service export	15	export possible without an individual professional services project
Export scope and formats	15	data classes, limits and formats sufficiently concrete
Post-termination window	15	access or extraction after contract end documented
Deletion and retention	10	retention, deletion or backup expiry traceable
Legal archive / switching	10	versioned terms or Data Act/switching documents findable

3.4 Scoring logic

1.0 = publicly clear and substantially documented for the aspect considered
0.5 = documented partially, scattered, plan-dependent or with material ambiguity
0.0 = not sufficiently evidenced in the public sources examined

[CRITICAL] “Not publicly evidenced” does not mean “does not exist”. It means that the information could not be counted as robust public evidence within the defined research process.

The score is:

score = sum(criterion fulfilment × weight)

There is no subsequent subjective award of bonus or penalty points.

---

II.4 Results

Vendor	Score / 100	Classification
Airtable	92.5	high public documentation maturity
Asana	92.5	high public documentation maturity
Atlassian	92.5	high public documentation maturity
HubSpot	87.5	high public documentation maturity
Intercom	87.5	high public documentation maturity
monday.com	72.5	solid public documentation maturity
Zapier	70.0	solid public documentation maturity
Slack	67.5	partial public documentation maturity
Notion	65.0	partial public documentation maturity
Webflow	62.5	partial public documentation maturity

4.1 Distribution

• [CALCULATED] Mean: 79.0
• [CALCULATED] Median: 80.0
• [CALCULATED] Range: 62.5 to 92.5
• [CALCULATED] Five of ten vendors reach at least 85 points.
• [CALCULATED] Three of ten fall below 70 points.

The relatively narrow range shows that at large SaaS platforms, public termination and export documentation is rarely completely absent. The difference arises above all in post-termination windows, versioned legal documents and the coverage of logic rather than mere data records.

---

II.5 Vendor observations

Airtable - 92.5

[OBSERVED] Airtable documents contract and renewal rules in its MSA or Terms and provides export paths for account data. Post-termination data access and deletion are addressed legally.

[LIMITATION] A table or CSV export is not automatically a restorable Airtable application. Automations, interfaces, permission details, integrations and external dependencies require a separate reconstitution check.

Asana - 92.5

[OBSERVED] The Subscriber Terms, cancellation notes, export documents and Data Act-related product terms examined form a comparatively closed public chain. A documented time window after contract end improves predictability.

[INTERPRETATION] The value shows that legal, billing and product documentation together can be stronger than a single export function.

Atlassian - 92.5

[OBSERVED] Customer Agreement, DPA and Legal Archive are well versioned and publicly traceable. Atlassian also provides product-related export and backup paths.

[LIMITATION] Jira, Confluence and Marketplace apps have different data and configuration surfaces. The score is not a claim that every app ecosystem is fully portable.

HubSpot - 87.5

[OBSERVED] Auto-renewal, renewal price reference, in-account processes and extensive export help are publicly documented. The information is, however, distributed across legal and product pages.

[RECOMMENDATION] Before a HubSpot Exit, CRM objects, marketing assets, workflows, custom properties, files, forms, reports, permissions and integrations should be treated as separate export classes.

Intercom - 87.5

[OBSERVED] Alongside general contract terms, a dedicated EU Data Act addendum exists. This improves the visibility of switching and export questions.

[LIMITATION] Plan, region, contract form and the specific product modules can influence the practical scope. An addendum does not replace a migration test.

monday.com - 72.5

[OBSERVED] Account cancellation and whole-account export are publicly explained. Post-termination access and versioned switching documents appear less consolidated.

[INTERPRETATION] The difference between “exporting the entire account” and “restoring the account with equivalent functionality in another system” has to be documented explicitly.

Zapier - 70.0

[OBSERVED] Zapier documents account data export, workflow JSON and run history export separately. That is stronger than a pure CSV data export.

