3 ms·

But if you answer it in a chat, you are going to have to answer it in that chat multiple times. At least with an issue tracker, you can search through the old i

by slaymaker1907 8y ago

But if you answer it in a chat, you are going to have to answer it in that chat multiple times. At least with an issue tracker, you can search through the old issues pretty easily.

meddlepal 8y ago
If we answer it enough times it's an indicator to fix our documentation. Also, often times the reason someone is asking something leads to learning useful stuff about the problem. The downside is repetition, but we invest in doc improvement to avoid that too much. The upside is personalized support which really means building trust with our users and also really trying to understand what they're trying to accomplish. Yea it has overhead but you just build that into your operations. In the long run we probably will start adding things like a Knowledge Base or using other support mechanisms but we've found so far that chat based support is very effective. Anecdotal. YMMV, etc.
antpls 8y ago
There are 2 sides : the customer/user side and the business side. Instant chats are a rather good experience on the user side, but it is more work on the business side (for now). As a customer, I used chat support to ask about information or support both on open-source projects and on profit-oriented businesses, and it's generally a good experience. As a paying customer, however, I feel like chat is not enough, especially if your issue is complex and requires several days to fix. Then the ticket system is better to have visibility as a customer. The underlying idea with chat is that you can first orient people with a conversational bot (name it as you want : a personal assistant, an expert system for searching in stack overflow) that will help them on frequent issues, and then redirect them to an human when the bot's knowledge is not enough. If you have a specific question, the conversational agent will look into different sources of knowledge (forum, documentations, stack overflow, even Google) and do the searches for the customer. It's like an interactive search session for the customer without having to navigate into different platforms and read a lot of text. Chat experience should only get better with improvements in NLP in the future.

But if you answer it in a chat, you are going to have to answer it in that chat multiple times. At least with an issue tracker, you can search through the old issues pretty easily.

meddlepal 8y ago

If we answer it enough times it's an indicator to fix our documentation. Also, often times the reason someone is asking something leads to learning useful stuff about the problem. The downside is repetition, but we invest in doc improvement to avoid that too much. The upside is personalized support which really means building trust with our users and also really trying to understand what they're trying to accomplish. Yea it has overhead but you just build that into your operations. In the long run we probably will start adding things like a Knowledge Base or using other support mechanisms but we've found so far that chat based support is very effective. Anecdotal. YMMV, etc.

antpls 8y ago

There are 2 sides : the customer/user side and the business side. Instant chats are a rather good experience on the user side, but it is more work on the business side (for now). As a customer, I used chat support to ask about information or support both on open-source projects and on profit-oriented businesses, and it's generally a good experience. As a paying customer, however, I feel like chat is not enough, especially if your issue is complex and requires several days to fix. Then the ticket system is better to have visibility as a customer. The underlying idea with chat is that you can first orient people with a conversational bot (name it as you want : a personal assistant, an expert system for searching in stack overflow) that will help them on frequent issues, and then redirect them to an human when the bot's knowledge is not enough. If you have a specific question, the conversational agent will look into different sources of knowledge (forum, documentations, stack overflow, even Google) and do the searches for the customer. It's like an interactive search session for the customer without having to navigate into different platforms and read a lot of text. Chat experience should only get better with improvements in NLP in the future.