top of page
Abstract Shapes

INSIDE

PUBLICATIONS

Sentiment Analysis for Brand Monitoring

Updated: 2 days ago

Sentiment Analysis for Brand Monitoring
Sentiment Analysis for Brand Monitoring

UIC emblem

UIC University 365 Institute of Communication

Series Marketing

Duration 25 minutes

Digital Communication, Marketing, Branding, Content Strategy, Media Studies


UNOP isochrone

UNOP Sound (University 365 Neuroscience Oriented Pedagogy)

Take five minutes to prepare your brain. Play the isochronous tone track (40Hz gamma frequency) with your eyes closed. Gamma-frequency tones before a learning session raise attention and make the material easier to absorb.

[Audio player: UNOP Pre-Lecture Isochrone (40Hz, 5 minutes)]

In this Lecture


Back to the TOC

The Hook: The Mention You Never Saw


A customer posted a two-paragraph complaint about your delivery times at 21:40 on a Tuesday. By Thursday it had 400 reactions and eleven replies, three of them from people saying they had the same experience. You found it on Friday, in a direct message from a friend.


The complaint was accurate. It was specific. It named the process, not the person. It was, in other words, the most useful piece of customer research you received that quarter, and it reached you three days late and second-hand.


The old answer was a media monitoring contract, priced for companies with a communications department and a budget line. The new answer is a pipeline you can build in an afternoon: collect the mentions, classify the sentiment, group them by theme, and alert a human when something is worth answering.


This lecture builds that pipeline. You will finish knowing how sentiment is actually measured, why most sentiment scores are less reliable than they look, and which decisions the score can support and which it cannot.

Back to the TOC

Step 1: What Sentiment Analysis Actually Measures


Sentiment analysis assigns a label to a piece of text. The labels look simple. The measurement is not.


Three levels of granularity, in increasing difficulty:


Level

Question it answers

Example

Document

Is this whole post positive, negative or neutral?

One review, one score

Sentence

Which sentence carries the feeling?

A positive review with one angry sentence

Aspect

How does the writer feel about each specific thing?

Delivery 2/5, product quality 5/5, price 4/5


Aspect-level analysis is what you actually want for brand monitoring, and it is the hardest of the three. A document-level score on a review that says "the app is excellent but support never replied" returns one number and loses the finding.


The three-value scale is a lie of convenience. Most tools output positive, negative or neutral. Real writing contains:


  • Mixed sentiment inside one sentence. "Faster than the old version, but it crashed twice."

  • Sarcasm. "Great, another mandatory update." The words are positive and the meaning is not.

  • Negation chained across clauses. "I did not expect to like this, and I still do not."

  • Domain shifts. "Sick" is negative in a hospital review and positive in a skateboard review.


A sentiment model reports confidence for a reason. A score of 0.62 positive is not the same finding as 0.96 positive, and treating them the same is how you end up reporting noise to a leadership team.


What the score is and is not. A sentiment score is an index, not a verdict. It is useful for tracking change over time within one brand and one topic. It is not a measure of your reputation, and comparing your score with another brand's score from a different tool measures the tools more than the brands.


The same sentence at document, sentence and aspect level, and what each one loses
The same sentence at document, sentence and aspect level, and what each one loses
Back to the TOC

Step 2: Where the Data Comes From


Collection is the part of the pipeline people under-plan, and it is the part that determines whether the rest is worth doing.


Four categories of source, each with a different access reality:


Source

Access route

What it gives you

Reality check

Owned channels

Your own platform APIs and exports

Reviews, support tickets, survey answers, comments

Highest signal, fully yours, usually a small volume

Social platforms

Official APIs where they exist

Posts, comments, replies, mentions

Access terms change; several platforms restrict retrieval and some prohibit scraping

Review and marketplaces

Published review pages and official review APIs

Star ratings with text, product-level detail

Structured and high-intent, often the best cost-benefit

News and blogs

RSS feeds and publisher APIs

Coverage and commentary

Low volume, high impact


The three collection rules.


  • Use the official interface wherever one exists. Platform APIs and export files give you stable, permitted access. Retrieval outside the published terms creates legal exposure and breaks without warning.

