How SaaS Vendors Hide 30-50% Price Hikes (2020-2026)

The problem: When SaaS vendors raise prices, the headline number on the pricing page is usually the smallest part of the story. A vendor that announces a 7% sticker bump is easy to spot. A vendor that holds the sticker steady, cuts its free tier, switches the unit of billing from per-seat to per-AI-resolution, or starts counting unsubscribed contacts toward your paid plan limit — that vendor is also raising prices, often by 30-50%, and the headline tracker won't see it. The pricing-page archaeology that catches the second pattern is the work that procurement teams, RevOps leads, and SaaS journalists keep asking for, and almost nobody does.

This is a documentary look at six years of pricing-page data on ten well-known B2B SaaS vendors, captured 2026-05-08. The headline finding: half of the sample raised sticker prices in line with inflation, three quietly restructured plans for an effective 30-50%+ hike without touching the headline number, and one — Figma — held its Professional plan at $12 per editor for seven straight years. The Figma flat-pricing claim is the only price point in this dataset directly verified against Wayback Machine snapshots; the other figures are sourced from contemporaneous press coverage and pricing-history publications, and that is disclosed throughout. Citation-worthy data depends on calibration as much as on the numbers themselves.

What is SaaS stealth pricing? A pricing-strategy pattern in which a SaaS vendor holds or even cuts the headline price displayed on its pricing page while raising the effective customer cost through plan restructure, free-tier reductions, billing-unit changes, or usage-based add-ons. The headline tracker doesn't see it; the customer's renewal invoice does.

Why it matters: Most public SaaS pricing analysis tracks sticker prices over time. Sticker-only tracking systematically understates real customer cost growth. In this 10-vendor dataset, three vendors raised effective customer cost by 30-50%+ over six years while the headline number rose less than 15% — a gap large enough to mislead procurement budgeting, board-level cost-of-software reporting, and SaaS-industry inflation commentary.

Use it when: You're ranking SaaS vendors by real cost growth, briefing procurement on renewal-cycle inflation, building a pricing-strategy comparison, or fact-checking an "AI is raising SaaS prices" narrative against six years of underlying pricing-page data.

Key findings

• 10 SaaS vendors tracked, 2020 to 2026. Sample is curated, not random — see methodology.
• Half raised sticker prices 7-33%. Salesforce +33%, Adobe +27%, Asana +10%, Slack +9%, Zoom +7%.
• Three restructured plans for an effective 30-50%+ hike. Mailchimp, HubSpot, Intercom — none touched the headline number.
• Two held flat for 5+ years. Notion held the entry-paid plan at $8/user/month; Figma held Professional at $12/editor/month for seven years.
• Three vendors are Wayback-verified at the price level. Figma 2020 ($12 / $15 / $45 — identical to 2026), Asana 2020 (Premium $10.99 — matches dataset), Zoom 2020 (Pro $14.99 — matches dataset). Other seven vendors' 2020 prices come from contemporaneous press coverage.
• US CPI rose ~22% from January 2020 to May 2026 (BLS CUUR0000SA0). Salesforce, Adobe, and the stealth restructurers all exceeded inflation; Slack, Zoom, Notion, and Figma fell below it.
• Intercom publicly destroyed ~$60M of legacy ARR to switch from per-seat to per-AI-resolution pricing, per its own Mostly Metrics interview.
• HubSpot's November 2024 restructure added per-additional-seat charges to a previously bundled tier — a 51% effective increase for a 10-seat team on Marketing Hub Pro.

In this article: The 10-vendor leaderboard · Three pricing archetypes · Stealth restructurers, decoded · Why Figma is the SaaS pricing anomaly of the decade · Notion: the cleanest restraint case · Methodology · Press lift-out for journalists · FAQ

Concrete pricing changes at a glance

The 10-vendor leaderboard

Pricing and plan-structure data based on publicly available information as of May 2026 and may change. Where the 2020 baseline column is sourced from press coverage rather than direct Wayback Machine snapshot, see methodology.

Three pricing archetypes

Mature SaaS doesn't price-hike on a single axis. The 10-vendor sample sorts cleanly into three patterns, and each pattern has a different signature on the customer's bill.

