Have to completely agree with this approach – did it recently for a client who put on an event which was – putting it mildly – not well received. I created a spreadsheet where the columns (23) represented each distinct issue and each row was online content mentioning (300) the event – ranging from articles, tweets, blogs and our own site comments. For each data point I assigned a -3 to +3 sentiment rating.

When summing vertically, we quickly saw, prioritised the top issues according to our customers and crafted an appropriate response. I ended up adding a weighting multiplier column for an important media company, influential individual or people who had gone to the trouble of analysing the situation. A horizontal summing allowed us to prioritise individual responses.

This level of analysis was certainly time consuming but also well beyond the capability of any automated tool I’ve seen. I feel we owed it to our customers to do it this way because at the end of it, we really did know all the issues inside out.