  • Capture the metadata, not just the text. Timestamp, platform, author type (customer, prospect, journalist, competitor employee), language, and a permalink. A mention with no timestamp cannot be trended; a mention with no permalink cannot be answered.

  • Record the sampling rule. If you collect only mentions containing the brand name, you miss the ones that describe the experience without naming you. Write down what your filter includes and excludes, and revisit it quarterly.


Volume honesty. A small brand may see 20 mentions a week. That is a good dataset for reading, and a bad dataset for statistics. Do not build a percentage change on 20 items: one extra complaint moves the number by five points. Read the mentions at that volume, and start trending when the weekly count is stable above roughly 100.

Back to the TOC

Step 3: The Three Methods, and What Each One Costs


There are three ways to get a sentiment label. They differ in cost, transparency and accuracy, and the right answer is usually a combination.


1. Lexicon-based scoring. A dictionary maps words to scores ("excellent" +2, "broken" -2) and the text is summed, with rules for negation and intensity.


  • Cost: effectively zero, runs locally, no data leaves your systems.

  • Strength: fully explainable. You can open the dictionary and see why a score came out where it did.

  • Weakness: blind to context. It misses sarcasm, domain shifts and mixed sentences.


2. Machine learning classifiers. A model trained on labelled text predicts the label. Available as hosted APIs or small open models you run yourself.


  • Cost: per-call pricing on hosted services, or compute if self-hosted.

  • Strength: handles context and negation far better than a dictionary.

  • Weakness: it is a black box on a per-item basis, and accuracy is domain-specific. A model accurate on restaurant reviews may be weak on enterprise software complaints.


3. Large language model classification. You pass a batch of mentions with a structured instruction and receive labels plus a one-line reason.


  • Cost: the highest per item, but falling, and zero training work.

  • Strength: handles context, sarcasm and mixed sentiment, and can classify aspect and intent at the same time. The reason field is what makes it auditable.

  • Weakness: cost at volume, output variability, and the need for a fixed instruction to keep labels comparable across runs.


Practical guidance.


  • Use a lexicon on very high volume where you only need a rough direction.

  • Use a hosted classifier when you have tens of thousands of mentions and a stable domain.

  • Use an LLM on the subset that matters: mentions of new products, escalations, journalists, high-reach posts, and anything the cheap layer scored as extreme.

  • Always run a human-labelled sample of 100 to 200 mentions against whichever method you choose, and compute agreement. A method you cannot measure is a method you cannot defend in a meeting.


Lexicon, classifier and LLM classification compared on cost, transparency and context
Lexicon, classifier and LLM classification compared on cost, transparency and context

The scoring brief that keeps LLM labels comparable. Fix the scale, the output shape and the tie rules once, then reuse it verbatim:


Classify each mention for brand monitoring. Output one JSON object per line: {"id": ..., "sentiment": "positive|negative|neutral|mixed", "aspect": "<product|delivery|support|price|other>", "intent": "<praise|question|complaint|churn_risk|none>", "reason": "<one short line>"} Rules: use "mixed" when one mention clearly contains both strong positive and strong negative statements. Mark intent "churn_risk" only when the writer states or clearly implies they will stop using the brand. Never infer a fact the text does not state.

Back to the TOC

Step 4: Building the Pipeline


Five stages. Each one is small, and each one can be run on a schedule.


Stage 1, collect. Pull from the sources chosen in Step 2 on a fixed cadence. Store raw items untouched, in a table with the raw text and the metadata. Keep the raw layer: your classification rules will change, and you will want to re-run them.


Stage 2, deduplicate. Cross-posting is the biggest source of inflation in brand monitoring. One complaint published to three platforms and shared four times is one complaint. Deduplicate on a normalised text similarity plus an author and time window, and record the cluster size as a separate field. Cluster size is a useful signal; counting it as seven mentions is not.


Stage 3, classify. Apply the method from Step 3. Store the label, the confidence, the aspect, the intent and the reason. Never overwrite a human correction: keep the machine label and the human label in separate columns so you can measure where the method drifts.


Stage 4, aggregate. Group by theme over a rolling window. A theme is a recurring subject, not a keyword: "delivery delay", "invoice confusion", "battery life". Roll up counts, the sentiment mix, and the reach of the top items per theme.