[LIMITATION] Connection credentials and third-party authorisations are not readily portable as reusable secrets, for security reasons. In addition, workflow JSON does not automatically mean a runnable reproduction outside the platform.

Slack - 67.5

[OBSERVED] Cancellation, retention and workspace export are documented. Scope and approval path vary by plan, workspace type and data class.

[INTERPRETATION] Messaging platforms show particularly clearly that export rights, data protection roles, private channels, direct messages and actual import capability are separate from one another.

Notion - 65.0

[OBSERVED] Plan changes, cancellation, workspace export and backup are publicly explained. An equally clear public post-termination and Legal Archive path was less visible in the pages examined.

[LIMITATION] Markdown, CSV or HTML exports preserve content, but not necessarily relation logic, database behaviour, permissions, comments, integrations and all presentation properties.

Webflow - 62.5

[OBSERVED] Plan cancellation and code export are documented. The exported code does not, however, reproduce every hosted CMS, ecommerce, form or platform function.

[INTERPRETATION] Webflow illustrates the most important Exit question of a platform: what is exported – and what remains a runtime function of the vendor?

---

II.6 The three most common procurement mistakes

6.1 Notice period without migration lead time

A company records only the last day for giving notice. That is too late. The internal decision point has to be calculated backwards from the following durations:

internal Exit start
= notice period
+ target system selection
+ data inventory
+ migration and tests
+ parallel operation
+ management buffer

6.2 Export as a synonym for restoration

An export can be complete in the vendor’s sense and still be insufficient for the target operation. The following layers have to be checked separately:

• business objects and content,
• relationships and IDs,
• files and attachments,
• workflows and automations,
• roles and permissions,
• audit and run history,
• integrations and webhooks,
• secrets and OAuth connections,
• dashboards and reports,
• branding and frontend,
• system states and schedules.

6.3 Deletion without proof

“Account closed” does not answer:

• when production data is deleted,
• how long backups rotate,
• which statutory retention obligations apply,
• whether subprocessors are affected,
• how a deletion or return process is confirmed.

For sensitive or regulated systems, the deletion confirmation should be an acceptance item of its own.

---

II.7 The FW Delta Exit Readiness Record

A machine-readable record should exist for every business-critical SaaS system:

system: "example-saas"
business_owner: "..."
technical_owner: "..."
contract_owner: "..."
renewal_date: "YYYY-MM-DD"
notice_period_days: 90
internal_exit_decision_date: "YYYY-MM-DD"
export_owner: "..."
export_last_tested: "YYYY-MM-DD"
export_artifacts:
  - data
  - files
  - workflows
  - permissions
  - audit_history
post_termination_access_days: 30
retention_summary: "..."
deletion_confirmation_required: true
target_runtime: "..."
parallel_run_days: 21
open_risks:
  - "..."

[RECOMMENDATION] This record does not belong in an individual’s personal collection, but in the contract, architecture or service portfolio repository.

---

II.8 Schedule for a controlled renewal

180 to 120 days before renewal

• Measure value, active Seats and actual process dependency.
• Secure the contract version and order form.
• Inventory the export scope.
• Review target architectures and alternatives.
• Name those responsible for Exit and renewal.

120 to 90 days

• run a complete test export,
• test import or reconstitution in an isolated environment,
• measure API limits and data volumes,
• calculate the cost of dual operation,
• verify termination requirements in writing.

90 to 60 days

• take the decision,
• trigger termination or renegotiation,
• secure support and export dates,
• plan a change freeze for critical data classes,
• draw up a communication plan.

60 days to contract end

• delta synchronisation,
• user and role switchover,
• migrate webhooks, domains, SSO and integrations,
• document acceptance criteria,
• trigger the deletion and retention process.

After contract end

• test access,
• archive final exports and hashes,
• check invoices and Credits,
• obtain deletion confirmation,
• revoke old tokens, OAuth grants and DNS connections,
• feed lessons learned back into procurement.

