Provide Support Skipped the AI Race to Keep Live Chat Simple

The chat bubble tucked in the corner of an online store is easy to ignore. Click it and someone answers - a person, usually, asking whether you found the size you wanted or why the checkout stalled. That small exchange is a whole software category, and one of its longest-running players is a company most of the internet has never heard of. Provide Support has been selling that chat bubble since 2003, from an office on Broadway in Lower Manhattan, without a splashy rebrand, an IPO, or a chatbot press tour.

The interesting thing about Provide Support is not what it built. Live chat is a solved problem; dozens of companies sell a version of it. The interesting thing is what it refused to do. It never split its features across tiers. It never metered conversations. It never bolted on an AI agent it could put on a billboard. For twenty-plus years it has sold roughly the same product to roughly the same kind of customer, and roughly 10,000 of them are still paying.

To understand why that is a choice and not an accident, it helps to look at the company standing on the other end of the category.

The LivePerson mirror

LivePerson claims to have invented online chat for brands back in 1995. It went public, landed on the NASDAQ as LPSN, and spent the last decade repositioning itself as an enterprise conversational-AI platform. Its pitch today involves a "Conversational Cloud," Facebook Messenger and WhatsApp integrations, bot automation trained on years of chat logs, and an acquisition list that includes Conversable, a bot platform once used by McDonald's and Whole Foods. LivePerson counts a large share of the Fortune 500 among its customers - telcos, airlines, banks, retailers.

That is one way to run a live chat company: go up-market, go public, and keep reinventing the product until it looks nothing like a chat window. It is a legitimate strategy, and it has kept LivePerson relevant across three eras of the internet.

Both companies are still here. Only one of them changed its name for the moment.

Provide Support took the other road. Instead of climbing toward the enterprise, it planted itself in the middle of the market - the online store, the university help desk, the mid-sized publisher - and stayed. Where LivePerson's story is about layers of new technology, Provide Support's is about keeping a promise: connect a website visitor to a real person, in real time, without making the buyer think too hard about pricing.

The pricing page is the pitch

Most SaaS companies treat the pricing page as a maze. Features get locked behind higher tiers. Seats get metered. Usage creeps into overage charges. The whole design nudges you toward the plan above the one you wanted. Provide Support does almost none of that, and the absence is the strategy.

Every plan includes every feature. The only decision a buyer makes is how many agents they need - one, three, ten, or an enterprise arrangement above that. Chats are unlimited. You can put the chat button on as many domains as you own without paying more. Plans start around $8.25 a month, there is a 30-day money-back guarantee, and if you want to pay in Bitcoin or by Alipay or wire transfer, you can.

It sounds almost naive next to the usage-based playbook everyone else runs. But a three-agent store does not need a conversational cloud. It needs a chat window that loads, an operator console that does not drop messages, and a bill it can predict. Provide Support figured out early who it was actually serving and priced for that person instead of the whale two segments up.

What you can actually do with it

For a business, the practical value is straightforward. You drop a snippet of code onto your site - or install one of the plugins for WordPress, Shopify, WooCommerce, Magento, Joomla, Wix and a couple dozen other platforms - and a chat button appears. When a visitor clicks it, one of your agents picks up in a console that runs on Windows, macOS, Linux, iOS, Android, or straight in a browser.

From there the feature set is the sort of thing that has accumulated over twenty years of listening to customers: real-time visitor monitoring, so you can see who is on the site and start a proactive chat before they bounce; pre-chat surveys and customizable windows; automatic translation, so an agent and a visitor can talk past a language barrier; chat statistics and transcripts; IP blocking and encryption for the security-minded; and an "eye-catcher" bubble to nudge the button into view. None of it is exotic. All of it is the stuff a support team actually reaches for on a Tuesday.

The best demo Provide Support runs is its own homepage - it sells live chat by answering you on live chat.

That detail is worth sitting with. The company staffs the same widget it licenses. If you have a question about the product, you ask it the way a customer of one of its customers would - through a chat bubble, to a person. It is a small act of eating your own cooking, and it doubles as the only advertisement the product really needs.

A category that keeps getting declared dead

Live chat has been written off on a roughly annual schedule. Social media would replace it. Then messaging apps. Then chatbots. Then, most recently, large language models that promised to answer every support question without a human in the loop. Each wave sent a round of "chat is dead" think-pieces into circulation.

Provide Support's answer to all of it was to keep a person on the other end and keep the software working. That is not a rejection of AI so much as a read on its own customers. The lesson buried in the company's longevity is not "avoid new technology." It is "know exactly who you serve." A mid-market store owner wants the chat answered well, not automated away.

Why boring lasts

There is a quiet category of software companies that never trend, never IPO, and never die. They pick a problem, solve it, and then have the discipline to stop. Provide Support belongs to that group. Its client list - Koss, Unilever, Nutricia, Air France/KLM, BullionVault, Florida State University, Princeton - is the kind of roster you build one renewal at a time, not one launch at a time.

The opposite of hype, in software, is a renewal. Provide Support has collected two decades of them by doing one thing and announcing almost nothing. It is older than the iPhone and older than most of the SaaS chat tools that came and went while it kept the button working. For founders drowning in pressure to pivot every eighteen months, the company is a useful counter-example: you can build something durable by refusing to complicate it.

Whether that holds for another twenty years is a fair question. AI-driven support is real, and enterprise buyers clearly want it - that is the ground LivePerson has staked out. But the small store, the campus help desk, the publisher who wants a human to answer a reader's question, is not going away either. As long as that customer exists, there is room for a company whose entire pitch is: click the bubble, a person answers.

Reporting by YesPress Newsroom · Sources: providesupport.com, LinkedIn, Capterra, liveperson.com