Stage 5, alert and route. Nothing in the first four stages matters unless it reaches a person. Route by rule, not by volume:


Trigger

Route to

Within

Churn risk intent detected

Customer success

Same business day

Journalist or high-reach account

Communications lead

2 hours

Safety, legal or regulatory claim

Legal or compliance

Immediately

Negative theme count up 50 percent week over week

Product owner

Weekly review

Everything else

Dashboard only

No alert


Cadence that fits a small team. Collect hourly or daily depending on volume, classify on the same run, aggregate nightly, send one digest each morning and immediate alerts only for the top three rules. A pipeline that alerts on everything is a pipeline people mute.


The five pipeline stages and the alert routing rules that reach a person
The five pipeline stages and the alert routing rules that reach a person
Back to the TOC

Step 5: The Dashboard That Answers a Question


Most sentiment dashboards fail for the same reason: they show what is easy to compute instead of what someone needs to decide.


Build the dashboard backwards from the decisions. For brand monitoring there are five real decisions:


  • Do we need to respond to something today?

  • Is a specific product or process generating complaints?

  • Did last week's change make things better or worse?

  • Which channel deserves more attention?

  • Is a narrative forming that we should address publicly?


Each decision has one view:


Decision

View

Metric

Respond today?

Alert queue

Open items by severity and age

Product or process problem?

Theme table

Mentions per theme, sentiment mix, week over week

Did the change work?

Trend line

Sentiment mix on the affected theme, before and after

Which channel matters?

Channel table

Volume, negative share, average reach

Is a narrative forming?

Narrative tracker

Co-occurring terms, top accounts, velocity


Two discipline rules for the numbers.


  • Always show the denominator. "Negative mentions up 40 percent" means nothing without "from 10 to 14 of 300". Small counts move in large percentages.

  • Never show a single composite score without its components. A "brand health index" that merges sentiment, volume and reach hides the three findings it was built from. Show the parts.

Back to the TOC

Step 6: From Signal to Response


A dashboard is not a communications strategy. What you do with a detected negative mention follows a small decision tree.


Classify the mention before you reply.


  • Accurate and specific. Thank the writer, state what you are doing, and give a route to follow up. One competent reply to an accurate complaint is worth more than ten defensive ones.

  • Accurate but general. Ask one clarifying question in public and move the detail to a private channel.

  • Inaccurate and specific. Correct the fact once, calmly, with a source. Do not repeat the wrong claim in your reply.

  • Abusive or coordinated. Do not engage. Record it, note the pattern, and move on. Engagement is the goal of coordinated abuse.


Timing. For a service complaint, reply within one business day. For a safety or legal claim, do not reply at all until the responsible function has cleared the wording. For a journalist, respond within hours, even if the answer is "we are checking and will come back today".


What not to do. Do not send a template that does not engage with the specific text. Do not argue about the sentiment score with the customer. Do not delete an accurate complaint. Do not turn a single mention into a public statement.


Feed the finding back. Every recurring negative theme belongs in one of three places: the product backlog, the process documentation, or the FAQ. A monitoring programme that produces replies but never changes a process is a cost centre.

Back to the TOC

Step 7: Where Sentiment Analysis Fails


The failures are predictable, and knowing them is what keeps the output trustworthy.


1. Sarcasm and irony. "Love that the app logs me out every hour" scores positive on a lexicon and often neutral on a classifier. LLM classification handles it better, not perfectly.


2. Negation scope. "Not bad at all" and "not good at all" differ by one word.


3. Domain and dialect drift. Product names that are ordinary words ("Excel", "Pages", "Teams") produce constant false matches unless the filter requires context. Regional slang, emoji and abbreviations vary by community.


4. Aspect confusion. A sentence praising the product and criticising the price must produce two labels, not one average.


5. Selection bias. People who write about a brand are not a random sample of its customers. Your sentiment mix measures the vocal subset, and quiet dissatisfaction never appears. Treat the pipeline as a detection system for themes, not as a satisfaction survey.


6. Translation loss. A pipeline that translates first and then classifies adds a second error on top of the first. Where volume justifies it, classify in the original language with a model that supports it.


7. Feedback loops. If you act only on what the pipeline flags, the pipeline's blind spots become your blind spots. Read a random sample of raw mentions every week, unfiltered, to test the system against reality.


