Why we're turning off our SaaS's live chat customer support

Senja is growing fast. Word of mouth is increasing, and one consistent reason is our high-quality customer support. Visitors, users and customers can hop on the live chat bot and speak to me or Wilson (the co-founders) in seconds. But, we've decided to turn off our live customer support and try switching to a ticket-based system. Here's why. #1. We want to cultivate a culture of self serve Right now, it's too convenient to reach out to us even if there's no reason to. We want people to fix their own issues. Our first goal is to improve the product UX and help center guides to reduce the need to speak to us, unless it's for a complex use case. We'd love to build a product so intuitive that any customer can get to success on their own, and still feel supported and confident in their actions. This is better for customers and better for us. #2. Our support is starting to suck Yes, customers repeatedly tell us they love our support - it comes up time and again in our testimonials. But as our user base has grown so has our number of enquiries - leading to a slower and inconsistent support experience. On one day you can have a 40-minute chat with a co-founder, and the next you won't hear back for 6 days. A message from a senja.io customer asking if there have been any updates on their questions. At the same time, attempts to time-block live chat responses are complex, because by the time you reply, the person is no longer available. #3. It sets the wrong expectation The majority of Senja users don't pay, and we have the most generous free tier in the market. We're proud of this, but live chat support for all users is a big promise that cultivates a culture of concierge support for free. We desperately need to move away from this. As well as this, when people see a live chat widget they expect instant answers and can become rude or demanding when they don't receive them. #4. There's so much noise Most users of the chatbot have a question they want answered and can clearly articulate it. But about 15% appear to simply want to have a chat, are curious, or fail to articulate their issue concisely. "Conversations" like this pop up more frequently than you might expect... A weird support conversation we had. We want to improve the quality of our support but that requires helping customers ask clearer, more specific questions. This is something we can facilitate with ticketed email support. #5. It kills productivity We're losing time improving our business and product to answer support questions. Role switching from deep work to 'fire fighting' in the chatbot is difficult. At the same time, improving the Senja product helps our customers generate more, high quality testimonials and grow their business. We need to be spending most of our limited time in the product and marketing to build a useful product, and sustainable business. #6. We want to build a calm company We're a bootstrapped 2-person team that isn't interested in raising money. We want to build a market-leading product while maintaining creativity, enjoyment and profitability. Our support needs to reflect our 'calm' ethos and quite simply, a chat bot doesn't. Next Steps Going forward, we'll be investing much more in self serve. Admittedly, our knowledge base and help guides could use some work. We still have lots of empty help articles, and we don't update it as frequently as we should. Over the next few weeks, we'll be fleshing out the knowledge base so users can answer 90% of their questions without having to contact support. We're moving from Crisp to a relatively new tool called Fernand. We picked Fernand because we love their philosophy. We should adore talking to our customers, not feeling dread about it. We're also going to invest heavily in the UX of our app. - making elements in the app idiot proof. - triggering help guides when the user is most likely to need them - being explicit about what the user can and can not do from within the product. By doing all these, we hope not just to reduce the number of customer requests we get but improve the quality of the conversations we do have.