Social Media
Social listening and brand monitoring
Social listening tells you what people say about your brand when they are not tagging you. We monitor mentions, track sentiment and competitor conversation, and turn it into a monthly brief you can act on.
Who it is for: Brands that want early warning on problems, honest feedback at scale, and sight of competitor conversation.
Everything in this service
- Brand and product mention monitoring across platforms
- Sentiment tracking over time
- Competitor mention and share-of-voice comparison
- Alerts for spikes and emerging issues
- A monthly insight brief with recommended actions
What to expect
- Problems caught while they are still small
- Honest customer feedback without running a survey
- A clear view of how conversation compares to competitors
How social listening actually works
Setting up monitoring that catches signal instead of noise
Badly configured listening produces thousands of irrelevant mentions, which is worse than no listening at all, because the alerts get muted within a fortnight and nobody looks again. The setup is most of the work.
The query set needs more than a brand name. It should include common misspellings and how people actually type your name, product names and model numbers, the names of your founders and any public-facing staff, your main campaign phrases and hashtags, your domain as a plain-text mention without a link, and the phrases people use when they are complaining about your category rather than about you. Then the exclusions, which matter just as much: your own accounts, your own staff posting, recruitment listings, aggregator spam, and any unrelated business or term that shares your name. Most brand names collide with something, and that collision is usually the source of ninety percent of the noise.
Coverage has honest limits and it is better to say so up front. Public posts on the open social platforms, forums, review sites, news and blogs are generally reachable. Private groups, direct messages, closed communities and much of what happens inside messaging apps are not, and any tool claiming full coverage is overselling. Some platforms restrict what third-party tools can access, and that changes from time to time, so the practical answer is to combine tooling with periodic manual checks of the places your customers actually gather.
- Track misspellings, product names, staff names and unlinked domain mentions
- Exclude your own accounts, job ads and same-name businesses
- Include category complaint phrases, not just brand terms
- Tune weekly for the first month, then monthly
- Be honest that private and closed spaces cannot be monitored
Sentiment analysis and where it stops being reliable
Automated sentiment scoring is useful as a trend line and unreliable as a verdict on any single mention. Knowing where it breaks is what stops a report being confidently wrong.
It fails predictably on sarcasm, which reads as positive to a machine and is usually the opposite. It fails on industry shorthand, where words like sick, mad or ridiculous can mean either thing depending on context. It fails on politely worded complaints, which contain no negative words at all, and on neutral factual mentions that get sorted into a bucket they do not belong in. It also struggles with mixed messages, where someone praises the product and criticises the delivery in the same sentence, since a single score cannot represent both.
The way to use it responsibly is as a change detector rather than a measurement. A sudden shift in the ratio is worth investigating, and that is a genuinely valuable early warning. The absolute number is not worth reporting as a fact. Every sentiment figure we publish is checked by reading a manual sample, and where the sample disagrees with the tool, we say so in the report. Tracking themes is usually more useful than tracking sentiment anyway. Knowing that complaints about delivery times tripled this month is actionable in a way that knowing sentiment fell four points is not.
- Treat sentiment as a trend line, never as a verdict
- Expect failures on sarcasm, shorthand and polite complaints
- Sample manually and report where the tool disagreed
- Track themes and volume alongside sentiment
- A sudden shift matters more than the absolute score
Listening to competitors and to the whole category
Most of the useful conversation does not mention you at all. Category listening, which means monitoring the problem your product solves rather than your brand name, is usually where the best material comes from.
On competitors, the value is in the complaints. What people consistently criticise about the alternatives is a direct list of what to emphasise in your own positioning, and it is far more reliable than a features comparison, because it comes from people who paid. Their launches, their pricing changes and how their customers react to those changes are all visible in public. Share of voice, meaning your share of total category mentions, is a fair benchmark as long as everyone understands that raw volume can be driven by a crisis just as easily as by success.
Category listening looks for people describing the problem in their own words, asking for recommendations, or announcing that they are switching. Those recommendation threads are worth watching closely, both because they show which names come up unprompted and because they show what criteria people actually use. One clear rule applies to all of this: participating in those conversations means identifying yourself as the brand, honestly, and adding something useful. Posting anonymous recommendations of your own product, or paying anyone to do it, is deceptive and in review contexts it is unlawful. We will not do it.
- Monitor competitor complaints for positioning material
- Watch recommendation threads for unprompted mentions and real criteria
- Track share of voice, with the caveat that volume can be bad news
- Listen for problem language, not only brand names
- Always identify yourself when joining a conversation
Feeding what you hear back into content, product and service
Listening only pays when the insight reaches someone who can act on it. A monthly report that circulates to marketing and stops there is an expensive newsletter.