---

II.9 Relationship to the EU Data Act

The EU Data Act strengthens the framework for switching between data processing services. For operational implementation, however, concrete questions remain:

• Is the product a covered data processing service?
• Which data and digital assets fall under the switch?
• Which deadlines and support services apply?
• Which export formats and interfaces are offered?
• Which components are third-party or marketplace services?
• Which costs are switching charges and which are normal usage?
• Which obligations lie with the customer?

[CRITICAL] This benchmark assesses public transparency signals, not the legal compliance of a vendor.

---

---

Part III - Per-Seat and AI scaling stress

III.1 Research question

How do the publicly visible annual software costs change when a company licenses the same business tools for 10, 50 or 200 chargeable identities – and how does AI change the billing unit?

The report does not answer which product is best. It examines the shape of the cost. A more expensive Seat can be more economical if it replaces several systems or demonstrably creates more value. A cheap Seat can be uneconomical if it remains unused on a large scale.

---

III.2 Sample and scope

2.1 Selection

The sample covers ten widespread categories:

• CRM and sales: HubSpot, Salesforce, Pipedrive
• Customer support: Zendesk
• Work management: Asana, monday.com
• Knowledge and collaboration work: Notion
• Low-code database: Airtable
• Business intelligence: Tableau, Power BI

In each case a publicly priced, feature-rich business plan was selected, not necessarily the cheapest or best-selling plan.

2.2 What a “Seat” means here

The vendors use different terms: user, seat, member, agent. The report normalises them as a chargeable identity. This does not mean that the rights are identical. A support agent, a CRM sales Seat and a BI user are not functionally comparable.

2.3 Not included

• individually negotiated enterprise discounts
• value added tax
• implementation, migration and training, except for explicitly stated mandatory onboarding
• premium support
• API, storage, data, automation or overage costs
• regional taxes and payment fees
• free viewer, guest or light roles
• bundle benefits, existing Microsoft/Salesforce contracts or reseller terms

---

III.3 Methodology

3.1 Data cut-off date

All price data was collected in July 2026 from publicly accessible vendor pages. For USD and GBP, the ECB reference rate of 28 July 2026 was used:

1 EUR = 1.1367 USD
1 EUR = 0.85550 GBP

The ECB points out that reference rates serve information purposes and are not recommended as transaction rates. For the comparison they are a reproducible normalisation, not a procurement rate.

3.2 Formula

annual price EUR per identity
= list price per month × 12 ÷ foreign currency per EUR

For euro prices, no currency conversion applies.

stress cost at N identities
= annual price EUR per identity × N

3.3 Linearity warning

The model extrapolates publicly visible prices linearly. Above public volume limits in particular, this is not a claimed quoted price. monday.com, for example, requires an individual quote for more than 40 users. The 50- and 200-Seat values there are exclusively a mathematical sensitivity.

---

III.4 Full price sample

Vendor	normalised plan	published price	EUR/year per identity	10	50	200
HubSpot	Sales Hub Professional	90 USD	950.12 €	9,501.19 €	47,505.94 €	190,023.75 €
Salesforce	Pro Suite	80 GBP	1,122.15 €	11,221.51 €	56,107.54 €	224,430.16 €
Pipedrive	Premium	59 EUR	708.00 €	7,080.00 €	35,400.00 €	141,600.00 €
Zendesk	Suite Professional	89 GBP	1,248.39 €	12,483.93 €	62,419.64 €	249,678.55 €
Asana	Starter	10.99 USD	116.02 €	1,160.20 €	5,801.00 €	23,204.01 €
monday.com	Pro	19 EUR	228.00 €	2,280.00 €	11,400.00 €	45,600.00 €
Notion	Business	19.50 EUR	234.00 €	2,340.00 €	11,700.00 €	46,800.00 €
Airtable	Business	45 USD	475.06 €	4,750.59 €	23,752.97 €	95,011.88 €
Tableau	Enterprise	35 USD	369.49 €	3,694.91 €	18,474.53 €	73,898.13 €
Microsoft	Power BI Pro	12.10 EUR	145.20 €	1,452.00 €	7,260.00 €	29,040.00 €

