This is really interesting for me, because I'm working on this problem right now with my startup: CustomerFind.
Right now, CustomerFind.com is a Twitter application that automatically follows users on Twitter. You specify a set of keywords, and then automatically follow users who have a tweet mentioning one of those keywords. Presently, on average of ~15% of users will follow back, with good keywords typically generating a 30% follow back rate.
Unfortunately, there's really no way to know if those users are real customers who will use your product/pay you. So I'm working on evolving the service. How many of you would be interested in the next iteration? The idea is to create personal conversations with customers. I am considering the following features for candidacy.
1. You get a list of users who have a tweet mentioning any one of some prespecified keywords.
1a. You have an option to auto-follow these users.
1b. You can filter the list by geography, user popularity, and emotional inflection of the message.
2. You get a CRM tool to enable you to personally engage with users. Sachin: you noticed someone is looking for "sell videogames"? You then have an opportunity to start an @reply/DM conversation with them, within the product. This coincides with Seth Godin's philosophy of permission marketing: you get to reach out to people who really want to hear what you have to say.
2a. You get automated notifications to encourage you to follow-up with more messages. Following up is critical to any good sales methodology.
2b. Each sales/conversation thread contains the user's information, and other facts mined from the tweeting history to help target the conversation.
3. You can type in a competitor's screen name (or your own!), and figure out the most popular words or phrases that its followers are tweeting about. Helps you do better targeting for your own list. For example, you type in @TheNorthFace, and learn that people are talking about "cheap snow jackets".
Anyone interested? Note: if you object by saying "that's too many conversations to keep up with!", I'd like to note that's a problem that you desperately want to have.
I was a user of your service. and when I signed up, I started following so many spammers I had to kill the oauth hookup. I don't know what was wrong with it, but it kept auto following all the horrible horrible spammers and my follower list just filled up with really bad people. I had to manually go through my follower list and delete people. Good thing I didn't have thousands of followers or else it could have taken me days.
The main issue was that zack's software would automatically make me follow someone. He could probably add some basic heuristics to prevent people from following spammers. Even basic bayesian filtering would have worked. Some of the spammer accounts have completely identical tweets. It would seem kind of stupid if I had to write the software to filter my followers... I could just not use customerfind, it seems easier that way.
You should consider disabling the service until you fix this issue. It wasn't just a bad experience for me, it was a negative one... one that cost me around 5 hours of work to clean up.
I'm guessing it's because you want to be the first to respond, e.g., if someone's looking for a project management app you want to suggest your product before a hundred other people suggest theirs. ;)
No, because context is king. If I ask a question and I get an answ3er within 15 minutes I still have the right context. IF I get answer next day I may no longer care as much.
Right now, CustomerFind.com is a Twitter application that automatically follows users on Twitter. You specify a set of keywords, and then automatically follow users who have a tweet mentioning one of those keywords. Presently, on average of ~15% of users will follow back, with good keywords typically generating a 30% follow back rate.
Unfortunately, there's really no way to know if those users are real customers who will use your product/pay you. So I'm working on evolving the service. How many of you would be interested in the next iteration? The idea is to create personal conversations with customers. I am considering the following features for candidacy.
1. You get a list of users who have a tweet mentioning any one of some prespecified keywords.
1a. You have an option to auto-follow these users.
1b. You can filter the list by geography, user popularity, and emotional inflection of the message.
2. You get a CRM tool to enable you to personally engage with users. Sachin: you noticed someone is looking for "sell videogames"? You then have an opportunity to start an @reply/DM conversation with them, within the product. This coincides with Seth Godin's philosophy of permission marketing: you get to reach out to people who really want to hear what you have to say.
2a. You get automated notifications to encourage you to follow-up with more messages. Following up is critical to any good sales methodology.
2b. Each sales/conversation thread contains the user's information, and other facts mined from the tweeting history to help target the conversation.
3. You can type in a competitor's screen name (or your own!), and figure out the most popular words or phrases that its followers are tweeting about. Helps you do better targeting for your own list. For example, you type in @TheNorthFace, and learn that people are talking about "cheap snow jackets".
Anyone interested? Note: if you object by saying "that's too many conversations to keep up with!", I'd like to note that's a problem that you desperately want to have.
-Zack