The case, explained simply
Customer support · field study
Give customer-service staff suggestions while they remain responsible for the conversation.
What actually happened
The published study Generative AI at Work examines the introduction of an assistant in customer support. It reports an average productivity increase of about 15%, measured as issues resolved per hour, with larger benefits for less experienced workers. Earlier working-paper versions reported different estimates; this case uses the published-study figure.
What the evidence does not tell us
A study in one business setting does not predict the result for every call centre or Greek-language service. The measure concerns resolution per hour, not simply the number of replies generated. Experience and task type affect the benefit.
What a Greek organisation can learn
- Start with an up-to-date set of approved answers for a narrow topic. Let staff review suggestions before sending them.
- Compare new and experienced staff separately. Track repeat contacts, incorrect advice and whether customers actually resolve their issue.
- Use the findings to improve training and the answer library. Do not reward faster replies that create more work for customers later.
Before spending money
For tourism and small businesses, begin with a few common questions in Greek and the languages customers actually use. Keep a clear route to a person.
Read the original evidence
Brynjolfsson, Li & Raymond · Quarterly Journal of Economics ↗The practical steps above are AIStrategy’s editorial suggestions. They are not reported actions or promised results from the organisation featured. Photography illustrates the subject unless its caption identifies the actual facility.