Machine-readable version:

data/FDR-2026-01_per_seat_stress.csv

4.1 Interpretation of the range

The range of a factor of 10.8 does not explain the difference in value. The products solve different problems. It does show, however, why a blanket “We have 200 employees” is unusable for budget planning. What matters is:

• How many people need write access?
• How many need viewer access only?
• Which roles are permitted in which plan?
• Are service accounts, bots or external partners chargeable?
• Are inactive accounts automatically revoked?
• Does a feature force all users onto a higher plan?

---

III.5 Stress test A: ten products, same Seat count

[SCENARIO] In this deliberately maximalist scenario, every chargeable identity holds a Seat in every product.

identical Seats per product	annual list price total
1	5,596.43 €
10	55,964.32 €
50	279,821.62 €
200	1,119,286.48 €

What this table states

It measures the price leverage of the identity count. It does not say that a company should buy all ten products in parallel. Duplications such as HubSpot/Salesforce/Pipedrive or Tableau/Power BI would be pointless in many architectures.

What it makes visible

With person-based billing, organisational growth automatically becomes software growth, even if:

• the data volume stays unchanged,
• the number of workflows stays unchanged,
• the number of dashboards stays unchanged,
• the technical infrastructure requires barely any more resources.

That can be justified if the vendor provides support, security, collaboration and product value per user. For TCO and make-or-buy, however, the coupling has to be modelled explicitly.

---

III.6 Stress test B: a more plausible five-product stack

The second model stack avoids direct CRM and BI duplicates:

1. HubSpot Sales Hub Professional
2. Asana Starter
3. Notion Business
4. Tableau Enterprise
5. Power BI Pro

Tableau and Power BI can also overlap in reality; they are both retained here in order to represent two widespread analytics licensing forms.

chargeable identities	annual Seat costs	one-off mandatory onboarding	first-year costs
10	18,148.29 €	1,319.61 €	19,467.90 €
50	90,741.47 €	1,319.61 €	92,061.08 €
200	362,965.89 €	1,319.61 €	364,285.50 €

[INTERPRETATION] A stack can look economically unremarkable at 10 users and become an architecture question at 200 users. The feature scope has not necessarily increased twentyfold; the number of licensed identities has.

---

III.7 Unused Seats as a budget risk

An “unused Seat” in this report is a paid identity without sufficient active usage within the defined measurement window. This is not a universal definition. A rarely used emergency or compliance access can be deliberately necessary.

For the 200-Seat five-product stack, the pure sensitivity gives:

assumed permanently unused share	committed annual budget
5 %	18,148.29 €
10 %	36,296.59 €
15 %	54,444.88 €
20 %	72,593.18 €
25 %	90,741.47 €

These values are not a claimed industry-wide waste rate. They are a decision table. A company has to determine its own share from login, event and role information.

Clean measurement

A robust licence usage analysis should distinguish at least:

• assigned
• signed in within 30/60/90 days
• core function used
• content created or changed
• consumed only
• automated/service-based access
• suspended but still billed
• external guest
• regulatory required emergency access

---

III.8 AI changes the model – but not in one direction

8.1 Seat-based AI surcharges

Vendor	AI unit	public price	EUR/year per paying identity	50	200
Salesforce	Agentforce for Sales	from 100 GBP/user/month	1,402.69 €	70,134.42 €	280,537.70 €
Zendesk	Copilot	40 GBP/agent/month	561.08 €	28,053.77 €	112,215.08 €

8.2 Usage-based AI units

For Custom Agents after the trial phase, Notion states 10 USD per 1,000 Notion Credits. HubSpot and monday.com show included Credits in the plans considered. Without a robust consumption profile, no serious annual figure can be derived from this.

