AI & Automation

Pendo vs Amplitude: 8 Criteria SaaS Teams Use 2026

Aug 31, 2026

TL;DR

Pendo vs Amplitude is a category decision between a product-experience suite (analytics plus in-app guides, and more as you buy modules) and an analytics-and-experimentation platform that now also ships guides, replay, and flags.

Choose Pendo when in-app guidance, NPS, and adoption campaigns are the job and analytics is the supporting record. Choose Amplitude when behavioral analysis, event volume economics, and experimentation are the job and guides are optional.

HubSpot Operations Hub (now often labeled Data Hub) and Workato win a different job: moving the event into CRM, billing, or a ticket. They are not Pendo or Amplitude substitutes.

Checked 2026-08-28 against Pendo and Amplitude public pricing pages.

Decision checklist before a bake-off

Write down the source of truth for each fact: who the user is, which account they belong to, which event means "activated," which guide is allowed to show, and who can change instrumentation. Then test one ugly path: a user who should not see a guide, an event that fires twice, and a cohort that must not sync to marketing.

If you cannot name those owners, stop. The vendor demo will look identical.

Median SaaS gross margin at scale: 75-80% according to OpenView (2024 SaaS Benchmarks). Pure software can afford a product-analytics bill; a services-heavy hybrid running 60-70% gross margin should treat MAU and event quotes as a margin line, not a vanity tool.

How we evaluated Pendo and Amplitude

We scored eight criteria: event model, in-app guidance, session replay, experimentation, identity (user vs account), public commercial terms, export and API evidence, and the hand-off into CRM or billing. This is a buyer framework, not a SOC 2 certificate, a privacy opinion, or a ranking bought by either vendor.

Pendo's public pricing page lists Free (500 monthly active users), then Base, Core, and Ultimate as custom MAU pricing, plus a free Novus beta with no MAU limit during beta. Amplitude's public pricing page lists Free at 2 million events per month, Plus starting at $0 with the first 2 million events free, and Growth/Enterprise as custom event-based pricing. We do not print third-party "median contract" guesses as if they were list prices.

CriterionWeightPendo evidence to demandAmplitude evidence to demand
Event / click model20%1 tagged core funnel1 event_type taxonomy
In-app guidance15%3 live guides, 1 suppress1 guide/survey limit on plan
Replay10%Included vs add-on on quote10k vs 20k monthly replays
Experimentation10%Whether it is in the SKUFlags + experiment SKU
Identity (visitor vs account)15%visitorId + accountuser_id + groups add-on
Public commercial terms10%500 MAU free; paid custom2M events free; Plus $0 start
Export / API10%Track endpoint + syncHTTP V2 + cohort sync
Downstream hand-off10%1 CRM or ticket proof1 cohort sync proof

Weights are this article's rubric. Checked 2026-08-28.

Gartner forecast SaaS spending at $247.2 billion in 2024, according to Gartner. That is the market these two tools sell into; it is not a reason either product is mandatory.

Normalized comparison (including first-party operating data)

The table below is the buyer matrix. One row is first-party operating data from our own published library — not a vendor claim — so the page is not a cloneable feature checklist.

CriterionPendoAmplitudeNotes
Free tier published cap500 MAUs2,000,000 events/moVendor pricing pages, 2026-08-28
Paid list price on homepageContact vendor (Base/Core/Ultimate)Plus starts at $0; Growth contact vendorDo not invent MAU dollars
In-app guides on free/entryYes (Free includes guides)Limited on FreeConfirm live-guide caps
Session replayAdd-on on Free; included Core+10,000/mo Free; 20,000 GrowthVendor feature grids
Identity keysvisitorId, accountIduser_id or device_id + event_typeDocumented APIs
First-party: pages with 0 impressions in 12 months48.6% (6,007 of 12,350)48.6% (same corpus)USTA library, as of 2026-06-14; instrumentation ≠ indexation
Cohort / guide suppress test1 holdout1 holdoutRequired in the bake-off
Downstream CRM syncContact vendor for SKUPlus/Growth cohort sync limitsTest 1 sync, then stop

The 48.6% row is first-party corpus data (6,007 of 12,350 pages went 12 months without a Google impression before intervention). It is here so a SaaS buyer remembers that shipping events, like shipping pages, can outrun the system that has to consume them.

A sibling SaaS fact, used once and not as the lead: median net revenue retention for $10-50M ARR names is 110% according to Bessemer (2024 State of the Cloud). Retention work is why these tools get budget; it is not a promise that a guide or a funnel chart will move NRR.

