Real World ServiceDesk Metrics: How Productive Can Bots Make Your Employees?

One of the exciting things about working in the enterprise bot space is that it is sometimes unclear what metrics you should actually be tracking for success. If bots are supposed to automate away some of your monotonous work, and make you more productive, you should be seeing less of certain activities.

As Talla has grown and we have a larger user base, our metrics have stopped jumping around so much and started to settle in to some interesting patterns. One of those patterns admittedly freaked us out a bit. What we saw was that tickets per active organization at Talla consistently decline.

Now, for most servicedesks this would be a bad problem. It means people aren’t using the product, right? But in our case, we saw a corresponding increase in questions answered by Talla.

Below is a sample graph of usage for a sub 50-person company, used with permission.

You can see that Talla automatically answered 106 questions in the last 30 days, while users still opened 56 tickets. This is a remarkably stable ratio over time. Almost every company that uses Talla for 60 days falls into this spot where Talla answers roughly 2/3 of the issues that come up for employees.

Below you can see what the top questions were for the past 30 days.

So, of the 106, only 3 were asked more than 3 times, for a total of 10. What that means is 96 questions were long tail, if you can call it that. Some of these 96 were answers Talla provided, and others were workflows Talla sent automatically to walk a user through something.

I don’t know what the average time spent per ticket was at this company, but here is how you can look at productivity metrics. Assume:

  • 5 minutes to work on a ticket for a service desk admin
  • 30 minutes of delayed work time and context shifting per ticket while the employee who submitted the request waits on an answer and repeatedly checks the status of the request

With 106 of those answered immediately by Talla, there was virtually no employee wait time, saving (106*30 min) = 53 hours of productive work. Add to that the time saving on the helpdesk admin side of (106* 5 min) = 530 minutes or roughly 8 hours of time your helpdesk responder got back this month.

So if your average knowledge worker makes $30/hr, and you just saved 61 hours of time this month, that’s (61*12*$30) = $21,960 in productivity gained throughout the year for a small company to implement Talla. Clearly these metrics are appealing, which is why we’ve seen larger and larger companies starting to look at our enterprise service automation capabilities.

If you are in the market for an I.T. or H.R. service desk, we hope you will give Talla a look.