Machine-readable version:

data/FDR-2026-01_ai_addon_stress.csv

8.3 The new cost vector

AI creates at least four possible models:

1. AI included in the existing Seat
2. AI as an additional surcharge per Seat
3. AI as a Credit/token/action model
4. AI as an outcome or agent price

[INTERPRETATION] The “death of the per-Seat model” is not evidenced by this sample. What is observable is a hybridisation. Companies pay for human access and machine usage at the same time.

8.4 Why agents do not simply mean “no logins”

A software agent can technically require a service account, API key or OAuth principal. Vendors can price such identities differently or limit them contractually. The economic question is therefore not only “Does the agent have a login?”, but:

• Which licensed function does it use?
• On whose behalf does it act?
• How many actions/calls/Credits does it consume?
• Does every human recipient still need a Seat?
• Is machine usage permitted under the contract terms?

---

III.9 Procurement model for renewals in 2026

[RECOMMENDATION] For every product, the renewal baseline should be available as a table:

Field	Mandatory question
Billing unit	user, member, agent, creator, workspace, Credit, action or outcome?
Seat basis	all employees, active users or only certain roles?
Minimum package	minimum count or fixed Seat buckets?
Viewer/guest	free, restricted or chargeable?
Inactivity	When is deprovisioning automatic?
AI	included, Seat add-on, Credits or overage?
Service accounts	permitted and priced how?
Contract term	monthly, annual or multi-year?
Price change	cap, indexation and renewal mechanics?
Exit	export, API, transition and deletion deadline?

Four metrics

Seat Utilization
= active core users / paid Seats

Cost per Active User
= annual costs / active core users

AI Cost per Completed Outcome
= AI costs / accepted outcomes

License-to-Infrastructure Ratio
= annual licence costs / estimated technical operating costs

The last metric does not prove that in-house development makes sense. It only shows when the difference becomes large enough to justify a build/buy/own review.

---

III.10 Architecture implications

Per-Seat makes sense if

• value and support genuinely rise with the number of users,
• collaboration creates the main benefit,
• functional and compliance requirements are high,
• the vendor provides strong economies of scale,
• switching and operating costs of an in-house solution would be higher.

An alternative cost base should be examined if

• a very large number of people only read,
• few workflows serve very many recipients,
• agents perform actions on behalf of numerous people,
• the same data logic is duplicated across several Seat products,
• the system represents a stable, company-specific core process,
• annual licence costs are permanently above build plus operations.

Possible alternatives are not limited to in-house development. Role clean-up, cheaper plans, capacity pricing, Open Source, self-hosting, consolidation or a different SaaS vendor can also be economically better.

---

---

3. Integrated procurement model

3.1 The Commercial Transparency Record

A versioned record should exist for every critical SaaS system:

product:
plan:
billing_currency:
billing_unit:
included_volume:
overage_rule:
seat_definition:
inactive_seat_policy:
ai_meter:
mandatory_addons:
onboarding_fee:
minimum_term:
renewal_date:
cancellation_deadline:
cancellation_channel:
price_change_notice:
self_service_export:
exported_state_classes:
post_termination_window:
retention_and_deletion:
contract_archive_url:
last_verified:
owner:

This record is not a substitute for the contract. It is the operational bridge between finance, procurement, legal and engineering.

3.2 180-day renewal plan

Point in time	Mandatory activity	Result
T−180	Inventory usage, Seats, workspaces, add-ons and contract documents	complete cost and dependency base
T−150	Run an export and restore test	evidenced portability gaps
T−120	Model alternatives and an internal ownership scenario	realistic BATNA
T−90	Calculate target state, parallel operation and migration effort	decision-ready business case
T−60	Approve negotiation or migration	signed plan
T−30	Execute termination, renewal or transition contract with legal certainty	documented resolution
T−0	Update contract, price and Exit Record	next controlled period

