Support Hero offers a lot of helpful statistics, but the most actionable is definitely the “failed searches” one.
Support Hero statistics show you the number of failed searches – searches performed by users who did not find their answer in the results, but more importantly, it also shows you the number of deflected tickets. The number of deflected tickets is the number of searches that lead to results users have clicked on and read, and for which, when asked if they have found the answer to their question, they didn’t click “no” or sent a ticket.
In a nutshell, a deflected ticket is a search for which results were displayed, content was read and no ticket was sent.
You can assume that if a search is performed and no content is clicked, the user probably didn’t find her answer. If a search is performed, content is clicked, but a ticket is sent, the user definitely didn’t find her answer.
The only scenario where the user highly probably found her answer is when a search is performed, content is displayed and read and no ticket is sent.
By looking at the deflected tickets menu, you can check what content was read for each search query. Based on our experience, the content read is exactly the answer to the corresponding search query, demonstrating the effectiveness of our approach to measure deflected tickets.
A Failed search is recorded every time a user performed a search and when asked if she’s found the answer to her question, clicks on “no”.
This may be because no content was displayed for that search query or because the content that was displayed did not address her specific question.
After clicking “no”, the user may or may not send a support ticket, this is why you may or may not see a support ticket associated with each failed search.
We do not record a failed search if a user performs a search click or doesn’t click on any of the results, but doesn’t click on “no” when asked if she’s found an answer to her question. She may go ahead and search different keywords for a 2nd search or may just be searching out of curiosity.
Let’s say that your account statistics page gives you the following numbers for last month:
- 30,477 widget views
- 15,658 Deflected Tickets
- 411 Failed Searches
And your ticketing system tells you that you received 2,500 tickets during the same period.
You might be tempted at first to subtract the Deflected Tickets, Failed Searches, and # of tickets sent from 30,477, and it still leaves you with 11,908 remaining instances where the widget was opened and did not result in either a deflected ticket, failed search, or a ticket sent.
But actually, those numbers cannot be added up together to link them to the number of widget openings. To sum things up:
- You had 30,477 widget openings –> number of time a user clicked on the “Need Help?” button
- 15,658 Deflected Tickets -> Number of tickets that were not sent because a user found the answer he was looking for after searching the knowledge base
- 411 Failed Searches -> Number of time a user searched for something and couldn’t find a satisfactory answer (i.e. no results returned, or clicked on “No” when asked if he could find what he was looking for, or sent a ticket)
- 2,500 tickets sent -> it will be the sum of the tickets sent through the contact forms and the articles feedbacks you got.
The information you’re looking for here is the widget openings. You have to keep in mind that when someone opens the widget, he might find the article relevant to his request right away thanks to the URL customization you made. Because of this, a user might open the widget, see filtered articles, and his answer will be just there. In this case, he won’t even search your KB, and then we’re not saving any failed search or saved ticket.
If someone opens the widget and just browse articles, it doesn’t create any failed search or deflected ticket. Failed searches and deflected tickets are only generated when someone searched your KB first. But in the end, the remaining widget openings are likely also deflected tickets of course.
That’s why the number of widget openings will be always bigger than the sum of the other counters.
“We’ve tried almost all support knowledge base products on the market, none came close to what Support Hero is offering. Beyond (beautifully executed) full in-app support, the insights are incredibly useful: 2 months after implementing Support Hero, we cut our volume of inbound support tickets by 50%. No other tool on the market will give you that!”