Steady raisers publish their hikes. Salesforce raised Sales Cloud Professional from $75 to $100 over six years — a documented increase that procurement teams can plan against. Adobe raised Creative Cloud All Apps from $54.99 to $69.99, with 2023 and 2025 cited explicitly in its pricing announcements. Asana removed its small-team discount in April 2020 and let the per-seat rate drift up afterwards. The hikes are visible, predictable, and roughly inflation-tracking when measured against US CPI's ~22% rise over the same window.

Stealth restructurers hold or cut the sticker price while changing the unit of sale. The headline number on the pricing page in 2026 looks similar to 2020. The renewal invoice doesn't. Three vendors in the sample fit this pattern, and they are decoded individually below.

Flat-pricers keep the sticker price unchanged for 5+ years. The strategy isn't restraint — it's a deliberate market-share bet. Figma operated through Adobe acquisition negotiations (2022-2023, eventually blocked by EU regulators), competitive pressure from Sketch and Adobe XD, and renewed entrants like Penpot and Modyfi. Holding $12/editor flat was the answer to all three.

Stealth restructurers, decoded

The interesting story isn't the steady raisers. It's the vendors whose pricing pages look almost unchanged but whose effective customer cost rose dramatically.

Mailchimp: free-tier cuts plus per-contact billing on unsubscribers

Mailchimp's Standard plan in 2020 cost $14.99/month for 10,000 contacts. The same Standard plan tier in 2024-2026 costs roughly $20-23/month for the same contact volume — a 33-53% effective increase, depending on how Mailchimp's tier reorganisations are mapped onto the 2020 structure (PriceTimeline, Mailchimp pricing).

The headline number isn't where the customer cost lives. Mailchimp made three structural changes that compound:

• Free plan limits cut. 12,000 emails per 2,000 contacts in 2019 → 10,000 emails per 2,000 contacts in 2020 → 500-contact cap and 1,000 sends per month in 2023.
• Tier reorganisation. A single per-contact pricing model became four discrete tiers, with steeper jumps between them at higher contact volumes.
• Per-contact billing including unsubscribed contacts. A customer with 9,500 active and 500 unsubscribed contacts is now billed at the 10,000-contact tier — even though 500 of those contacts will never receive another email.

A customer running the same email program in 2020 and 2026 pays meaningfully more in 2026 without ever seeing a price-increase email.

HubSpot: November 2024 core-seats-plus-add-on-seats restructure

HubSpot Marketing Hub Pro went from approximately $800/month (single-seat-era pricing, pre-November-2024) to $890/month base in late 2024 — a +11% sticker increase that sounds modest. The 2020 baseline is sourced from contemporaneous press coverage rather than direct Wayback verification (the actor's content-diff captured zero pricing-categorised changes for HubSpot, suggesting the pricing page is heavily JS-rendered and the CDX-level diff missed structural changes). The structural change is the per-additional-seat charge.

The new Pro tier includes only three core seats. Each additional seat costs $45/month (HubSpot pricing, Encharge — HubSpot pricing breakdown). A 10-seat team that previously paid $800 now pays $890 + (7 × $45) = $1,205. That's a 51% effective increase — and it shows up only on a customer's seat-utilisation invoice, not on the public pricing page.

The pattern: the public sticker is still close to the 2020 number; the customer cost is computed from a different formula.

Intercom: deliberate $60M ARR destruction to switch to outcome-based pricing

Intercom's restructure is the most aggressive in the sample, and the only one whose financial impact has been publicly confirmed by the vendor. The 2020 baseline was a per-seat model — Pro plans included two seats with a base price around $87/month, plus $39/month for each additional seat per contemporaneous Intercom pricing reviews. In 2023-2024, Intercom cut headline seat prices on its legacy support-tools plans aggressively, then introduced $0.99 per AI resolution as the new billing primitive for its Fin AI Agent product.

Intercom's CFO discussed the move in detail with Mostly Metrics: the company describes destroying approximately $60M of legacy ARR to reposition pricing around AI-driven outcomes rather than per-seat headcount. For customers whose support volume scales with usage, the effective price floor rose substantially even as the per-seat headline came down.

The cross-vendor pattern across all three: when sticker increases would risk churn, the vendor restructures the unit of sale so the headline stays stable while the formula behind the customer's bill changes. This isn't sloppy SaaS pricing — it's a documented strategy discussed openly in pricing-strategy literature.

