More than a chatbot - an AI agent that works your marketing funnel 24/7.
In 2009, four people who had been building a small piece of chat software as a side project made a decision that would look strange to almost every startup that came after them. They raised a little money - roughly a hundred thousand dollars from Y Combinator and a few angels - and then they mostly stopped. No Series A. No board full of investors nudging them toward a bigger, faster, riskier version of the plan. They just kept selling the product and kept the lights on.
That product is Olark, and the odd thing is that it worked. More than fifteen years later, the little chat box sits on the websites of over 12,000 paying businesses - technical colleges, sports academies, e-commerce shops, software startups - most of them too small to afford a fancy customer-support suite. Olark is legally still incorporated under the name it almost launched with, Habla, Inc. It is a certified B Corporation. It has no headquarters in the usual sense; the team of around 40 works from the US, Canada, Brazil and the UK. And for a category that venture capital eventually turned into a battlefield, Olark's story is unusually quiet.
At its simplest, Olark is live chat for websites. A business pastes a snippet of code onto its site, and a chat box appears in the corner. When a visitor types a question, someone on the other end - a support rep, a salesperson, the owner - answers in real time. Around that basic loop, Olark built the machinery that makes chat useful at scale: canned responses for the questions you get fifty times a day, chat transcripts you can search later, visitor tracking so you know who is on the pricing page right now, targeted triggers that start a conversation automatically, agent groups, routing, and reporting.
The pitch has always been ease over enterprise. You do not need an engineering team to run Olark. You do not need a six-week onboarding. That was the founding insight, and it is worth pausing on, because in 2009 real-time chat was mostly the province of big companies with big call centers. Olark made it something a two-person shop could switch on before lunch.
The seed was planted long before the company existed. Ben Congleton, Olark's CEO, started a web-hosting business as a high school freshman in 1998 and grew it to around $170,000 a year. Running support for hosting customers meant running chat - and the experience of doing it badly, with clunky tools, stuck with him. Years later he and Roland Osborne began building something better, folding in Matt Pizzimenti and Zach Steindler as the other two founders. The earliest code dates to 2007. By 2009 it was a real company, accepted into Y Combinator's Summer batch.
They very nearly called it Habla. When it came time to launch properly, they tried the stranger name - Olark - as an experiment, and it stuck. The legal entity never changed, which is why the fine print still reads Habla, Inc. dba Olark.
The growth story that followed is a small master class in pricing. In the early days the entry plan was five dollars a month. Revenue crept along until, in 2011, the founders did the uncomfortable thing and raised prices - the Bronze plan climbed toward seventeen dollars. Customer counts barely moved, but revenue took off. They had been undercharging the whole time.
Roughly four out of five Olark customers are small businesses, and the company never really tried to change that. You will find it on the sites of a state technical college fielding admissions questions, a sports academy handling camp registrations, a training center, an online store. Named customers include IMG Academy, Chippewa Valley Technical College, and HBA Learning Centres. The common thread is a team that wants to talk to the person on their website without hiring a call center to do it.
That focus cuts against the industry's gravity. Most chat companies, once they can, sprint toward enterprise - bigger contracts, bigger logos, bigger sales teams. Olark stayed pointed at the businesses that started with it, and that choice quietly shaped everything downstream: the self-serve pricing, the one-minute setup, the refusal to make the product something you need a consultant to install.
The newest chapter is the one that could have easily gone wrong. As AI reset expectations for what a chat box should do, Olark rebuilt its pitch around an AI agent - one that trains on a company's website in under a minute, with no code, then answers visitor questions from the live content of the site, engages based on behavior, qualifies leads, and, crucially, escalates to a human when it hits the edge of what it knows. The framing on the homepage is deliberately unfussy: more than a chatbot, an agent that works the marketing funnel around the clock.
It plugs into the tools small teams already run - HubSpot, Salesforce, ActiveCampaign, and custom webhooks - so a conversation that starts in the chat box lands in the CRM without anyone copy-pasting. The optional add-ons, which Olark calls PowerUps, layer on extra capability for teams that want it, priced roughly between $29 and $99 a month on top of the base plan.
The model is unglamorous and durable: subscription SaaS, billed per agent. The Standard plan runs around $29 a month per seat, dropping to the low twenties or high teens for annual and multi-year commitments, with PowerUps as the upsell. Because Olark never took a large venture round, revenue was never subsidized by someone else's money - it had to cover the company from early on. The founders have described the result as multi-millions in annual revenue, reached without the boom-and-bust arithmetic that governs most of the category.
| Approach | Olark | Typical VC-backed rival |
|---|---|---|
| Outside funding | ~$100K, then bootstrapped | Tens to hundreds of millions |
| Primary customer | Small business (~80%) | Mid-market & enterprise |
| Setup | Minutes, no engineer | Onboarding & implementation |
| Structure | Remote, B Corp | HQ-centric, growth-first |
The competitive field is crowded and well-funded: Intercom, Zendesk, Drift, LiveChat, Tidio, Crisp, Freshchat, Zoho SalesIQ and others all want the same chat box. Several are deeper if what you need is chat sitting next to a full ticketing system with macros and SLAs. Olark's answer has never been to out-feature them. It is to be the thing a small business can actually turn on, understand, and afford - and, increasingly, an AI agent that does the first shift before a human ever has to.
For a company that avoids noise, Olark had one very loud moment. In 2017, a developer named Madalyn Parker emailed her team to say she was taking two days for her mental health. Congleton wrote back to thank her - to say that using sick days this way should be standard practice, and that she was helping cut through the stigma. She posted his reply. It was retweeted more than ten thousand times and covered by CNBC, Money and Global News, and for a while Olark was less known for chat software than for how its CEO answered an email.
It fit a company that had already chosen to be remote before remote was fashionable, and to become a Public Benefit Corporation with a stated focus on web accessibility. None of that is marketing garnish on the product - it is roughly the same instinct that made the founders build chat for the businesses everyone else ignored.
Olark's place in the market is not the biggest chat company, or the flashiest, or the one with the largest logos. It is the one that proved a different route was possible: raise a little, charge fairly, stay small and remote, serve the customers at the bottom of the market, and last. In a category defined by well-funded giants trading blows over the enterprise, that is a quieter argument - and Olark has been making it, one chat box at a time, since 2009.