Koala Found the Buyers. Then It Found the Harder Problem.

Imagine the most promising customer in your territory has spent the morning reading your technical documentation. No form filled. No demo requested. To your sales team, this person may as well be invisible. Koala's original proposition was that invisibility was an expensive administrative error. It watched activity across a company's website, docs and product, connected that activity to people or accounts where it could, and told a rep when the interest looked worth a conversation.

That sounds obvious now, but it rearranges the seller's day. A cold list says, these companies fit the profile. An intent signal says, this company may be thinking about the problem today. The distinction is the difference between knocking on every door on a street and noticing a light in one window.

The clue that looked like a customer

Koala was founded by Tido Carriero, Netto Farah and Matthew Shwery. Their early pitch grew from a simple observation made with fast-growing software companies: sellers were searching cold account lists while people from potentially useful accounts were already browsing the companies' own sites, blogs, documentation and apps. Koala became a customer-intent platform for making those visits legible to the people who could respond.

The plumbing mattered. Koala described a business-to-business data layer that tracked page views and active session time, joined anonymous and known activity across devices, and organized it around visitors and accounts. If somebody had supplied an email address, the system could recognize that person. If not, its documentation said it used Clearbit Reveal to infer a company from an IP address, presenting an anonymous visitor from that company. That is a useful lead, though it is not the same thing as knowing the individual at the keyboard.

Once data arrived, a team could decide which behavior merited an alert: a pricing page, an enterprise feature in the docs, a long case-study read or a product event. Koala could send the result into Slack, update HubSpot or Salesforce, route a lead, or enroll a prospect in an outbound sequence. Its place in the market sat between web analytics, enrichment, CRM and sales engagement. It tried to make those parts behave like one conversation rather than four browser tabs.

The signup that lied

Retool offers an unusually sharp example of the problem Koala set out to solve. According to Koala's published case study, Retool already notified sales when somebody visited its pricing page. The notification lacked useful identity and context. After trying Koala with five business development reps, Retool expanded access to 75 people across its go-to-market team. Koala's account of the engagement says Retool's net-new pipeline rose 25% in less than a year, with dozens of deals won or accelerated. That is a customer case-study result, not a controlled experiment, but the underlying workflow is easy to inspect.

Retool's team found that guides to broad company rollout, documentation on single sign-on and self-hosting, and integration pages could be stronger clues than a generic signup. In the case study, new signups made up 85% of the leads flagged by a previous tool, yet those signups converted to sales-qualified opportunities at half the rate of an average homepage visitor. Many had simply chosen the self-service route. A tidy database field, it turned out, could be a poor substitute for understanding what the buyer was trying to do.

This explains the difference between Koala and a plain visitor-identification tool. It was not enough to reveal a company logo beside a page view. Koala argued that a sales team needed a way to judge the behavior, relate it to its own ideal-customer profile and existing accounts, and pass it to the right person while the timing still mattered. For a business with plenty of anonymous traffic and a considered purchase, that could change the order of a rep's calls. For one with little site or product activity, the same machinery would have far less to work with.

The dashboard became the problem

Koala also learned an uncomfortable lesson from its own users. The company said it had imagined an “analytics playground” in which reps would explore signals and find opportunities. Users wanted help deciding which paths deserved exploration. As Koala added more data sources, including product and community activity, that complaint became harder to ignore: richer context could also mean a more crowded morning.

“Users really wanted to know which paths were even worth exploring.”Koala's account of the move to Koala 2.0

Its answer, Koala 2.0, had three named pieces. Plays let a team define when a lead should enter a workflow, using triggers, CRM criteria and AI-assisted qualification. Magic Inbox grouped those leads by play so a rep could work a batch with the right context instead of jumping between unrelated alerts. Coach researched a person or account, suggested a next step and drafted a message. Koala claimed work that once took 20 to 30 minutes of manual research could take about 30 seconds in Coach. The claim came from the company, but the design change is concrete: the product moved its value from showing the evidence to helping a seller act on it.

Sanity's sales team supplied another glimpse of the routine. In a Koala interview, SDR Jordan Crosby described scanning activity from accounts in his region, then reaching out after several stakeholders from a prospect showed sustained interest. That prospect was weighing Sanity against an incumbent CMS near renewal. The timing made the outreach useful. The signal did not write the deal; it helped a person notice when a conversation might be welcome.

The price of knowing, and the price of ending

Koala sold that workflow as subscription software. Its published monthly plans offered two seats and 100 credits for free; Starter listed two seats and 1,000 credits for $200; Growth listed three seats and 5,000 credits for $1,000. Business was custom-priced. The paid plans opened more automation, retention and support, and larger teams could pay for additional seats. These are historical list prices: Koala is no longer a product a new customer can adopt.

The company was not a tiny experiment by its final year. Its customer page claimed thousands of companies and reps, $400 million in accelerated pipeline and a billion intent signals per year. Those are Koala's own figures. Public customers included Retool, Sanity, Mintlify, Vercel and Statsig. In February 2025, it raised a $15 million Series A led by CRV, with Recall Capital, Afore Capital and HubSpot Ventures participating.

Five months later came the odd ending. Cursor's maker, Anysphere, acquired Koala. TechCrunch reported that several of Koala's engineers would join a team focused on making Cursor ready for enterprise customers, while Cursor would not integrate Koala's core CRM product. Koala said it would shut down by September 30, 2025. Tido Carriero pointed customers toward Common Room, which worked with Koala on their transition. The acquisition price was not disclosed.

The facts resist an easy moral. Koala had named customers, a funded balance sheet and a product that users described as part of their daily work. It also had a product that its buyer chose not to carry forward. The founder's public goodbye said this was not how the team had imagined the chapter ending. The result is a reminder that customer value, fundraising and a durable independent company are related, but they are not synonyms.

The part worth stealing

For a sales leader, the copyable idea is smaller than Koala's whole platform. Pick one behavior that plausibly signals a purchase: sustained reading of enterprise documentation, several visits from an account to an integration page, or product usage near a paid limit. Compare it with actual opportunities, then route a small number of alerts to named owners. Ask whether those alerts led to conversations, not merely whether a dashboard counted more visitors. Koala's own evolution suggests the next test: give the rep enough context to write a relevant note without sending them on a 20-minute research expedition.

That loop needs a certain kind of business. It depends on meaningful website or product activity, an identifiable account, and a team able to follow up thoughtfully. Anonymous traffic alone does not certify intent, and automated outreach based on a weak signal can make a company sound alarmingly well informed. Koala's best examples worked because the seller treated the data as a reason to ask a useful question, not proof that a stranger was ready to buy.

Koala's short life traced a path from seeing the buyer to understanding the work left to do. A warm visitor is a clue. A good play is a decision. The gap between them was the company.