Brand Monitoring: From Public Mention to Published Proof

Image URL

Do not index

AI summary

Brand monitoring involves actively searching for public mentions of your product, triaging each match to decide whether to publish, use for product insights, support, or ignore, then obtaining permission before republishing as proof on your site; it requires a focused keyword list, deduplication, sentiment tagging, and a workflow that turns alerts into usable assets while avoiding legal and privacy pitfalls.

Title

Brand Monitoring: From Public Mention to Published Proof

Date

Description

Brand monitoring that doesn't stop at the alert. Find unsolicited mentions, ask permission, and publish them as proof.

Status

In progress

Current Column

Ready to review

Person

Writer

Someone on Reddit named your product without being asked.

You screenshot it. It is still in Photos.

A customer tweeted you last quarter. You liked the post. A stranger on the pricing page will never see it.

Brand monitoring is watching public surfaces for people naming you. The useful version of that job does not stop at the alert. It ends with proof on your site — original author, original words — or a decision to leave the mention alone.

The ask is a different pipeline. If that is the job, start with the whole testimonial pipeline. This page is the mention you did not request.

What brand monitoring is (and is not)

Brand monitoring is a search. You pick a handful of words. You watch the public places those words appear. You decide what to do with each match.

It is not a Google review programme. Structured reviews on Google, G2 or Capterra live under someone else's ranking and someone else's takedown rules. The difference between reviews and testimonials is that split. "Monitoring your Google reviews" is the other job.

It is also not a "best social listening tools" roundup. Those pages survey dashboards for PR and community teams. This page is for a founder or a marketer who wants unsolicited public praise to become something a prospect can see. I am not ranking listening platforms here.

Crisis response is a slice of the job, not the whole job. A negative spike is worth seeing the day it happens. Most of the mentions you will actually use are quiet and positive, and they die because nobody moved them.

Why most programmes stop at the alert

Three ways this dies.

You never saw it. A Reddit thread, an HN comment, a LinkedIn post that did not tag you. The mention existed for a day. By Monday the thread has moved.

You saw it and did nothing. The screenshot sits in Slack. Nobody owns "turn this into a site asset," so it does not become one.

You published it without asking. Cropped screenshot, no name, no link, no permission. That is the version that gets you a DM you did not want, and it is the version the legal side of testimonials exists to prevent.

Almost every guide on this topic ends at "set up alerts and respond." Fine for a comms team. Wrong for a proof programme. An alert is a ticket. The asset is the mention on a Wall of Love, with the original author, the original link, and a screenshot of the original post.

Four beats, not two:

Find the match.

Triage it — publish, product, support, or ignore.

Ask before it lives on your site.

Publish it, or leave it.

Skip permission and you do not have a testimonial. You have a scrape. Skip publish and you have a folder of screenshots. Skip triage and the inbox is noise.

Design Bot — diagram-mention-to-wall-of-love-light.png. What: a four-step flow, left to right — mention surfaces → triage (publish / product / support / ignore) → permission email → Wall of Love card with author, link, screenshot. The permission step is visually longer than the others; do not draw four equal boxes. Why here: the angle is that the job does not stop at the alert, and the drawing has to show the extra beat. Palette: purple-600 #5D5DFF, gray-900 #151719, gray-500 #707D86. Light theme. Naming per the assets README.README.

Where this framing breaks. Some mentions should never be republished. A complaint that only makes sense inside a 40-comment thread. A joke. A customer who would be furious to see their face next to a pricing button. The pipeline is allowed to end at "leave it." That is triage working, not triage failing.

I would rather have twelve mentions a month and publish two than a dashboard of 800 that nobody reads.

What to watch (you do not get unlimited keywords)

Brand Monitor on Ultimate tracks 10 keywords. Ultimate + tracks 15. As of 31 August 2026, that is the cap on the feature page — a real limit, not a suggestion. Spend them.

The list I would start with, in order:

Exact product or company name. The thing a stranger actually types.

The obvious misspelling. The one you wince at in search logs.

The founder or public-face name. People mention people more reliably than products.

One category phrase a buyer uses. "Social proof tool" if that is you — not five synonyms of the same idea.

One competitor handle, only if you will do something with those matches. Lurking is not a use. A comparison post you can ask to be in, is.

Drop campaign hashtags unless you are in the middle of the campaign. Drop the CEO's childhood nickname. Drop every synonym of your category that will match every thinkpiece published this week.

A small team cannot watch everything. Pretending to is how the inbox becomes noise. The rest of keyword strategy is a later post. The constraint is the point of this one.

Where people actually name you

The public surfaces that matter for a B2B product are not symmetric. Treat them as different rooms, not one firehose.

X. Fast, public, easy to search, easy to vanish. A post already looks like a quote, which is why it makes a good card — and why the author will notice if you republish it.

LinkedIn. Slower, more professional, worse public search. Company-page mentions and personal posts are different objects. A lot of the good ones never use your product name in a way a keyword will catch.

Reddit. High signal when it is real. A reply from the brand account is often the wrong move. Subreddit rules on self-promotion are not decorative. A positive mention you did not ask for is still not an invitation to drop a collection-page link.

Hacker News. Thin volume, high intensity. The culture penalty for a defensive reply is real. Read twice. Reply once, or not at all.

YouTube and TikTok. Longer artifacts. A 40-second clip of someone using the product is proof you cannot write. Also harder to drop into a grid without looking like a stolen video.

The open web. Round-ups, comparison posts, "best X" lists. Useful. Usually not something you republish as a testimonial. You open the page. You decide whether to ask the author for a line. Different motion.

Brand Monitor watches those seven: X, LinkedIn, Reddit, Hacker News, YouTube, TikTok, and web search. As of 31 August 2026, that is the list on the feature page.