The professional habit. For any claim you make from this data, be able to say: how many items, over what period, collected how, classified by which method, with what measured agreement against a human sample. If you cannot answer those five, you have an impression, not a finding.

Back to the TOC

Step 8: Listening Without Reading Everything


A monitoring pipeline that covers every mention is a pipeline nobody reads. The discipline that makes sentiment analysis useful is a filter you design on purpose, before the data arrives.


The three levels of listening


Most teams run one of these and call it monitoring. They answer different questions and they cost different amounts.


Level

Scope

What it answers

Cost

Direct mentions

Your name, your handles, your domains

What is being said about you, today

Hours per week, mostly triage

Category mentions

The problem you solve, and the words your customers use for it

Where the market's attention is moving

Hours per day to do properly, or a paid tool

Competitor and partner mentions

Named rivals, integrators, the analysts who compare you

What the alternatives claim, and where you are placed

Requires a maintained entity list


The trap is starting with the category level because it sounds thorough. It produces a firehose, most of it irrelevant, and the team stops reading it within a month. Start with direct mentions, tune the filter until every item is worth a glance, then expand one level.


Designing the query set


A monitoring query is not a keyword. It is a boolean expression with a scope and an exclusion list. Build it in four parts:


  • The entity. Every spelling a human actually uses: the formal name, the short name, the domain, the handle, common misspellings, and any acronym that collides with another meaning. For a generic-word brand, add a required context term next to it.

  • The exclusion list. The homonyms that produce noise. If your brand name is also a place, an artist or a common noun, you will otherwise spend your week reading about somebody else. Keep the list in a file, and add an entry every time the same false positive appears twice.

  • The scope. Language, geography, and the platforms that matter for your audience. Do not monitor every platform; monitor the ones where a conversation would change a decision.

  • The priority tier. Mark a small subset as "always read" (named analysts, key accounts, the accounts that reach your customers) and treat the rest as batch.


Review the query set monthly and count two things: the false positive rate and the missed mentions you found by accident. Both numbers tell you whether the filter is working, and neither is reported by a monitoring tool.


The signal that matters is a change, not a score


A sentiment score of 0.62 is not actionable. A change in that number is. Three patterns are worth an alert, and nothing else should wake anyone up at night.


  • A sustained shift, not a spike. One angry post is a Tuesday. Fourteen days where the same complaint appears twice a day is a trend, and it usually names a specific product or process.

  • A new theme inside known volume. The mention count is flat but the vocabulary changed. A new cluster of phrases inside the same volume is often the leading edge of a real issue.

  • A shift in who is speaking, not how much they say. You lose a supportive voice, or an account that reaches your customers starts repeating a criticism. Reach and sentiment move independently, and reach is usually the more important of the two.


Write the alert rules down as plain sentences and keep them next to the query set. If a rule cannot be stated in one sentence, it is not a rule, it is an intention.


Sampling the raw text, or you will misread the chart


An automated classifier assigns a label. It does not tell you why, and it is wrong in predictable ways on sarcasm, on mixed reviews, and on comparisons ("better than the alternative, but still slow").


Once a week, read twenty raw items at random, including items the classifier scored at the extremes. Two questions:


  • Of the items labelled negative, how many are actually negative about you?

  • Of the items labelled positive, how many contain a complaint inside the praise?


Twenty items cost fifteen minutes, and they correct the two errors that make a dashboard misleading: misattributed negative sentiment and concealed dissatisfaction. Record the answers, because the ratio is the calibration of every number above it.


The three levels of listening and the rules that produce an alert
The three levels of listening and the rules that produce an alert
Back to the TOC

Step 9: From Alert to Response, Without Overreacting


Monitoring has no value until a signal changes what somebody does. That step is a decision procedure, and it is where most teams either overreact to noise or miss the one mention that mattered.


The triage grid


Sort every item that reaches a human into one of four boxes. The box determines the response, and the response time.



