Short answer: if you want a published self-serve price, the 2026 user onboarding shortlist is UserGuiding (from $174/Mo billed yearly), Chameleon (from $279 for 2,000 MTUs) and Userpilot ($299/mo). WalkMe, Whatfix and Pendo are quote-only above a tiny free tier. But the number you need to check is not the price. It is the gap between the MAU range a plan advertises and the MAU allowance the price actually covers, and in this category that gap runs as high as 50x. And no tool here can tell you why users quit, which is what Koji is for.
The 2026 shortlist at a glance
| Tool | Entry price | Meter | What the meter counts |
|---|---|---|---|
| UserGuiding | From $174/Mo, billed yearly | MAU | Price covers up to 2,000 MAU |
| Chameleon | Starts at $279 for 2,000 MTUs | MTU | Profiles you send to Chameleon |
| Userpilot | $299/mo | MAU | Includes 2k MAU on Starter |
| Appcues | Not published | MAU | Everyone who signs in to your product |
| Pendo | Free to 500 MAU, then custom | MAU | Custom MAU volume |
| Whatfix | Not published | User licenses | MAUs or all provisioned users |
| WalkMe | Not published | Not stated | Third-party contract data only |
| Userflow | Could not verify | MAU | Pricing page returned HTTP 403 |
| Intercom | Not comparable | Seats plus usage | Tours bundled into messaging |
| Koji | EUR 29/mo, free to start | Credits | Only interviews scoring 3+ |
The finding: the advertised MAU range is not the MAU you get
Every vendor in this category meters monthly active users. Three of them publish a plan price and an MAU range in the same table, and in each case the range is dramatically wider than what the price includes.
Userpilot prints $299/mo for Starter, which includes 2k MAU against a printed range of 0-2,000. Those agree. Growth prints From $849/mo and includes 10k MAU, against a printed range of 0-100k. The advertised ceiling is ten times the included allowance.
UserGuiding is the starkest version. Starter is From $174/Mo and Growth is From $349/Mo, both billed yearly. The plan table has an MAU Range row reading 0-10K for Starter and 0-100K for Growth. Under each plan, the same page prints that the price covers up to 2,000 MAU and that above it you should book a demo. So Starter advertises a ceiling 5x the priced ceiling, and Growth advertises one 50x the priced ceiling. Both statements sit on the same page, under the same plan.
Chameleon does not have that contradiction, but it has a different one. Startup starts at $279 for 2,000 MTUs. Pro is from $750/mo for 5k MTUs billed monthly, and Growth is from $1,250/mo for 5k MTUs billed annually. Both Pro and Growth are quoted at the same 5,000 MTUs, so Growth costs 1.67x the price for identical capacity. The extra money buys feature unlocks, not headroom.
The practical consequence is the same in all three cases. Above roughly 2,000 to 10,000 monthly active users, almost every tool in this category becomes a book-a-demo product, including the ones that look self-serve. Budget on that basis.
Then check what the meter is counting
The unit matters more than the rate, and the definitions are not interchangeable.
Appcues meters monthly active users, and its own documentation defines an MAU as the total number of unique users who sign in to your platform in any given month, on a rolling 30-day basis, updated nightly. Read that again: it is not everyone who saw an Appcues tooltip. It is everyone who signed in to your product, whether Appcues ever spoke to them or not. Appcues also documents what happens when you exceed the plan: nothing breaks, nothing throttles, and you are charged a prorated upgrade fee to move to the correct MAU tier for the remaining months of the contract. A good growth month silently repriced the rest of your year.
Chameleon meters uniquely identified user profiles sent to Chameleon in the prior 30 days. If your integration identifies every authenticated user on page load, your bill tracks your logged-in user base no matter how narrowly you targeted the tours themselves.
Whatfix publishes no dollar figures, but it does publish the structure: a flat fee plus user license fees. The definition of a user license then forks. For customer-facing applications, a license equals the monthly active users with access. For employee-facing applications, a license equals the total number of users with access to the application. For an internal rollout, activity stops being the meter and provisioning becomes the meter, so you pay for people who never log in.
WalkMe publishes nothing, and the third-party data does not agree with itself. Vendr reports a median WalkMe contract of $39,000 per year across 45 deals handled, with a range from $14,400 to $197,118, while the same source describes the smallest deployment it recognises, 100 to 500 users on one or two applications, as costing $40,000 to $150,000 annually. The typical customer therefore spends less than the floor of the smallest deployment. Both figures cannot guide your budget, and we work through which to trust in the WalkMe alternatives breakdown.
Pendo prints one number, 500, which is the MAU ceiling on its free plan. Base, Core and Ultimate are all custom MAU volume, request pricing.
Userflow we cannot price for you honestly. userflow.com/pricing returned HTTP 403 Forbidden when we fetched it for this article, so rather than repeat an aggregator figure we have not read on the vendor page, we are telling you to check it directly.
Intercom is on this list because teams shortlist it, not because it is really a peer. Product tours ship as one part of a customer messaging suite rather than as a standalone onboarding tool, which makes a like-for-like MAU comparison misleading. See the Intercom alternatives breakdown for the trade-off.
What all ten have in common
Every tool above is an intervention engine. It shows a tooltip, a checklist, a tour or a survey, and it measures whether the step completed. That is genuinely useful, and if your activation problem is that users cannot find a feature, a tour will fix it.
