How useful is Copilot for social media monitoring? We explored this question as part of a research project covering a large number of social media channels. The aim was to find out how efficiently relevant news and posts could be captured and documented. We collected information such as publication date, headline and a short summary of each post.
Our source list was stored in an Excel spreadsheet, and individual posts that had already been identified as relevant were also collected in Excel. One particularly useful feature was the ability to let Copilot work directly with links. Once a link had been entered into an Excel cell, Copilot could be asked to generate a short summary of the corresponding post. In practice, however, the generated texts were not always precise or concise enough. Additional instructions were often needed, for example to shorten the summary, focus more closely on specific content or adjust the tone. On the positive side, once a summary had been approved, it could be inserted directly into the desired cell, which helped streamline the workflow.
There were also limitations when processing content from certain platforms. Copilot was unable to access Facebook posts and therefore could not generate useful summaries for them. LinkedIn posts, by contrast, could be processed.
Overall, Copilot can support monitoring tasks, particularly by helping to create and structure summaries quickly. For larger volumes of content, such as several dozen social media posts, the process is not yet fully efficient or reliable. With clear prompts, some fine-tuning and increasing familiarity with the workflow, however, the integration can still provide useful support in day-to-day monitoring.
Our conclusion: Copilot is already a useful supporting tool for structured monitoring tasks in Excel, but it does not yet replace manual quality control. Particularly for cross-platform research and larger datasets, review and post-editing remain necessary.
By Andrew Colclough