It does not watch Discord, Slack communities, email, or what an AI answer says about you. If a news wire or AI-answer tracking is the job, this is the wrong page. We do not do that.

The free path, if you want nothing from us

You can run a crude version of this with tools you already have.

Google Alerts on the company name and the misspelling. It emails you. It is late, and it misses most social.

Site search: site:reddit.com "your product", and HN Search. Manual. You will forget to run it.

Native search on X and LinkedIn. You see what the platform wants to show you. You will not see everything.

Three problems with that stack:

It is a reminder, not an inbox. Nothing is deduplicated. The same round-up will hit you three times.

Nothing in it turns a match into a site asset. You still screenshot, still ask, still paste.

You will not run it on a Friday.

That is enough for a one-person company named twice a month. It is not a programme.

Triage

Every match is one of four jobs. Pick one. Do not leave it in "I will deal with this."

Publish candidate. Specific, public, positive, and the person looks like a customer a prospect would recognise as "someone like me." What makes a testimonial good is specificity in a small space; the same bar applies here. These are rare. Treat them that way.

Product note. Mixed, or a feature request wearing a mention. Do not put it on the site. Do put it in the product channel.

Support. Negative, or a billing complaint in public. Reply as a person. Take it off the thread if they will come with you. Do not hide the original. Routing a detractor to support is legitimate. Suppressing the mention is not, and I will not write that.

Ignore. The job listing that matched your company name. The same round-up, again. The bot. This is most of the volume. Bulk-ignore exists for it.

Automated sentiment — positive, neutral, negative, mixed — is a sort order. It is not a verdict. Sarcasm, jargon, and "this is fine" will fool it. Read the ones you might publish.

Permission before you republish

A public post is public. Putting it on your marketing site, next to a pricing button, is a different act.

Ask. In writing. Be specific about where it will appear.

Ask separately for:

The words

The name and role

The company

The face or avatar

The original link, kept visible

Plenty of people will give you the words and not the photo. A card with a role, a company, and a link back still works. A cropped screenshot with no attribution barely does.

If anything changed hands — a discount, a free month, a gift card — the connection has to be disclosed. That is not a nicety. In the US, the FTC Endorsement Guides (updated June 2023) treat a material connection as something a significant minority of readers would want to know. Check your own jurisdiction. The legal side of testimonials is the primer, not legal advice.

Do not scrape a platform against its terms to get the quote. Search the platform, or use a product that already has access. If it will live on your site as proof, you asked.

The ask can be short:

Subject: can I put this on our site?

Hi [Name] —

You wrote [the specific line] on [platform]. It is the most useful sentence I have heard about [task] this quarter.

May I add it to our Wall of Love, with your name, [company], and a link back to the original?

If you'd rather I anonymise the company, say so. If you'd rather I didn't use it, also say so.

— [Your name]

If they say no, you leave it. That is the whole policy.

The social half of asking — how not to make it weird — is how to ask for a testimonial. Use that when you still need a quote that does not exist yet.

Once a mention is approved, the Wall of Love guide is the display half: where the card actually sits on the site.

Doing this with Testimonial

The Google Alert plus the email above is a working system. You do not need us for it.

What Brand Monitor changes is the gap between "it was said in public" and "it is sitting in an inbox you will actually open."

You save keywords. The first scan starts when you switch it on — including an all-time sweep of the web and TikTok, so you do not begin from empty. Matches from X, LinkedIn, Reddit, Hacker News, YouTube, TikTok and web search land in one inbox, sentiment-tagged and deduplicated. Filter by source if you only want Reddit today. Filter by sentiment if you are hunting for Wall of Love candidates.

Approve a mention and it imports onto the social-media inbox and the Wall of Love — original author, original link, screenshot intact. Ignore or delete the rest. Bulk-ignore the noise. Mention volume and sentiment sit on 7 / 30 / 90-day windows, so a spike is visible the week it happens.

Design Bot — brand-monitor-inbox-light.png. What: Brand Monitor inbox, light theme, desktop, 1440×900 capture cropped to the inbox. A handful of mentions with positive / neutral / negative / mixed tags visible; one row showing the "Add to Wall of Love" action; source labels (X, Reddit, etc.) readable. Redact any real customer names, handles, avatars. Why here: the reader has just been told the loop ends at published proof; this is the screen that loop lives on. Naming per the assets README.README.

Brand Monitor is on Ultimate and above. As of 31 August 2026, Ultimate is $50/month/space ($600 billed annually) with 10 keywords and 2,000 mentions a month. Ultimate + is $95/month/space ($1,140 a year) with 15 keywords and 5,000 mentions. Ultimate and Ultimate + are priced per space. The plans and pricing page is the live list. Mention credits can be topped up; I would treat that as a relief valve, not a plan.

If you already have the URL of a specific post, you do not need Brand Monitor to import it. Free includes unlimited X and LinkedIn imports. Brand Monitor is the watch, not the paste.

Where it breaks, including for us.

Ten keywords will not cover every misspelling, every product, and every competitor. I would pick the five above and stop. The product will not stop you wasting slots on vanity terms.

Two thousand mentions sounds like a lot until a category round-up matches every week. Dedup helps. It does not make a noisy keyword smart.

Sentiment will mis-tag. Read before you publish.

The one-click import does not ask permission for you. Approval in the product is "I want this on the Wall of Love." Approval from the person is still an email. Skip that and you have a pretty card you should not have shipped.

It does not watch Discord, email, or AI answers. Do not buy it expecting a media-monitoring desk.

If you are named twice a month and you already run Google Alerts on a Friday, stay there. The tool earns its place when mentions arrive faster than you can search for them, and some of them should be on the site.