Low reach

High reach

Factually wrong

Correct if it will be found by others; otherwise ignore

Correct publicly, quickly, with the source named

Legitimate complaint

Log it, reply briefly, fix the cause

Reply publicly, then fix the cause, then say what was fixed

Opinion you disagree with

Leave it

Leave it, unless it misstates a fact

Praise or useful signal

Acknowledge lightly or not at all

Thank the person, and add it to the case evidence


Two rules make the grid practical. First, separate "the claim is false" from "the claim is unflattering". Only the first warrants a correction. Second, decide the response time per box once, in advance, so nobody has to negotiate urgency while a thread is spreading.


Reasonable response times, agreed in advance: high reach and factually wrong within two hours; high reach complaint within one working day; everything else within one week, or never.


The correction that does not inflame


A public correction has four parts and no adjectives.


  • State the claim you are correcting. Quote it or name it precisely. Do not paraphrase it into something easier.

  • State the fact, with its source and its date. A number and a link. If the source is internal, state the method.

  • State what you are doing about the cause, if there is one. If the mistake was yours, say so in one sentence and stop. A long explanation reads as an excuse.

  • Give one route to follow up. A named address or a form, not a general inbox.


Then stop. Do not reply to every subsequent message in the thread, and do not correct the same claim in five places unless it appears in five places with independent reach.


When not to respond


Most negative mentions should not get a reply, and deciding this in advance protects the team from the instinct to defend every statement.


Do not respond when: the mention has no reach; the claim is an opinion; the author is arguing in obvious bad faith; the thread is already closed; or the only thing you would say is a restatement of your marketing. A reply that adds no information only adds reach to the original.


One exception: a mention from a paying customer, however small, is worth a reply even when the reach is zero. They will notice, and the reply is customer service rather than public relations.


The weekly one-page report


A monitoring programme is judged by the decisions it produced, not by the volume it processed. Write one page a week, with fixed sections:


  • Volume and shape. The mention count against last week, split direct and category. One line.

  • The three items that moved a decision. Each with a link, the action taken, and the owner.

  • The theme to watch. One emerging phrase or complaint, with its first-seen date.

  • What the pipeline missed. Any mention found by accident, and the query change it implies.

  • What we stopped tracking. Queries removed, and why.


The last two sections are the ones that keep a programme honest. A monitoring system that never loses a query is a system nobody has pruned.


Measuring the programme itself


Four numbers describe your own pipeline, and none of them is a sentiment score.


Number

What it tells you

Healthy direction

False positive rate

Share of surfaced items that were about somebody else

Falling

Time to first read

Median delay from mention to a human seeing it

Falling, and stable

Response coverage

Share of high-reach items that got the response the grid requires

Rising toward 100 percent

Decisions produced

Items that changed an action, a page, a product or a policy

Rising, and never high in absolute terms


A programme that surfaces 4,000 mentions a month and changes nothing is an expense. A programme that surfaces 400 and changes four things a month is a control system, and it is worth its cost.


The escalation path, written down


Decide in advance, in one paragraph, who is woken up and when. A workable default:


  • The community responder reads every high-reach item within one working day and handles the grid responses.

  • The comms lead is told within two hours about any high-reach item that is factually wrong, or that names a legal, safety or regulatory issue.

  • The responsible function (product, legal, support or security, depending on the claim) is engaged the same day when the item implies a defect, a breach or a breach of contract.

  • Nobody else is copied. A wide distribution list turns a single complaint into a meeting, and the meeting is the delay.


Write the names and the numbers into the same document as the query set. The value of a monitoring pipeline is measured entirely by what happens in the two hours after it fires.


Training the classifier on your own language


A sentiment model trained on general text will misread your customers, because your customers do not speak in general text. They use product names as verbs, they compare you to competitors you know, and they express approval and complaint in the same sentence.


Four steps bring a general classifier closer to your reality, and none of them requires a machine learning team.


  • Collect two hundred real items. Randomly, from your own mentions, across at least six weeks. Not the worst and not the best: a random draw.

  • Label them by hand into three classes. About you positively, about you negatively, not about you at all. The third class is the one most teams forget to label, and it is the one that produces the inflated negative counts.

  • Measure where the classifier disagrees with your labels. The confusion is usually concentrated: sarcasm read as positive, comparison read as negative, a competitor's name inside a quoted reply misattributed to you.

