How to prioritise customer feedback: volume, negativity and trend

Why volume is not priority

When teams work out how to prioritize customer feedback, they usually start by sorting topics by how often they are mentioned. The top of the list is usually something like "staff" or "product", which customers mention constantly, mostly positively. Volume tells you what customers talk about. It does not tell you what is going wrong.

The three measures

1. Negative volume: how many reviews mention the issue negatively in the period. This is the size of the problem.
2. Negative share: of all mentions of the issue, what fraction are negative. This is how bad it is when it comes up.
3. Trend: how the negative share changed against the previous period. This is whether it is getting worse.

An issue that is big, mostly negative and rising goes to the top. One that is big but mostly positive is a strength. One that is small but suddenly rising is an early warning.

A simple prioritisation grid

• Fix now: high negative volume and rising share.
• Plan a fix: high negative volume, stable share. A known, persistent problem.
• Investigate: low volume, sharply rising share. Read the reviews before it grows.
• Monitor: low volume, stable share.
• Protect: high volume, mostly positive. What customers value; do not break it.

Two overrides: sample size and severity

Two things change the order. First, small samples: a share that doubled from 2 reviews to 4 is not a trend. Flag issues below a minimum number of reviews and look at them again next period. Second, severity: safety, hygiene, discrimination, fraud and legal threats go to the top whatever their volume. One credible food-safety review outranks fifty about parking.

A worked example

• Staff friendliness: 900 mentions, 12% negative, stable. Protect.
• Billing wait: 340 mentions, 61% negative, up from 48%. Fix now.
• Exchange policy: 120 mentions, 70% negative, stable. Plan a fix.
• App order pickup: 40 mentions, 55% negative, up from 20%. Investigate.
• Parking: 80 mentions, 65% negative, stable, outside your control at most stores. Monitor, and reply well.

Adding cost and control

Once issues are ranked by customer impact, add two practical questions before committing: can we change it, and what will it cost? A cheap fix to a high-priority issue goes first. An expensive fix to an issue we only partly control may wait for a business case. Keep these as a second step, so customer evidence sets the order and practicality adjusts it, not the other way round.

Store-level versus network-level

Run the same method at two levels. Network-wide, it tells theme owners which issues to tackle centrally. Store by store, it tells regional managers which issues to raise at which stores. An issue that ranks high everywhere usually has a central cause; one that ranks high at a handful of stores usually has a local one.

Pitfalls

• Ranking by volume alone and spending a quarter on a topic customers mostly praise.
• Chasing every rise. A share that rises on a handful of mentions is usually noise; flag it and wait.
• Letting the loudest internal voice reorder the list. Keep the evidence visible in the meeting.
• Forgetting strengths. Topics customers praise are worth protecting when budgets are cut or processes change.
• Never removing items. Once an issue is fixed and stays fixed for a few months, take it off the list.

Presenting the priorities

Present the list as a short table each month: issue, negative mentions, negative share, change, group in the grid, owner and next step. Keep it to the ten or so issues that matter, and attach two reviews to each. A list that fits on one screen gets discussed; one that needs scrolling gets skimmed.

Adding other signals

Reviews are one view of customers. If you also have complaint logs, call-centre categories or survey comments, run the same three measures on them and look for agreement. An issue that is rising in reviews and in call volume is a stronger priority than one visible in only one source. Keep the sources separate in the numbers, so a change in one channel is not mistaken for a change in customer experience.

Doing this with Falcon

Falcon labels each topic in scope as needing attention, emerging or a strength, from its negative and positive counts and the change against the previous period. Each insight names the issue, why it matters (from the summary of the most-mentioned issue in scope) and a next step that names the taxonomy owner. Actions can be tracked to done, and because reviews keep arriving, the same signals show whether the issue's negative share falls afterwards.