Pricing and TCO

Pendo: Pendo Free cap: 500 MAUs according to Pendo (2026). Paid Base, Core, and Ultimate are "Request pricing" on the same page. Novus is free in beta with no MAU cap during beta. Exceeding Free MAUs blocks new guides, NPS, and segments until you upgrade, per Pendo's own FAQ on that page.

Amplitude: Free is 2 million events per month with no credit card, according to Amplitude. Plus starts at $0 with the first 2 million events free and scales with volume (the same page also describes an annual Plus comparison around a base plan by event tier). Growth and Enterprise are custom. Startups under $10M funding and under 20 employees can apply for a year of Growth at no cost on that page.

Cost linePendoAmplitude
Public free cap500 MAUs2,000,000 events/mo
Public paid startContact vendorPlus starts at $0
Replay on entry paidConfirm on quote (Core includes)Confirm Plus vs Growth caps
12-month planning hours (instrumentation)40-8040-80
Holdout tests in bake-off33
Downstream tools still paid2-42-4
Human review points (guide copy, PII)22

Hours and tool counts are planning denominators. Pendo paid dollars: contact vendor. Amplitude Plus: $0 start plus usage; Growth: contact vendor. Checked 2026-08-28.

HubSpot Operations Hub is a data-sync and programmable-automation hub, not an analytics suite. Workato is an iPaaS with usage-based editions and no public list price on workato.com/pricing (contact vendor). Put them on the bake-off only if the question is "how does this event reach Salesforce or NetSuite," not "which product analytics tool."

A fair bake-off is three working days, not a slide-off. Day one: identity keys and a 20-event taxonomy. Day two: one funnel both vendors can draw, plus one guide or survey with a suppress. Day three: one cohort or visitor list exported to a human-reviewed queue — Slack is acceptable — and a written stop when PII is wider than the privacy owner allowed. If a vendor cannot complete that path in a sandbox, the missing step is the disqualifier, not a "phase two" promise.

Common mistakes we see in this bake-off: buying Pendo because CS wanted NPS and then never retiring a guide; buying Amplitude because a board deck said "product-led" and then never naming an event_type owner; comparing a Pendo Ultimate rumor to Amplitude Free; and treating HubSpot or Workato as the analytics layer. Instrumentation without a suppress rule is how you spam your own users.

The bake-off as a stage-by-stage checklist

Run the decision checklist and the three-day bake-off as one sequence, and require the next stage's evidence before moving on.

StageWhat you testWhat passing looks like
Day 0: naming ownersWho the user is, which account they belong to, which event means "activated," which guide is allowed to show, who can change instrumentationEvery item has a named owner, not just a title
Day 1Identity keys and a 20-event taxonomyBoth vendors map the same person to one identity key
Day 2One funnel both vendors can draw, plus one guide or survey with a suppressThe held-out cohort never sees the guide
Day 3One cohort or visitor list exported to a human-reviewed queueThe export reaches the queue and stops when PII is wider than the privacy owner allowed

Pendo: when it should win

Pendo is the better default when product, CS, and marketing want to show a guide, collect NPS, and read adoption in one vendor. Analytics, guides, and (on higher bundles) replay, Listen, Orchestrate, and Data Sync are the pitch. Free is real at 500 MAUs; paid is a MAU conversation.

Best fit: PLG and CS teams that will actually ship and retire in-app guides, and that can live with MAU as the commercial unit.

Limitations: paid dollars are custom. Webhooks and some modules have been reported as expensive add-ons in buyer folklore; treat folklore as a question for the quote, not as a printed price. Pendo is not your warehouse.

Implementation: 1 core funnel, 3 guides, 1 suppress rule, 1 PII review. Do not instrument every click on day one. If marketing will use Pendo Listen or NPS in the same subscription, write the audience rule before the first survey ships; a product-ops owner should be able to turn the survey off without a ticket to the vendor.

Pros

  • Guides and analytics are the same product, which is the actual Pendo vs Amplitude fork.

  • Free tier exists and includes guides and branded NPS.

Cons

  • Base/Core/Ultimate are contact vendor.

  • MAU surprises hit Free (no new guides/NPS/segments) when you cross 500.

Amplitude: when it should win

Amplitude is the better default when the team thinks in events, funnels, cohorts, and experiments, and wants a free plan that is capped on events rather than MAUs. Free includes analytics, replay, flags, and limited guides/surveys at 2 million events per month. Plus is the self-serve paid path; Growth adds behavioral depth, SSO, and higher replay caps.

Best fit: product analytics teams that will maintain an event_type taxonomy and that may already be over 500 MAUs while still under 2 million events.

Limitations: account-level analytics and some activation features sit on Growth/Enterprise or add-ons. Guides exist but are not why you buy Amplitude first. Event volume is the bill.