  • Write explicit handling rules for the recurring cases. If sarcasm is the problem, add a rule that a positive label on a phrase containing a known sarcastic marker is downgraded to "review". If misattribution is the problem, require the entity name in the same sentence as the sentiment word.


Record the accuracy of the classifier on your labelled set, and re-measure it every quarter. A classifier without a measured error rate is a decoration, and the error rate is exactly what you need in order to know how much weight to put on any single score.


Aspect-level sentiment: the split that changes the response


A single score for a mention hides the part that matters. The same review can be positive about your support and negative about your pricing, and the aggregate label tells you nothing about which team should act.


Score four aspects separately on every high-reach item:


Aspect

The question it answers

Product or service

Does it do what the customer needed

Support and response

Did somebody help, and how quickly

Price and value

Is the cost defensible against the alternative

Trust and conduct

Do they believe you behave honestly and competently


The split is what turns a dashboard into a work queue. A quarter where product sentiment is flat, support sentiment is rising and price sentiment is falling is a pricing conversation, not a support conversation, and the aggregate score would have hidden it.


Keep the aspect scores in the same row as the item, and read them as trends over a quarter rather than as daily values. Daily aspect scores on small volumes are noise.


Where the numbers stop being trustworthy


Three conditions make sentiment numbers unusable, and each has a visible signature.


  • Very low volume. Below about thirty items a month for a given aspect, a percentage is not a measurement. Report the raw items and quote two or three of them instead.

  • A platform that hides the denominator. If a tool reports sentiment shares without telling you how many items it classified, the percentages cannot be checked. Ask for the counts.

  • A query that captures a homonym. If your brand name matches a place, a person or a common word, an unfiltered query will mix unrelated conversation into your numbers and pull them in a direction that has nothing to do with you. The false positive rate is the check.


When one of those conditions holds, say so in the report rather than publishing the percentage. "Fourteen items this month, eleven about support and three about pricing; no reliable trend yet" is honest and useful. A single number with a hidden denominator is neither.


The quarterly review that keeps the programme useful


Once a quarter, sit down with the query set, the alert rules and the reports from the previous three months, and answer six questions:


  • Which queries produced the most noise and the fewest decisions?

  • Which alert fired most often, and did each firing lead to an action?

  • Which high-reach items went unanswered, and why?

  • What did the aspect scores say, quarter over quarter, for each of the four aspects?

  • What did the manual sample reveal that the classifier got wrong?

  • What is one thing we will stop tracking before the next quarter?


The last question is the load-bearing one. Monitoring programmes fail by accumulation, not by defect: every query sounds reasonable when it is added, and nobody removes anything. A quarterly removal is what keeps the reading time proportionate to the value.


The triage grid, when not to respond, and the weekly report
The triage grid, when not to respond, and the weekly report
Back to the TOC

Feynman Summary: Explain It Like You Are 12


Imagine your football club has a team page and hundreds of people write comments about it.


Some comments say "great match". Some say "the coach is terrible". Some say "I loved the first half but the goalkeeper was awful". If you only count "good words" and "bad words", the last comment goes wrong: it has good words and bad words in the same sentence.


So you do three things. First you gather all the comments in one place, without copying the same person's comment five times. Then you read each one and decide: happy, angry, or a bit of both, and about what? The goalkeeper, or the coach, or the price of tickets? Then you count them by subject and week, so you can see that people are getting angrier about ticket prices but happier about the goalkeeper.


The important part is what you do next. If someone is angry about a real problem, you fix the problem. Counting angry comments without fixing anything changes nothing. And you must remember that only the loudest people write comments. The quiet ones never say anything, so you cannot say "everyone is happy".

Back to the TOC

Mindmap: The Complete Picture


Complete mindmap of the brand monitoring sentiment pipeline
Complete mindmap of the brand monitoring sentiment pipeline

The mindmap shows the four source categories, the three classification methods with their trade-offs, the five pipeline stages, the alert routing rules, the five dashboard decisions, and the seven known failure modes with the human check attached to each.



UNOP isochrone

UNOP Sound (University 365 Neuroscience Oriented Pedagogy)

Take five minutes to consolidate your memory. Play the isochronous tone track (10Hz alpha frequency) with your eyes closed. Alpha-frequency tones after a learning session support consolidation, helping move what you just learned from short-term to long-term memory.