3.3 Minimum questions before signing

1. What is the smallest technically and contractually billable unit?
2. Which usage creates additional units?
3. Which plan functions become mandatory as the company grows?
4. Which fees are one-off, which recurring, which consumption-dependent?
5. How are price changes announced?
6. When and how does the contract renew?
7. Which channel is binding for termination?
8. Which data and state classes can be exported?
9. Which configurations, permissions, automations and audit data are missing?
10. How long does access persist after contract end?
11. When are backups and residual copies deleted?
12. Is there a documented import or reconstitution path?
13. Which functions or data are released only via professional services?
14. How are service accounts, bots, agents and external users priced?
15. Which AI costs are charged by Seat, Credit, token, Task or outcome?

---

4. Visualisation specification for fwdelta.com

Every chart must be generated from the enclosed CSVs. No values may be transferred manually from the running text.

Chart 1 - Pricing transparency by product

• Type: horizontal bars
• Data: FDR-2026-01_pricing_transparency_scores.csv
• X axis: score 0–100
• Sorting: descending
• Note: “Assesses public documentation, not price/performance”

Chart 2 - Pricing vs. renewal/Exit transparency

• Type: scatter plot
• Data: FDR-2026-01_commercial_transparency_composite.csv
• X: pricing transparency
• Y: renewal/Exit transparency
• Label: vendor
• Quadrants: 80/80 as editorial orientation, not as a standard

Chart 3 - Seat stress 10/50/200

• Type: grouped bars or small multiples
• Data: FDR-2026-01_per_seat_stress.csv
• Y: annual EUR list price costs
• Caveat visible: no discounts, no implementation, no taxes

Chart 4 - Four clocks of the Exit

• Type: timeline/swimlane
• Data: editorial, from section 2.2
• No numeric claims without contract data

Chart 5 - Commercial Transparency Funnel

price visible
→ meter understood
→ growth modellable
→ renewal controlled
→ export verified
→ operations reconstructable

This chart is a conceptual model and must be labelled as such.

---

5. Reproducibility

Data files

• data/FDR-2026-01_pricing_transparency_scores.csv
• data/FDR-2026-01_renewal_exit_transparency.csv
• data/FDR-2026-01_per_seat_stress.csv
• data/FDR-2026-01_ai_addon_stress.csv
• data/FDR-2026-01_commercial_transparency_composite.csv

Reproduction steps

1. Open every official pricing, billing, terms, export and retention page on the new cut-off date.
2. Record URL, document version, region and currency.
3. Score the criteria exclusively against the published rubric.
4. Do not phrase “not found” as “does not exist”.
5. Document the currency conversion with cut-off date and source.
6. Repeat the Seat stress only for publicly priced, genuinely comparable plans.
7. Calculate the composite exclusively for the intersection of the modules.
8. Log all changes in the changelog and claim ledger.

---

6. Limitations

1. The samples are purposive and not representative of the SaaS market as a whole.
2. Public documentation can deviate from individual contractual rights.
3. A high transparency score proves neither cheap prices nor good Exit capability.
4. A low score does not prove that the vendor withholds information from customers; it may be available in the sales process or in the contract.
5. Prices, AI packages and meters change frequently.
6. Currency conversion creates an additional dependency on the cut-off date.
7. Seat scenarios ignore discounts and minimum commitments unless publicly stated.
8. The composite weights pricing and renewal/Exit at 50 percent each; this weighting is an FW Delta methodological decision.
9. Export documentation was not validated by real enterprise migrations for every vendor.
10. Legal interpretation of contracts is not the subject of the report.

---

7. Citation notes

Permitted:

“In the FW Delta SaaS Commercial Transparency Index 2026, the median of the 20 pricing surfaces examined was 80 out of 100 points; the index assesses public documentation, not price/performance.”

Not permitted:

“80 percent of all SaaS vendors are transparent.”

Permitted:

“In a selected public plan sample, the normalised annual costs at 200 Seats ranged from roughly 23,000 to 250,000 euros per product.”

