[1] How long should my bot be/how many questions should my bot have/how long should the conversation take?
The recommended length for your conversation depends on the type of conversation you are designing as well as whether the conversation is Web or SMS.
A note for all conversations:
- The longer (more questions) your conversation is, the more likely it is that a candidate will “drop off,” or not complete the conversation (this is especially true for SMS conversations).
Suggestions for Pre-Screening conversations:
- Pre-screening conversations over the web can be as many questions as you need them to be. There is no recommended length.
- Again, be mindful of the fact that the longer the conversation, the higher the chance that a candidate might not complete the flow.
- Pre-screening conversations over SMS shouldn’t go over ~10 questions. Keep in mind you may have branches that lead to different outcomes. The 10 questions recommendation is for the path a candidate will go down in a single conversation, so you might have more than 10 total questions in your bot.
- Questions do not refer to opening or closing messages.
Suggestions for Data Enrichment conversations:
- SMS conversations: 3-5 questions are ideal (not including greeting, closing, etc.)
- Webchat conversations: 5-7 questions are ideal (not including greeting, closing, etc.)
[2] Should I tell the candidate that they are being contacted by a bot, or omit that information and lead them to believe they are interacting with a human?
You should let the candidate know that they are interacting with a bot, or virtual assistant, in your opening message. Transparency is an important part of building trust with the candidate and managing their expectations for the conversation.
Here’s an example of how to introduce the bot in your opening message:
This is Reva, a virtual recruiting assistant, from Blue Way staffing.
[3] Help! How do I write an effective opening message that will hook candidates into the conversation?
Writing a really good opening message for your bot can be challenging. To make it easier, here are a couple of tips:
- Add a dynamic <firstName> parameter to make the greeting more customized to the candidate.
- Introduce Reva, the virtual recruiting assistant & introduce your company.
- For SMS conversations, tell the candidate why you are reaching out to them. For Web conversations include a reminder about what the conversation is about.
- Give the candidate a heads-up on what will happen once they answer all the questions. Give them information on what to expect when they complete the conversation. Will a recruiter reach out to them? Will they get an email confirmation? Will their profile be updated so they can get more tailored recommendations? Etc.
- For Web conversations remember that the user has just come from an email. Make sure that the first message of the bot flows well from the content of that email.
[4] How do I disqualify a candidate in a pre-screening conversation?
There are two ways to disqualify a candidate.
- Create a branch off of a qualifying question. One branch will lead candidates who answered with the preferred information go down the “happy path” to continue the conversation. Candidates who answer with the not-preferred answer will be led down the “unhappy path,” usually to a consolation message.
- Use a conditional node at the end of your conversation. Configure your conditional node so that if a candidate has passed all your knock-out questions, they will be given the result for passed pre-screened candidates, and if not, they get a consolation message. Using a conditional node to qualify candidates has the added benefit of being able to continue the conversation past the question that a candidate did not pass, allowing you to still fully collect data from that candidate, even if they didn’t pass this specific pre-screen.