Why Figma is the SaaS pricing anomaly of the decade

Figma kept its Professional plan at $12 per editor per month from launch in 2018 through 2025 — seven years of zero sticker-price increases during a period when the average SaaS in this dataset raised headline prices 10-33%. When Figma finally raised in 2025, it didn't raise the unit price. It bundled FigJam and Slides into the existing $12 plan, delivering more product for the same money (SaaSPricePulse — Figma Pricing History, Pricing SaaS — Inside Figma's Pricing Evolution).

This is the only price point in this dataset that has been directly verified against Wayback Machine snapshots. Snapshots of the Figma pricing page from April, July, and September 2020 explicitly mention $12, $15, and $45 — identical to the live 2026 Figma pricing page (Professional $12 annual, $15 monthly, Organization $45). The Wayback verification was run via the wayback-machine-search Apify actor against the Figma pricing URL with dateFrom: 2020, dateTo: 2026, and includeContent: true.

The competitive context explains the choice. Figma operated under existential pressure from Adobe XD and Sketch in the early years, then through Adobe acquisition negotiations from 2022-2023 (eventually blocked by EU regulators), then under renewed competitive entry from Penpot and Modyfi. Holding the sticker flat was a deliberate market-share strategy, and bundling FigJam + Slides into the existing $12 plan in 2025 reportedly produced the opposite reaction to the stealth-restructure pattern: customers received it as a value increase.

For pricing-page archaeologists, Figma is the control case. It proves that SaaS price increases are a strategic choice, not an inevitable inflation effect.

Notion: the cleanest restraint case

Notion renamed its team-tier plan from "Team" to "Plus" and held the price at $8/user/month from 2020 through 2026 (Notion pricing). No structural change. No per-contact billing. No additional-seat surcharges. No usage-based add-ons in the base plan.

Plan-rename is sometimes a cover for a price hike. Asana's Premium-to-Starter rename in 2023 came after a 10% effective increase. Notion's didn't — the rename was cosmetic. In a six-year window where Salesforce raised 33% and Mailchimp restructured for an effective 53%, Notion is the cleanest example of a SaaS vendor holding the headline price across the inflationary 2021-2023 period.

The caveat: Notion launched the Notion AI add-on in 2023 at an additional $8-10 per seat per month. Customers who adopted Notion AI saw their base $8/user cost roughly double. Then in May 2025, Notion discontinued the standalone AI add-on entirely and bundled full AI access (Notion AI Agents, Ask Notion) only into the Business tier at $20/user/month. Customers who wanted full AI access went from "Plus + AI add-on at $16-18/user" to "Business required at $20/user" — a structural change disguised as feature simplification. The Plus plan's headline price remained $8; the platform's effective price for AI-using teams rose 2.5×.

Methodology

Tool. wayback-machine-search — an Apify actor that wraps the Internet Archive Wayback Machine CDX API and returns structured snapshot timelines for any URL. Run details for this analysis: 10 vendor pricing URLs, dateFrom: 2020, dateTo: 2026, outputMode: timeline, includeContent: true, changeIntelligence: true. Run ID yJBcW4d7JmCbcecmg. Runtime 1,219 seconds. 4,442 dataset items returned.

Vendors. Ten well-known B2B SaaS products with publicly-discussed pricing histories: Salesforce, Adobe, Asana, Slack, Zoom, Mailchimp, HubSpot, Intercom, Figma, Notion. Selection criteria: name recognition with US/EU buyers, English-language pricing pages, URLs that have remained stable since 2020.

Plan tier tracked. Entry-tier paid plan — Professional, Standard, Pro, Plus, or equivalent. This is the tier most commonly used as the "ungated" benchmark in SaaS pricing comparisons.

Currency and billing. All prices in USD, annual-billed unless noted. Where vendors offer monthly-billed surcharges (e.g., Slack $7.25 annual / $8.75 monthly), the annual rate is used for consistency.

2020 baseline. "2020 price" means the price effective during calendar year 2020. Where prices changed mid-year (Asana's small-team-discount removal in April 2020), the post-change number is used and annotated.

2026 reference. Prices effective May 2026, captured 2026-05-08 from the live vendor pricing pages.