[Audio player: UNOP Post-Lecture Isochrone (10Hz, 5 minutes)]

Back to the TOC

Practical Exercise: Monitor One Brand for 24 Hours


Exercise: One Brand, One Day, Five Numbers


Choose a brand you can observe legally from public pages. Your own is the best choice. If you have no brand, use a small local business whose reviews are public.


  • Write the sampling rule first. One sentence: which pages, which search terms, which language, which time window. Save it.

  • Collect 50 to 200 mentions from two of the four source categories. Copy the text, timestamp, platform, author type and permalink into a spreadsheet.

  • Deduplicate by eye. Record how many mentions collapsed into clusters and what the largest cluster was.

  • Label 100 mentions by hand, if you have them, using the four labels positive, negative, neutral, mixed, plus an aspect and an intent.

  • Run the same 100 through one automated method, using the scoring brief from Step 3 verbatim.

  • Compute agreement. How many labels matched? Write down the three items you disagreed on most and why.

  • Group by theme. Name three recurring subjects and count mentions per theme.

  • Draw the two views you would actually look at each morning: the alert queue and the theme table.

  • Pick one theme and write the sentence you would say to the product or process owner.


What to Observe


  • How many of the 100 items were mixed or sarcastic? Was that more or fewer than you expected?

  • Which failure mode from Step 7 appeared most often in your own sample?

  • Did the automated method get the easy cases right and the hard cases wrong, or did it fail randomly?

  • How many mentions would it have taken before the percentage change became meaningful?

  • Which single theme, if fixed, would remove the largest share of negative mentions?


Applied AI Connection


You chose the sources, wrote the sampling rule, labelled the sample and measured the method against your own judgement. The tools classified at volume and grouped the text. That is the CI-First pattern: CI = HI + (AI x HI). The human sets the standard the machine is measured against, and the machine handles the volume the human cannot.

Back to the TOC

Glossary


Term

Definition

Sentiment analysis

The automated assignment of a positive, negative, neutral or mixed label to a piece of text.

Aspect-based sentiment

Sentiment measured per subject inside a text, such as delivery, product quality and price separately.

Lexicon-based scoring

Sentiment scoring from a dictionary of scored words with negation and intensity rules, fully explainable and context-blind.

Classifier

A model trained on labelled examples to predict a label for new text, more context-aware and less transparent than a lexicon.

Confidence score

The model's own estimate of how sure it is, which matters as much as the label itself.

Mixed sentiment

One text containing both strong positive and strong negative statements, which a single label cannot represent.

Churn risk

An intent signal where the writer states or clearly implies they will stop using the brand.

Deduplication

Collapsing the same content reposted or shared across platforms into one item with a cluster size.

Cluster size

The number of copies one mention produced, a useful reach signal that must not be counted as separate mentions.

Sampling rule

The written definition of which content a collection filter includes and excludes.

Selection bias

The distortion created because people who write about a brand are not a random sample of its customers.

Denominator

The total from which a percentage is drawn, without which a change figure is meaningless.

CI-First

Co-Intelligence First: the U365 principle that the human is the orchestrator and the AI is the amplifier. CI = HI + (AI x HI).

UP-Context-Engineering Method

University 365 Prompting-Context Method: a structured approach to prompting with explicit context blocks for consistent output.

Back to the TOC

Quiz: TEST YOUR UNDERSTANDING


1. Why is aspect-level sentiment more useful than document-level sentiment for brand monitoring?


A) It runs faster


B) It separates feelings about different subjects inside one text, such as delivery versus price


C) It removes the need for deduplication


D) It produces one number instead of several


2. A review reads "the app is excellent, but support never replied". A document-level classifier returns one label. What is lost?