Not permitted:

“At 200 employees, SaaS always costs at least 23,000 euros.”

---

8. Version and correction history

• 1.0 · July 2026 · Flagship edition consolidated from three separate modules; eight vendors in the joint composite; data files and reproduction rules harmonised.
• Live version open · Re-check all volatile pricing, terms and help centre pages before publication.

---

9. Sources

1. Zapier, “Plans and pricing”, https://zapier.com/pricing, accessed in July 2026.
2. Zapier Help, “How is task usage measured?”, https://help.zapier.com/hc/en-us/articles/8496196837261-How-is-task-usage-measured-in-Zapier, accessed in July 2026.
3. Make, “Pricing”, https://www.make.com/en/pricing, accessed in July 2026.
4. Make Help Center, “Credits”, https://help.make.com/credits, accessed in July 2026.
5. n8n, “Pricing”, https://n8n.io/pricing/, accessed in July 2026.
6. Activepieces, “Pricing”, https://www.activepieces.com/pricing, accessed in July 2026.
7. HubSpot, “Sales Hub pricing”, https://www.hubspot.com/pricing/sales, accessed in July 2026.
8. Salesforce, “Sales pricing”, https://www.salesforce.com/uk/sales/pricing/, accessed in July 2026.
9. Pipedrive, “Pricing”, https://www.pipedrive.com/en/pricing, accessed in July 2026.
10. Zendesk, “Pricing”, https://www.zendesk.com/pricing/, accessed in July 2026.
11. Intercom, “Pricing”, https://www.intercom.com/pricing, accessed in July 2026.
12. Asana, “Pricing”, https://asana.com/pricing, accessed in July 2026.
13. monday.com, “Pricing”, https://monday.com/pricing, accessed in July 2026.
14. Notion, “Pricing”, https://www.notion.com/pricing, accessed in July 2026.
15. Airtable, “Pricing”, https://airtable.com/pricing, accessed in July 2026.
16. Twilio Segment, “Pricing”, https://segment.com/pricing, accessed in July 2026.
17. Twilio Segment, “Connections pricing”, https://segment.com/pricing/connections, accessed in July 2026.
18. Fivetran, “Pricing”, https://www.fivetran.com/pricing, accessed in July 2026.
19. Tableau, “Pricing”, https://www.tableau.com/pricing, accessed in July 2026.
20. Microsoft, “Power BI pricing”, https://www.microsoft.com/de-de/power-platform/products/power-bi/pricing, accessed in July 2026.
21. Datadog, “Pricing”, https://www.datadoghq.com/pricing/, accessed in July 2026.
22. New Relic, “Pricing”, https://newrelic.com/pricing, accessed in July 2026.
23. Webflow, “Pricing”, https://webflow.com/pricing, accessed in July 2026.
24. Webflow, “Pricing calculator”, https://webflow.com/pricing/calculator, accessed in July 2026.
25. HubSpot, Archived Customer Terms of Service – 2 February 2026, https://legal.hubspot.com/archive-tos-222026, accessed in July 2026.
26. HubSpot, Product Specific Terms, https://legal.hubspot.com/product-specific-terms, accessed in July 2026.
27. HubSpot Knowledge Base, Export your content and data, https://knowledge.hubspot.com/account-management/export-your-content-and-data, accessed in July 2026.
28. Slack, Customer Terms of Service, https://slack.com/terms-of-service, accessed in July 2026.
29. Slack Help, Export your workspace data, https://slack.com/help/articles/201658943-Export-your-workspace-data, accessed in July 2026.
30. Slack Help, Customize data retention, https://slack.com/help/articles/203457187-Customize-data-retention-in-Slack, accessed in July 2026.
31. Slack Help, Change or cancel your paid plan, https://slack.com/help/articles/48764458651795-Change-or-cancel-your-paid-Slack-plan, accessed in July 2026.
32. Notion, Terms and Privacy, https://www.notion.so/terms, accessed in July 2026.
33. Notion Help, Change your Notion plan, https://www.notion.com/help/upgrade-or-downgrade-your-plan, accessed in July 2026.
