Every enterprise software rollout contains a moment of theater. Executives gather around a slide that says go-live. The system is switched on. The consultants leave. Then Susan in procurement cannot find the approval button, Amir in sales keeps a private spreadsheet, and a support queue fills with variations of “where do I click?” WalkMe built a company inside that anticlimax.
Its software sits above other software. It watches how people move through a workflow, spots where they stall, and places help in the same window: a tip beside a form, a step-by-step path through Salesforce, a prompt inside SAP, an automated action that removes five dull clicks. The administrator gets analytics; the employee gets a nudge; the chief information officer gets a chance to prove the expensive new system is earning its keep.
That sounds modest because the interface is modest. A balloon is not a moonshot. But the underlying problem is enormous and stubborn: corporations are good at acquiring technology and uneven at changing behavior. WalkMe says its platform now reaches about 35 million users in more than 42 countries, serving roughly 1,600 customers and 27 percent of the Fortune 500. The names include IBM, Deloitte, Accenture, Walgreens, BMO, KeyBank and Thermo Fisher Scientific. Most users are not shopping for WalkMe. Their employer is buying fewer mistakes, faster onboarding and less training debt.
The original trickMove the manual onto the screen
WalkMe was founded in 2011 by Dan Adika, Rafael Sweary, Eyal Cohen and Yuval Shalom Ozanna. Its first public product, launched in 2012, was described with less polish and more clarity: an interactive, on-screen user guidance system for the web. Instead of sending someone to a manual, a course or a support agent, it would guide the person through the task while the task was happening.
The company reached $1 million in annual recurring revenue in 2013 and kept widening the proposition. Mobile, desktop and analytics arrived. Smart Walk-Thrus turned processes into guided routes. ActionBot gathered information conversationally and kicked off actions. Discovery exposed unused applications and licenses. Surveys, segmentation, enterprise search, automation and governance made the product less like a tour builder and more like an operating layer for behavior.
This product loop is the company's real difference. Appcues and Userpilot are associated with customer onboarding inside a product. Pendo combines product analytics and in-app engagement. Whatfix is WalkMe's closest broad enterprise rival. Traditional learning systems teach away from the work; robotic process automation removes work from the person entirely. WalkMe tries to connect the middle: understand behavior across applications, intervene in context, automate what is safe, and report what improved.
“Applications, processes, data and people are the four key elements of a successful business transformation.”Christian Klein, SAP CEO
The customersSell the result, not the speech bubble
The clearest examples are pleasantly unromantic. KeyBank used WalkMe to guide commercial-banking customers through complex online tasks; targeted workflows reached self-service completion increases of 150 percent or more and a 90 percent self-service adoption rate. Deloitte says its internal deployment spans more than 90 applications and 165,000 end users, with support tickets down 30 percent. IBM reported a sixfold increase in retention and a 300 percent improvement in product adoption in its customer program.
Schneider Electric offers the more useful implementation story because it includes the messy middle. A three-person team began in North America in 2017, helping employees use Salesforce. Positive results earned an executive sponsor in 2021, then budget, people and access to global business units. The program spread into HR, procurement, SAP and Coupa, with regional centers of excellence. By 2026, Schneider reported 80 percent user engagement and a 60 percent decline in training-related support tickets.
What changed their mind? Evidence, then sponsorship. The team did not open with an enterprise-wide manifesto. It showed that one workflow could behave better, found a senior leader willing to turn local proof into a global program, and personalized guidance by role. That sequence is copyable. It also reveals why a DAP is rarely a plug-in-and-celebrate purchase: somebody must choose the workflows, author the interventions, maintain them when the underlying app changes and decide what success means.
The public resetWhat failed first was the growth story
WalkMe raised more than $300 million privately before listing on Nasdaq in June 2021. Its initial public offering priced at $31 a share, implying a valuation around $2.5 billion. Revenue kept rising, from $193.3 million in 2021 to $267 million in 2023, but the market stopped rewarding expansion at any cost. WalkMe cut 43 jobs in 2022 and another 112, about 10 percent of staff, in April 2023. Management said customers were buying more cautiously and the organization needed to match near-term growth.