A) Nothing, the label is sufficient


B) The negative support finding, which is the actionable part


C) The author's identity


D) The platform metadata


3. Why deduplicate before counting?


A) To reduce storage cost


B) Because one complaint reposted across platforms would otherwise be counted as many mentions


C) Because platforms require it


D) To make the sentiment score more positive


4. You see "negative mentions up 40 percent". What must accompany that figure?


A) A confidence interval only


B) The denominator, such as 10 to 14 of 300


C) The brand's revenue


D) The competitor's score


5. Which failure mode means your sentiment mix should never be read as a satisfaction survey?


A) Negation scope


B) Sarcasm


C) Selection bias, because only the vocal subset writes publicly


D) Translation loss



Answers: 1-B, 2-B, 3-B, 4-B, 5-C

Back to the TOC

Related Resources


U365 INSIDE Publications



External Resources


  • Brandwatch: Enterprise social listening with aspect-level classification

  • Talkwalker: Consumer intelligence and sentiment dashboards

  • Google Alerts: Free mention monitoring for news and blogs

  • Trustpilot and Google Business Profiles: Structured review sources with export options

  • Hugging Face: Open sentiment classification models you can run on your own hardware


Related U365 Lectures (Coming Soon)


  • Lecture 2: AI for SEO: Content That Ranks (UIC, Marketing Series)

  • Lecture 6: The AI Press Release: Automating PR (UIC, Content Strategy Series)

  • Lecture 10: Crisis Communication in the AI Age (UIC, Branding Series)

Back to the TOC

U.Copilot for This Lecture


Discuss this lecture with U.Copilot, your AI chat companion trained on this content.


Copy and paste the following prompt into the U.Copilot chat on university-365.com:


You are U.Copilot for Lectures, an AI chat companion specially trained on University 365 lecture content. You are helping a Fellow who just completed the lecture "Sentiment Analysis for Brand Monitoring" from the Marketing Series at the U365 Institute of Communication (UIC). Your role is to help the Fellow build and read a brand monitoring pipeline. You can: - Help them write the written sampling rule for their own brand before they collect anything - Compare lexicon, classifier and LLM classification for their specific volume and languages - Review their labelling brief and explain how to measure agreement against a human sample - Design the five dashboard views backwards from the decisions their team actually makes - Write the alert routing rules and the response decision tree for their context - Stress-test any claim they want to make from the data by asking for the count, period, method and agreement Always maintain U365's CI-First approach: the Fellow defines the standard and the machine handles the volume. Remind them to show denominators and to read a random unfiltered sample every week. Use the UP-Context-Engineering Method: provide context-rich, role-aware responses that account for the Fellow's data maturity and goals.

Back to the TOC

Next Steps


Now that you can monitor a brand with real data instead of impressions, here is what to do next:


  • Write your sampling rule today, before you collect anything, and save it where future you can find it

  • Label 100 mentions by hand and keep them: they are the only honest benchmark your pipeline will ever have

  • Build the alert queue first, not the dashboard, so something reaches a person this week

  • Pick one recurring negative theme and take it to a product or process owner with the count attached

  • Continue with Lecture 2 in this series, "AI for SEO: Content That Ranks", to apply the same discipline to search demand

  • Explore the Marketing Series tag on INSIDE for more measurement workflows built on the CI-First approach


The mention you never saw on Tuesday was data. By next Tuesday, it will reach someone who can act on it.

Back to the TOC

IMPORTANT NOTICE


This lecture is published by University 365 as part of its INSIDE Publications Hub. The content is free to read for all visitors. Lectures in this series may be part of a structured academic program leading to a Micro-Credential for your Career (MCC). To enroll in an academic program, visit university-365.com/tuition.


This content is for educational purposes. While we strive for accuracy, platform access terms and tool pricing change frequently. Verify current legal terms and pricing against primary sources before professional use.


Copyright University 365, Inc. All rights reserved. This content is protected under University 365's copyright policies. For permissions or inquiries, contact uda@university-365.com.


Published by the Department of Academics, University 365.

Lecture delivered by the University 365 Institute of Communication (UIC).

Lea Loringam, Dean of Communication, UIC

Signed for the academic year 2026.

Comments

Rated 0 out of 5 stars.
No ratings yet

Add a rating
Image by Erik  Lucatero

Become Superhuman

Master AI to stay irreplaceable in every field.

 

 

 

​

​

Apply for Admission Today.
Select Your Initial Access Level.


Become a DISCOVERY, INSIDER, or SUPERHUMAN Fellow.

Image by Milad Fakurian

Master Your Life with a Digital Second Brain

Turn overwhelm into clarity with LIPS + CARE
U365’s unique framework to organize your goals, projects, and knowledge into a superhuman system for success

bottom of page