► The first quantitative exploration of services innovation by users. ► 55% of today's computerized commercial banking services were initially developed by service users. ► 44% of today's computerized retail banking services were initially developed by service users. ► Our findings differ significantly from prevalent producer-centered views of service development. Many services can be self-provided. An individual user or a user firm can, for example, choose to do its own accounting – choose to self-provide that service – instead of hiring an accounting firm to provide it. Since users can ‘serve themselves’ in many cases, it is reasonable to suspect that they can also innovate with respect to the services they self-provide – possibly without the assistance of service providers. In this paper, we conduct the first quantitative exploration of the importance of services innovation by users, focusing on the field of commercial and retail banking services. We find that 55% of today's computerized commercial banking services were first developed and implemented by non-bank firms for their own use, and 44% of today's computerized retail banking services were first developed and implemented by individual service users rather than by commercial financial service providers. Manual precursors to these services – manual procedures that carried out functions similar to computerized services in our sample – were almost always developed by users as self-services. Our empirical findings differ significantly from prevalent producer-centered views of service development. We speculate that the patterns we have observed in banking with respect to the major role of users in service development will prove to be quite general. If so, this will be an important matter: on the order of 75% of GDP in advanced economies today is derived from services. We discuss the implications of our findings for research and practice in service development.