Four routes cover most of the value. Content gets the questions people keep asking, in the exact words they use, which is the best possible source of article topics and headings because it is unfiltered demand rather than a keyword tool guess. Product gets the recurring complaints and the feature requests, with a count against each so priority is based on frequency rather than on whoever complained loudest. Customer service gets the problems that appear in public before they arrive as tickets, plus any gap in the documentation that the same repeated question exposes. Sales gets the objections and the comparisons people make between you and the alternatives.
The routing has to be a habit rather than an intention, which means a named recipient and a fixed monthly slot. It also helps to close the loop publicly. When a complaint that surfaced through listening leads to a change, saying so where the original conversation happened is one of the cheapest trust-building actions available, and it makes people more willing to tell you things next time.
- Recurring questions become article topics, in customer wording
- Complaints and requests go to product with frequency counts
- Emerging issues reach support before the tickets arrive
- Objections and comparisons go to sales
- Tell people publicly when their feedback changed something
Crisis detection: seeing it early enough to matter
The value of listening is highest in the hours before a problem becomes widely known. Most issues that end up damaging a brand were visible in a small number of posts well before anyone official noticed.
Detection depends on alerting on the right thing. Volume alone is a weak trigger, because a marketing success creates a spike too. What works better is a change in the ratio: a rise in negative mentions relative to the normal baseline for that time of week. Other triggers worth setting include a mention by an account with unusually large reach, any post carrying the words scam, fraud, lawsuit, dangerous or recall alongside your name, a cluster of similar complaints appearing within a short window, and any journalist mentioning you at all.
Alerts are only useful if the response is agreed in advance. Who receives them, including out of hours. Who assesses whether it is genuinely an issue or one loud account, which is a real judgement and one worth making calmly. Who has authority to pause scheduled posts, which should happen immediately and can be reversed later at no cost. And what gets said in the first hour, which is normally an acknowledgement rather than an explanation. Not every negative spike is a crisis, and treating one every time is its own problem, because a business that reacts publicly to every complaint teaches people that complaining publicly is the fastest route to attention.
- Alert on the ratio of negative mentions, not raw volume
- Set specific triggers for scam, lawsuit, recall and safety wording
- Flag mentions from high-reach accounts and any journalist
- Agree who assesses, who pauses the schedule, and who speaks
- Distinguish a real issue from one loud account before responding
Reporting: what a monthly listening brief should contain
A listening report is judged on whether anyone does anything differently after reading it. A dashboard export is not a report, because it hands the interpretation back to the reader.
Five sections do the job. Volume and trend, with the reason for any notable spike explained rather than just marked. Themes, which are the three to five subjects people actually discussed, with real quotes attached because a quote lands where a summary does not. Sentiment with its caveats, presented as direction rather than a precise figure. Competitive comparison, meaning share of voice and anything meaningful a competitor did. And recommended actions, split by who has to do them: content, product, support, sales.
The part that gets remembered is usually the quotes. Two or three verbatim comments, positive and negative, communicate more to a leadership team than any chart, because they are unarguable. The report should also record what changed as a result of the previous month, which turns a monthly document into a visible chain of cause and effect and makes it much harder for the work to be quietly deprioritised.
- Volume and trend, with spikes explained
- Three to five themes, each with real verbatim quotes
- Sentiment as direction, with its limits stated
- Share of voice and notable competitor activity
- Actions by owner, and what last month actions produced
A clear path, step by step
- 01
Set up tracking
We configure monitoring for your brand, products, competitors and the terms that matter.
- 02
Monitor
Mentions and sentiment tracked continuously, with alerts when something spikes.
- 03
Analyse
We separate signal from noise: what people praise, what they complain about, what is changing.
- 04
Act
A monthly brief with recommended actions for content, product and customer service.
Why choose us for this
Human analysis on top of the tooling, not raw dashboards
Alerts tuned so you hear about real issues, not noise
Insights routed into your content and social plans
Common questions
What is the difference between social listening and monitoring?
Monitoring collects the mentions. Listening interprets them: sentiment, themes and trends over time. We do both, and the monthly brief is where the value lands.
Which platforms can you track?
The major social networks, forums, review sites and news sources. Coverage varies by platform API access, and we are upfront about what can and cannot be tracked.
Explore related work
Want this for your business?
Book a free visibility call and I will tell you honestly whether I can help.
How this is delivered
One person leads every project. Where a job genuinely needs a specialist, I bring in people I have worked with before and manage them, so you get one point of contact and one invoice rather than three suppliers blaming each other.
- You talk to the person responsible for the work, not an account manager
- Specialists are briefed and managed by me, and their work is checked before it reaches you
- One contract, one invoice, one place to chase