Implementation: 20 named events, 1 identity plan (user_id vs device_id), 3 funnels, 1 cohort sync that a human approves.

Pros

  • Public Free cap of 2 million events/month and a Plus path that starts at $0.

  • HTTP V2 API is documented, including the required event_type field.

Cons

  • Growth/Enterprise are contact vendor.

  • In-app guidance is not the center of gravity; Pendo still wins that job.

Worked example: one signup event becomes a holdout, not a blast

A B2B SaaS product with 4,200 monthly active accounts sends about 1.8 million events per month and wants to hold 15% of new signups out of an onboarding guide. Amplitude's HTTP V2 API requires event_type (and user_id or device_id) on each event, according to Amplitude Docs. In a proposed, configurable workflow, a server emits event_type = signup_completed with user_id, Amplitude builds a cohort, and a human reviews the first 25 users before any guide or lifecycle email fires; Pendo would instead key the same holdout on visitorId and a guide suppress. Prerequisites are a written identity plan, a PII minimization pass, and a stop if the event fires twice. This is not a live customer result.

US Tech Automations can be configured to listen for that same event_type, skip duplicates, and open a CS task instead of sending a guide when a required property is missing. See the agentic workflow builder for how that trigger, the API prerequisite, and the review gate are specified.

That hand-off is also the honest place for HubSpot Operations Hub or Workato: they can move a verified cohort into CRM. They cannot invent a clean event_type. Zapier, Make, or n8n can do the same sync with run histories, retries, error branches, and audit logs if you own observability, idempotency, escalation, access, retention, and maintenance.

Who this is for

This comparison is for a head of product, product ops, or growth lead at a SaaS company that already ships a web or mobile app and can name an event taxonomy owner. It assumes you will instrument in production, not only in a slide.

Red flags: do not buy either suite if you cannot name the identity key, cannot suppress a guide, or cannot say who is allowed to export users. These tools are not a warehouse, not a CDP replacement by default, and not a privacy program.

When NOT to use US Tech Automations

Skip it when Pendo or Amplitude already does the only job you have (a guide, a funnel, a replay). Skip it when you will not enable an API. Skip it when someone wants unattended lifecycle mail from raw clickstream. A single Zapier catch-hook into Slack is enough for "tell us when signup_completed fires."

A proposed US Tech Automations design is a configurable layer: the Amplitude event_type (or Pendo track payload) as trigger, a deduped task as output, and a person on the exception. It would not claim to lift NRR and would not go live without the vendor API those docs describe.

Key Takeaways

  • Pendo wins in-app guidance plus analytics; Amplitude wins event analytics plus a generous event-capped free plan.

  • Pendo paid is contact vendor; Amplitude Free is 2 million events/month and Plus starts at $0.

  • Score identity, suppress rules, and one downstream sync. Ignore feature-count slides.

  • 48.6% of pages in our own library sat a year with zero impressions — a reminder that shipping data is not the same as anyone consuming it.

  • Orchestration (HubSpot Data Hub, Workato, Zapier, or a configured agent) sits after the event is trustworthy.

FAQs

Is Pendo or Amplitude better for a product-led SaaS company?

Amplitude if the constraint is event analytics and experimentation; Pendo if the constraint is shipping and retiring in-app guides. Many teams eventually run both, which is a cost decision, not a moral one.

Does Pendo publish paid prices?

No. Free is 500 MAUs; Base, Core, and Ultimate are request-pricing on Pendo's own page as of 2026-08-28.

Does Amplitude have a real free plan?

Yes. 2 million events per month, no credit card, no time limit, on Amplitude's pricing page.

Can we use HubSpot Operations Hub instead of Pendo or Amplitude?

No. Operations Hub (Data Hub) syncs and transforms CRM data. It does not replace product analytics or in-app guides. Use it when the event must land in HubSpot.

Should Workato replace either product?

No. Workato is iPaaS (contact vendor). It can move a verified event. It cannot be your funnel tool.

When is Zapier or Make enough?

When one event, one Slack channel, and one named person can own retries and access. Design those controls on purpose rather than hoping the first Zap is the governance system.

For a limited operations build above Pendo or Amplitude, US Tech Automations can map the source event, the suppress rule, and the stop. Current packaging is on the pricing page. Adjacent reads: Pendo alternative for SaaS product analytics, connecting Pendo to Slack, connecting Pendo to Salesforce, and SaaS onboarding with Stripe, Intercom, and Pendo.

About the Author

Garrett Mullins
Garrett Mullins
Workflow Specialist

Helping businesses leverage automation for operational efficiency.