Verification status. Three vendors are directly Wayback-verified at the price level: Figma 2020 (multiple snapshots show $12, $15, $45 identical to the 2026 live page), Asana 2020-05-30 (rendered snapshot mentions $10.99 Premium plus other tier prices, matching the dataset's post-April-2020 figure), and Zoom 2020-06-01 (rendered snapshot mentions $14.99 Pro plus other tier prices, matching the dataset). Verification was run via the wayback-machine-search Apify actor with renderJs: "on" to capture JS-rendered pricing tiers from archived pages. The other seven vendors' 2020 prices come from contemporaneous press coverage and pricing-history publications: SaaSPricePulse, Mostly Metrics, DPReview, Encharge, and PriceTimeline. A fuller verification pass on those seven — Salesforce, HubSpot, Slack, Notion, monday.com, Atlassian, Mailchimp — would require Wayback re-runs against alternate near-2020 dates (the 2020-06-01 target produced empty SPA shells, navigation timeouts, or no-snapshot-available errors for these vendors) and is flagged as a pre-publish gap.

Pricing-categorised version changes per vendor (out of 122 total in the dataset): monday.com 25, Asana 15, Notion 14, Slack 12, Atlassian Jira 8, Salesforce 8, Zoom 7, Figma 4, Mailchimp 1, HubSpot 0. The high-change vendors track with the steady-raisers hypothesis. The low-change count for Mailchimp and HubSpot underlines the stealth-restructure dynamic — the change happens in plan structure, not in the $N mentions a diff-tracker captures.

Data caveats journalists will probe

• Plan structure changes mean apples-to-apples is hard. HubSpot's 2020 Pro and 2026 Pro are not the same product. The closest equivalent tier is used; structural changes are disclosed.
• Volume discounts and enterprise pricing are not captured. All prices are list-tier published rates. Real customers — especially at the enterprise tier — pay negotiated rates that vary widely.
• Currency rates affect non-US comparisons. All prices in this dataset are USD; non-US vendors may have raised in their local currencies on a different schedule.
• AI-feature surcharges are increasingly opaque. Notion AI, Asana AI, and Atlassian Intelligence added separate AI add-on charges in 2024-2025. Base-plan figures here exclude those add-ons; real total-cost-of-ownership has risen for customers who adopted them.
• The 10-vendor sample is not random. Selection favours name-recognisable English-language US/EU SaaS. A different sample — vertical SaaS, Asian SaaS, infrastructure-as-a-service — would produce different patterns. The findings should not be generalised to "all SaaS."

What are the alternatives to direct pricing-page archaeology?

There are four practical ways to track SaaS pricing changes over time. Each has a different cost structure and different blind spots.

Pricing and feature data based on publicly available information as of May 2026 and may change.

For multi-vendor longitudinal pricing research, the CDX API approach is the only one that scales without a six-figure tooling budget. The browser UI is fine for a single vendor; pricing-history publications are convenient but editorial; manual checks miss everything except the sticker. The trade-offs depend on how many vendors and how many reference dates the analysis covers.

Common misconceptions

"SaaS prices are rising in line with inflation." The 10-vendor sample shows wide dispersion. Slack, Zoom, Notion, and Figma fell below US CPI's ~22% rise. Salesforce and Adobe exceeded it on the sticker. The stealth restructurers exceeded it on the effective customer cost while the sticker fell below it. There is no single "SaaS inflation rate" — only a sticker rate, a structure-adjusted rate, and the divergence between them.

"Stealth restructuring means the vendor is hiding something." The pricing-strategy literature treats unit-of-sale changes as a normal part of mature SaaS pricing. Vendors disclose the changes in renewal communications, support docs, and earnings calls. The point of this analysis is that headline-price tracking systematically misses those changes — not that the changes are concealed.

"Figma must have raised prices when it bundled FigJam and Slides." It didn't. The 2020 Wayback snapshot shows $12, $15, $45. The 2026 live page shows $12, $15, $45. The bundling moved more product into the existing tier rather than shifting the price.