One company, two market prices
The $14 offer was a 45% premium to the pre-announcement close, but less than half the 2021 IPO price.
The response was not a new consumer product or a fashionable rebrand. WalkMe went further toward big enterprises, partner-led services and efficiency. In 2023 it produced positive free cash flow for the full year and narrowed its non-GAAP operating loss sharply. By March 2024, 42 customers generated more than $1 million each in annual recurring revenue and 536 generated more than $100,000.
Then SAP arrived. It agreed to pay $14 a share in cash, approximately $1.5 billion in equity value, and completed the acquisition in September 2024. The deal cost public shareholders a clean comparison: the purchase price represented a 45 percent premium to the previous close, yet less than half the IPO price. It was both a strategic exit and a reminder that timing owns part of every valuation.
SAP paid roughly $1.5 billion. Customers receive custom quotes, not a public rate card. The honest budget includes licenses, implementation, content ownership, analytics and ongoing maintenance - not merely the overlay.
The SAP logicAn interface for AI that follows the work
Why would SAP want a layer designed to sit on top of SAP and everybody else? Because the layer is where a cross-application assistant can see the user. SAP has said WalkMe will help make Joule, its AI copilot, available inside both SAP and non-SAP applications. In that framing, WalkMe is no longer only the yellow brick road through a difficult form. It becomes the eyes and delivery surface for context, recommendations and actions wherever work is happening.
The product has moved accordingly. WalkMeX introduced contextual AI assistance. The 2026 Action Bar provides a persistent entry point for guidance and actions. Smart Highlights can pull useful information from another system into the screen at hand. A Solutions Gallery offers prebuilt AI workflow templates. The Q2 2026 release added contextual chat, more detailed governance, and a unified console for administrators.
WalkMe Learning Arc extends the same thesis from doing to learning. It uses AI to author courses, lessons and individual passages, then delivers simulations, quizzes, learning journeys and reinforcement inside applications. It is designed to complement a learning management system, not evict it. The distinction matters: formal training establishes a concept; WalkMe tries to catch the employee three Tuesdays later, when memory has left and the expense form has not.
“Learning should move, adapt, and flow seamlessly with the way we work.”Dan Adika, co-founder and CEO
The stealable bitsStart with one expensive hesitation
The obvious lesson is “put help in context.” The better lessons are operational. WalkMe turned a small visual intervention into enterprise software by wrapping it in measurement, governance, cross-application reach and a partner ecosystem. Founders and internal transformation teams can borrow the sequence without copying the product.
Choose one costly workflow with errors, abandonment or support volume. “Adoption” is fog; a failed requisition is evidence.
Record completion, time, errors and tickets first. Without a before state, every improvement becomes a testimonial.
Place guidance where confusion occurs. Do not make the user leave the work to learn how to do the work.
Use one local result to secure budget, ownership and access. Scale follows proof more reliably than enthusiasm.
The other steal is commercial. WalkMe sells to the person accountable for return on software, not only to the person who enjoys the interface. Its business model is enterprise SaaS with custom pricing, supplemented by implementation and managed services delivered with partners such as Deloitte, Accenture, KPMG and Capgemini. Technology partners including Salesforce, Oracle, Microsoft and Celonis widen where the product can act. The end user gets a prompt; the economic buyer gets reduced risk, fewer tickets, faster proficiency and usage data.
When the playbook does not work
- The underlying process is broken, contradictory or unnecessary.
- The team needs a light product tour for one simple app.
- No owner will maintain guidance after interfaces change.
- User volume is too small to justify enterprise deployment overhead.
- The buyer cannot baseline errors, completion, time or support demand.
An overlay can make a good process easier to follow. It cannot rescue a policy nobody understands, an approval chain nobody owns or a product that should have been redesigned. Too much guidance can also become wallpaper. The platform works best where the workflow is important, repeated, measurable and stable enough to maintain - yet complicated enough that timely context changes the outcome.
That boundary makes WalkMe more interesting, not less. The company did not eliminate confusing software. It made confusion observable, editable and billable. The first era was a speech bubble pointing to the next button. The next is an AI assistant that knows what is on the screen, what sits in another system and which action should come next. Same wedge, larger ambition: close the last mile between a technology purchase and a person doing useful work.