But a completion rate is not a reason. When 40% of new accounts abandon setup, your onboarding tool tells you which step and how many. It cannot tell you that they stalled because the integration screen asked for a production API key before anyone had decided the product was worth adopting, or because the pricing page they opened in another tab answered a different question than they expected.
That gap is the expensive one, and it is getting more expensive. The Condens 2026 report on AI in research analysis, built on a survey of 332 research practitioners and published in May 2026, found the top pain point researchers named was that it takes too long to get from data to insights, at 47%, with stakeholders needing faster answers than they can provide close behind at 43%. In 63% of organizations, the most common way a stakeholder gets research is still to ask a researcher directly. Meanwhile Maze, across nearly 500 responses in its Future of User Research report, found the share of organizations where research is essential to all levels of business strategy nearly tripled in a single year, from 8% in 2025 to 22% in 2026.
So the question gets asked more often, by more senior people, and the answer still takes weeks.
Koji: the layer that explains the drop-off
Koji is not an onboarding tool and will not render a tooltip. It is an AI-native customer research platform that interviews the users who stalled and tells you why in their own words.
The workflow is the point. Your onboarding tool identifies the cohort that abandoned step three. Koji runs AI-moderated voice or text interviews with those people, in parallel and on demand, then codes every transcript and returns a themed report with quotes. No scheduling, no moderator, no transcription backlog, and no moderator bias, because the AI interviewer asks everyone the same questions and probes follow-ups the same way.
Koji also carries the quantitative half, so you do not need a second tool for the numbers. Structured questions support six types, which are open_ended, scale, single_choice, multiple_choice, ranking and yes_no, which means a satisfaction scale and an open probe into the reason behind the score live in one interview and one report. The mechanics are in the structured questions guide.
And the price is published. Koji starts free with 10 credits and no card. Insights is EUR 29 per month with 29 credits, Interviews is EUR 79 per month with 79 credits, extra credits are a flat EUR 1 each, and a text interview costs 1 credit against 3 for voice and 5 for a report refresh. A quality gate means only conversations scoring 3 or better consume a credit. There is no MAU meter, so your bill does not move because your product grew.
Where to go next
- User onboarding research covers how to interview new users about activation.
- Diagnosing onboarding drop-off with AI interviews is the step-by-step version.
- Product adoption guide is the wider playbook.
- Onboarding survey guide covers the quantitative instrument.
- B2B onboarding experience surveys handles the time-to-value case.
- Appcues alternatives, Chameleon alternatives, UserGuiding alternatives and Whatfix alternatives go deep on each meter.
Fix the step you cannot explain
Take the onboarding step with the worst completion rate and the weakest theory. Run a Koji study on the people who abandoned it. You will have themed reasons in hours instead of weeks, with no research expertise required, and your onboarding tool can then fix the right thing. Start free with 10 credits, no card needed.
Frequently asked questions
What are the best user onboarding tools in 2026?
Userpilot, UserGuiding and Chameleon are the strongest self-serve options because they publish a price. Appcues remains the most established mid-market choice, Pendo is best where analytics and guidance belong together, and WalkMe and Whatfix dominate large enterprise and employee-facing rollouts. Pair any of them with Koji, which interviews the users who dropped off and explains why.
How much do user onboarding tools cost in 2026?
UserGuiding starts From $174/Mo and UserGuiding Growth From $349/Mo, both billed yearly. Userpilot is $299/mo for Starter and From $849/mo for Growth. Chameleon starts at $279 for 2,000 MTUs, with Pro from $750/mo and Growth from $1,250/mo. Appcues, Pendo, Whatfix and WalkMe do not publish paid prices. Critically, the published prices usually cover only about 2,000 monthly active users even when the plan advertises a far higher range.
Why is my onboarding tool bill higher than the advertised plan price?
Because the advertised MAU range is not the MAU allowance the price includes. UserGuiding prints a 0-100K range on Growth while the same page says the price covers up to 2,000 MAU. Userpilot prints a 0-100k range on Growth while including 10k MAU. Appcues additionally charges a prorated upgrade fee to move you to the correct tier for the remaining months of your contract once you exceed the plan, so a single good growth month reprices the rest of the year.
What does MAU actually mean in onboarding pricing?
It depends on the vendor, and the differences are large. Appcues counts every unique user who signs in to your platform, whether or not they ever saw an Appcues message. Chameleon counts uniquely identified user profiles you send to it in the prior 30 days. Whatfix counts monthly active users for customer-facing apps but every provisioned user for employee-facing apps. Always get the definition in writing before signing.
Can user onboarding tools tell you why users drop off?
No. They measure whether a step was completed and can show you exactly where people stall, but the reason exists only in the account the user gives of it. Koji closes that gap by running AI-moderated voice and text interviews with the cohort that abandoned, then producing a themed report with quotes automatically, so a completion rate becomes a cause you can fix.
Do I need an onboarding tool and a research tool?
Usually yes, because they do different jobs. An onboarding tool changes the experience; a research tool explains it. Running tours without research means guessing which friction to remove, and guessing is expensive at enterprise MAU rates. Koji is priced in credits rather than monthly active users, so adding the research layer does not scale your bill with your user base.