34. Notion Help, Export your content, https://www.notion.com/help/export-your-content, accessed in July 2026.
35. Notion Help, Back up your data, https://www.notion.com/help/back-up-your-data, accessed in July 2026.
36. Asana, Subscriber Terms – May 2026, https://asana.com/terms/subscriber-terms-052026, accessed in July 2026.
37. Asana, Subscriber Agreement, https://asana.com/terms/subscriber-agreement, accessed in July 2026.
38. Asana, Product-Specific Terms – July 2026, https://asana.com/terms/asana-product-specific-terms-071026, accessed in July 2026.
39. monday.com, Terms of Service, https://monday.com/l/terms, accessed in July 2026.
40. monday.com Support, Cancel and close your account, https://support.monday.com/hc/en-us/articles/360018133779-How-to-cancel-and-close-your-account, accessed in July 2026.
41. monday.com Support, Export your entire account’s data, https://support.monday.com/hc/en-us/articles/360002543719-How-to-export-your-entire-account-s-data, accessed in July 2026.
42. Intercom, Terms and Policies, https://www.intercom.com/legal/terms-and-policies, accessed in July 2026.
43. Intercom, EU Data Act Addendum, https://www.intercom.com/legal/eu-data-act-addendum, accessed in July 2026.
44. Airtable, Master Subscription Agreement, https://www.airtable.com/company/msa, accessed in July 2026.
45. Airtable, Terms of Service, https://www.airtable.com/company/tos, accessed in July 2026.
46. Airtable Support, Exporting data from Airtable, https://support.airtable.com/docs/exporting-data-from-airtable, accessed in July 2026.
47. Webflow, Terms, https://webflow.com/legal/terms, accessed in July 2026.
48. Webflow Help, Downgrade or cancel a Site plan, https://help.webflow.com/hc/en-us/articles/33961220482707-Downgrade-or-cancel-your-Site-plan, accessed in July 2026.
49. Webflow Help, Export Webflow site code, https://help.webflow.com/hc/en-us/articles/33961386739347-How-do-I-export-my-Webflow-site-code, accessed in July 2026.
50. Atlassian, Customer Agreement, https://www.atlassian.com/legal/atlassian-customer-agreement, accessed in July 2026.
51. Atlassian, Data Processing Addendum, https://www.atlassian.com/legal/data-processing-addendum, accessed in July 2026.
52. Atlassian, Legal archives, https://www.atlassian.com/legal/archives, accessed in July 2026.
53. Zapier, Terms of Service, https://zapier.com/terms, accessed in July 2026.
54. Zapier Help, Export or delete your account data, https://help.zapier.com/hc/en-us/articles/8496289406349-Export-or-delete-your-Zapier-account-data, accessed in July 2026.
55. Zapier Help, Import and export Zap workflows, https://help.zapier.com/hc/en-us/articles/8496308481933-Import-and-export-Zap-workflows-in-your-Team-or-Enterprise-account, accessed in July 2026.
56. Zapier Help, Export your Zap history, https://help.zapier.com/hc/en-us/articles/8496294549005-Export-your-Zap-history, accessed in July 2026.
57. European Central Bank, “Euro foreign exchange reference rates”, reference rates of 28 July 2026, https://www.ecb.europa.eu/stats/policy_and_exchange_rates/euro_reference_exchange_rates/html/index.en.html, accessed in July 2026.

---

Disclosure and disclaimer

FW Delta builds custom software and self-controlled infrastructure and can benefit economically from SaaS replacements. This position shapes the research question. To make the conflict of interest verifiable, criteria, raw data, weightings, limitations and sources are disclosed.

The report is not legal, tax, accounting, purchasing or investment advice. List prices are snapshots, not binding offers. A specific contract must be reviewed with the relevant qualified professionals.