Most contact center agents are happy with Deepdesk, research shows. It makes them type faster and takes away redundant tasks. Our survey also makes clear that agents need tutorials and ongoing training. And they want our AI to be even more personalized.
Fitter, happier, more productive?
At Deepdesk we claim to help agents at contact centers to be more effective when chatting with customers. So, does the agent indeed feel being helped? To gain more insight, we launched a survey among agents working at two of our customers. The survey was completed by no fewer than 100 agents.
Nearly 70% of agents rated Deepdesk with a 7/10 or higher. The ease of use gets a 7.3, and the speed a 7. This is pretty high already. We started from scratch, only two years ago. 18% of agents give us a 5 or lower. We have some room for improvement here. More on this below.
Growth in use
This appreciation is also reflected in the use of Deepdesk. We measure which part of all your text has been typed, and which part has been completed by Deepdesk. We call the latter Assisted Text. For one customer, the percentage of Assisted Text has grown from 4.4% to more than 11% in only seven months, peaking to even more than 16% in the last couple of weeks. For another customer, the figures have always been a bit higher, with some weeks an AT percentage above 20% and even days when almost a quarter is Assisted Text.
Happier with less redundancy
What agents especially like about Deepdesk is that they don’t have to keep typing the same thing all the time, say 84 of the respondents. Deepdesk autocomplete also radically decreases the amount of typo’s, something that almost a quarter of the respondents highly appreciated. For reference, an agent on average deletes 15% of his written text. There's already great value in reducing typos.
Faster? More than 60% say yes.
Asked if they think Deepdesk speeds up conversations with customers, over 60% says they do. And only a small percentage (less than 9%) doesn’t think so.
Those subjective answers are supported by our data. On average, agents type twice as fast while using Deepdesk autocomplete and reply suggestions.
The need for tutorials and ongoing training
However, a pretty large group doesn't feel comfortable using Deepdesk yet, giving us a low rating overall (see above) or reporting not to be faster using Deepdesk. We need to help them. We are rapidly evolving in terms of our own customer care, by creating smooth tutorials and improvements based on agent recommendations. We asked the agents what they need from us:
More personal chat conversations
Our AI suggests the best answers from over a million conversations. However, the best answer in general might not always be the best answer a particular customer needs or a particular agent is willing to give. We strongly believe in the added value of the human agent. So, personalization of suggestions is a big thing for us. It turned out to be a big thing for the agents as well. Right now, we are working on adapting our algorithms to make autocomplete and reply suggestions even more personal. More about this later!
Want to save on redundant tasks without the disadvantage of chatbots? Check out our features or get in touch with us.