Build a competitive intelligence report from live X mentions, not a vibe

You are building a competitive intelligence report from live X mentions of named competitors, using Grok's real-time post access. Every comparative claim needs a source; an impression that people seem to like a competitor more than us, with no post backing it, is not a finding — it's a guess wearing a findings format.

OUR BRAND
Fieldnote, a field-service scheduling app for small contracting businesses

NAMED COMPETITORS
Jobber and Housecall Pro

TIME WINDOW
Past 30 days

SPECIFIC COMPARISON DIMENSIONS
Mobile app reliability complaints, and response time when users publicly complain about a bug

DECISION THIS REPORT FEEDS
Product team deciding whether mobile stability should be next quarter’s top engineering priority over the two features currently ahead of it in the backlog

REPORT RULES
Search live X mentions for our brand and each named competitor separately across Past 30 days, then compare only along the specific dimensions listed — do not drift into a general who's-winning narrative if the ask was about one specific thing, like response time to complaints or a specific feature people are requesting. Match volume to context before drawing any conclusion from it: a competitor with ten times the mention volume might simply have ten times the user base, so a raw mention-count difference is not evidence of a sentiment or quality gap on its own — normalize your read by mentioning the scale difference explicitly rather than letting it silently inflate a comparison. Separate what users are praising a competitor for from what users are merely mentioning neutrally — a competitor named in passing while a user is still deciding is a different data point than one being actively praised, and folding both into one tally overstates the competitor's actual standing. Look specifically for churn signals — users explicitly stating they left one product for another and why — since these are the highest-value data points in a competitive scan and are easy to lose inside a larger volume of generic mentions if you're not searching for that pattern directly. Note anything a competitor appears to be doing that's generating notably positive reaction and that we are not currently doing, since that is the most actionable output of a report like this — a comparison that only confirms existing assumptions without surfacing a concrete gap has not earned its own existence. State your confidence per dimension separately rather than one blanket confidence for the whole report — a comparison might be well-supported on pricing sentiment and thin on support-quality sentiment within the same time window, and collapsing that into one number hides which conclusions are safe to act on.

OUTPUT FORMAT
1. A table: Dimension | Us | Jobber and Housecall Pro | Confidence | Key supporting quote per brand.
2. Churn signals found, quoted, with the stated reason for switching.
3. The single most actionable gap: something a competitor is doing well that generated real positive reaction and we are not currently doing.
4. What this report cannot tell you — the specific question Product team deciding whether mobile stability should be next quarter’s top engineering priority over the two features currently ahead of it in the backlog needs answered that live X mentions alone cannot settle.