Limitations

• Three vendors are Wayback-verified at the price level: Figma, Asana, Zoom. The other seven 2020 baselines come from press coverage and pricing-history publications. Closing the remaining seven requires Wayback re-runs against alternate near-2020 dates — the 2020-06-01 target produced empty SPA shells (HubSpot, Atlassian), a navigation timeout (Mailchimp), a wrong-date snapshot (Salesforce — Wayback's nearest match was 2024), or no-snapshot-available errors (Slack, Notion, monday.com).
• The dataset is 10 vendors. A 25-vendor sample (adding Atlassian Jira, monday.com, Microsoft 365, Google Workspace, Stripe, Webflow, Loom, Calendly, Typeform, ConvertKit, Klaviyo) would be the canonical reference benchmark. Not in scope for this run.
• AI add-on costs are excluded from base-plan figures. Notion AI, Asana AI, and Atlassian Intelligence added per-seat AI surcharges in 2024-2025 that meaningfully raise total-cost-of-ownership for customers who adopted them.
• Effective-cost figures depend on customer-specific usage. HubSpot's 51% increase assumes a 10-seat team. A 3-seat team on the same plan sees a +11% increase. Mailchimp's 33-53% range depends on contact volume. The ranges are illustrative, not single-customer guarantees.
• Enterprise-negotiated pricing is invisible to this analysis. Customers above the published tiers pay rates that none of these sources can see.

Key facts about the SaaS pricing time machine

• The sample tracks 10 SaaS vendors from 2020 to 2026.
• Half the sample raised sticker prices 7-33% over six years.
• Three vendors (Mailchimp, HubSpot, Intercom) restructured plans for an effective 30-50%+ hike.
• Two vendors (Notion, Figma) held entry-paid plan stickers flat for 5+ years.
• Figma's $12/editor Professional plan held for seven years (2018-2025).
• Figma is the only vendor in the sample with direct Wayback-snapshot verification at the price level.
• US CPI rose approximately 22% from January 2020 to May 2026.
• Intercom destroyed approximately $60M of legacy ARR to switch from per-seat to per-AI-resolution pricing.

Glossary

• Stealth restructuring — a pricing change in which the headline sticker is held or cut while the unit of sale (per-seat, per-contact, per-resolution) or plan structure changes in a way that raises effective customer cost.
• Steady raiser — a vendor that publishes its sticker increases on a roughly annual cadence, generally tracking inflation.
• Flat-pricer — a vendor that holds its entry-paid plan sticker price unchanged for five or more years, often as a deliberate market-share strategy.
• Effective customer cost — the total amount a customer actually pays under the new structure, including additional-seat charges, per-contact billing on unsubscribers, AI resolution charges, and other usage-based add-ons. Distinct from the headline sticker.
• Sticker price — the headline number displayed on the vendor's public pricing page for the entry-paid plan.
• CDX API — the Internet Archive's free, open endpoint for querying its index of archived web snapshots. The basis for programmatic Wayback Machine research.

Why pricing archaeology applies beyond SaaS

The patterns in this dataset apply beyond B2B SaaS to any subscription business with a public pricing page and a renewal cycle. Five universal principles:

• Sticker-price tracking systematically understates real cost growth in any market where vendors can change the unit of sale without changing the headline number.
• Plan-structure changes are pricing decisions even when the affected price points stay the same.
• Free-tier reductions are price increases for users who depended on the free tier.
• Usage-based add-ons compound with sticker stability to create the largest gap between headline and effective cost.
• Holding the sticker flat is a strategic choice, not an absence of pricing strategy.

Streaming services, telecoms, cloud infrastructure, consumer subscription apps, and B2B utility software all show the same dynamics. The dataset here is SaaS-specific; the methodology generalises.

When you need this analysis

• You're briefing procurement on multi-year SaaS renewal-cycle inflation
• You're a SaaS journalist fact-checking an "AI is raising SaaS prices" narrative
• You're a CFO modelling cost-of-software growth into a 2027-2030 budget
• You're a pricing strategist benchmarking your own price changes against the field
• You're a board member asking "is our software spend rising faster than the market"

You probably don't need this if:

• You're tracking a single vendor on a single plan (use the vendor's pricing page directly)
• You're shopping for the cheapest entry-tier plan (sticker price is fine)
• You only care about list price, not effective cost

Press lift-out for journalists

Copy-paste-ready paragraph for citation:

A 2026 ApifyForge analysis of pricing-page changes at 10 major SaaS vendors
since 2020 found three distinct strategies: steady sticker-price increases
of 7-33% (Salesforce, Adobe, Asana), stealth restructures that effectively
raised customer cost by 30-50%+ without raising headline prices (Mailchimp,
HubSpot, Intercom), and flat-pricers that held sticker prices steady for 5+
years (Notion, Figma). Figma kept its Professional plan at $12 per editor
per month for seven straight years — 2018 through 2025 — making it the
SaaS pricing anomaly of the decade.

For follow-ups, contact ApifyForge through the contact page on apifyforge.com. The underlying Wayback snapshots are reproducible against the Internet Archive CDX API using the parameters disclosed in the methodology section.

Frequently asked questions

Is the 30-50% effective hike figure verified?

Partially. The HubSpot 51% figure is computed from publicly listed prices: the base $890/month plus seven additional seats at $45/month, applied to a 10-seat team that previously paid the bundled $800 single-seat-era rate. The Mailchimp 33-53% range is computed against the 2020 single-tier model versus the current four-tier-plus-unsubscribed-billing structure. The Intercom restructure is publicly described by Intercom itself, with the ~$60M ARR-destruction figure cited from the Mostly Metrics interview. Customer-specific effective cost varies with usage, team size, and contact volume.

How is this different from existing SaaS pricing trackers?

Most pricing trackers report sticker prices over time. They miss stealth restructuring entirely. The contribution of this analysis is the archetype framing — separating steady raisers from stealth restructurers from flat-pricers — and the disclosure that headline-only tracking systematically understates effective cost growth. Three vendors' 2020 prices in this dataset are directly Wayback-verified at the price level (Figma, Asana, Zoom); the other seven 2020 baselines come from press coverage.

Why isn't Atlassian / Microsoft / Google in the sample?

The 10-vendor sample is curated, not exhaustive. Top candidates for a 25-vendor expansion: Atlassian Jira (acknowledged October 2020 increase), monday.com Standard tier, Microsoft 365 Business Standard, Google Workspace Business Starter, Stripe payment processing rates, Webflow CMS, Loom Business, Calendly Standard, Typeform Plus, ConvertKit Creator Pro, Klaviyo. The 25-vendor expansion is in scope for a follow-up dataset.

Did Figma really hold $12 for seven years?

Yes, and it is the only price point in this dataset directly verified against Wayback Machine snapshots. The Figma pricing page from April, July, and September 2020 explicitly mentions $12, $15, $45 — the same three numbers on the live 2026 pricing page. The seven-year window covers 2018 through 2025; the actual Wayback snapshots verified here are from 2020 through 2026.

What's the relationship between sticker prices and US inflation?

US CPI rose approximately 22% from January 2020 to May 2026. Salesforce (+33%), Adobe (+27%), and the stealth restructurers' effective-cost increases all exceeded that benchmark. Asana (+10%), Slack (+9%), Zoom (+7%), Notion (0%), and Figma (0%) all fell below it. There is no single "SaaS inflation rate" in this sample — there is a wide dispersion that depends on archetype.

Can I reproduce this analysis on my own SaaS vendor list?

Yes. The methodology is reproducible against any URL the Wayback Machine has snapshots for. The Internet Archive's CDX API is free and open. The wayback-machine-search Apify actor is one way to query it programmatically and return structured snapshot timelines; an earlier ApifyForge post on querying the CDX API walks through the parameter shape. The HTML-parsing-by-reference-date follow-up step is the part that takes the most engineering work and is where the next version of this dataset is heading.

How is this relevant to ApifyForge readers?

ApifyForge tracks the underlying-data side of SaaS — pricing-page archaeology, web-archive research, and programmatic historical-data extraction are recurring topics on the blog. This post is the first multi-vendor longitudinal study published here. Future analyses are likely to expand the sample, add the AI-add-on-cost column, and re-derive every 2020 baseline from a Wayback snapshot at a fixed reference date.

Ryan Clinton publishes Apify actors and MCP servers as ryanclinton and writes about pricing archaeology, web-archive research, and developer tooling at ApifyForge.

Last updated: May 2026

This guide focuses on B2B SaaS, but the same pricing-archaeology patterns apply broadly to any subscription business with a public pricing page and a renewal cycle.