Finding Out Last: Media Monitoring for Small Businesses on a Real Budget

Picture finding out your business was mentioned in a trade publication three weeks after it ran. Not from the journalist, not from a Google Alert, but from a supplier who happened to read the magazine in a waiting room. The piece is positive. It names your business as an example of good practice in your industry. You could have put it on the website, pitched a follow-up, told your team, shared it with the client it was actually about. Instead, by the time you know it exists, it’s old news everywhere except in your own head.

That’s the quiet cost of not monitoring your own coverage. Rarely a disaster, just a missed opportunity, repeated often enough that it adds up. Somebody mentions your business in a podcast. A customer vents on a Facebook group you’re not in. A competitor’s announcement changes what a journalist wants from you next. None of it reaches you unless something is actively watching for it, and “I’ll just notice” has never once worked as a system.

This post is about building that system without the enterprise price tag. It covers what’s actually worth monitoring, a weekend workflow using AI chatbots and free tools, what the paid options cost when you outgrow the DIY version, and the specific point where AI stops being able to help.

What Is Media Monitoring, and Does a Small Business Actually Need It?

Media monitoring is the ongoing practice of tracking where and how your business gets mentioned, across news sites, blogs, podcasts, social platforms, and review sites, so you know about it close to when it happens rather than weeks later. Any business that has ever pitched media, cares what customers say publicly, or operates where a competitor’s move affects it needs some version of this. The question isn’t whether to monitor. It’s how much system the size of the business actually justifies.

A solo consultant with no media profile yet can get by on a handful of free alerts. A business that pitches journalists monthly, fields public reviews, and competes in a crowded category needs something closer to a weekly habit, with a tool doing the catching so a person only has to do the reading.

What You’re Actually Trying to Catch

Four things matter more than the rest, and they need different kinds of watching.

Earned media mentions: articles, podcast appearances, and broadcast segments that name the business, a spokesperson, or a product. This is the one most small businesses think of first and the one that’s easiest to miss, because most coverage doesn’t come with a courtesy email. Understanding why journalists run some releases and delete others helps explain which coverage is worth watching for.

Public reviews and complaints: Google reviews, Trustpilot, industry forums, and social posts where a customer names the business without tagging it. A complaint caught within a day is a service recovery. The same complaint caught three weeks later, after it’s been screenshotted and shared twenty times, is a different problem entirely.

Competitor and category moves: a rival’s funding announcement, product launch, or executive hire that changes the story a journalist is interested in telling about the sector. Missing this doesn’t cost you coverage directly. It costs you the chance to be part of the story while it’s still being written.

Being named inside an AI answer: when someone asks ChatGPT, Gemini, or Claude a question in the business’s category and the business’s name comes up, or doesn’t, in the answer. This is a newer category of mention and a genuinely strange one, because there’s no link to click and no notification waiting. It only shows up if someone goes looking, which is covered in more depth in a related post on this.

The Workflow: A DIY Monitoring System Built in a Weekend

This assumes a solo business or small team with no existing tool, using free alerts, manual checks, and a general AI chatbot to do the reading so a person doesn’t have to scan everything by hand.

1. Set up free alerts for every name variant (20 minutes). Create a Google Alert for the exact business name, the founder’s name, and any product names, each in quotes for an exact match. Add a misspelling or two if the name is commonly typed wrong. Set delivery to “as-it-happens” rather than daily digest, because a digest buries a time-sensitive mention among routine ones.

2. Set up a standing search on social platforms (15 minutes). Save a search for the business name on X and LinkedIn, and check whichever platform the audience actually uses, directly in-app, since a saved search costs nothing and catches posts that never show up in a general alert.

3. Check reviews weekly, not when you remember (10 minutes a week). Claim the Google Business Profile and Trustpilot listing if the business doesn’t have them set up, then check both on a fixed day each week rather than waiting for a notification that may not arrive.

4. Build a weekly AI-assisted digest (15–20 minutes a week). This is where a chatbot earns its place. Paste in whatever turned up that week, links, review text, social posts, and ask for a short summary sorted by what needs a response, what’s worth sharing, and what’s just noise. A chatbot with web search enabled, such as Claude or ChatGPT’s browsing mode, can also run a fresh search on request and summarise what it finds, though it won’t catch everything a dedicated tool would.

5. Ask the AI chatbot directly what it says about you, monthly (10 minutes). Separately from monitoring the open web, ask the chatbot itself what it would tell someone researching the business in its category. This isn’t monitoring in the traditional sense, since you’re checking the AI’s own answer rather than something it found elsewhere, but it’s increasingly where some buyers start.

6. Log anything worth acting on in one place (5 minutes). A simple spreadsheet with date, source, link, and action taken is enough. The point isn’t a sophisticated system. It’s not relying on memory for something that happened three weeks ago.

Total time for a first setup: around an hour. Ongoing time: 25 to 35 minutes a week, mostly the review check and the digest.

A copyable starting prompt for the weekly digest:

“Here’s everything I found this week mentioning [business name]: [paste links, review text, social posts]. Sort these into three groups: needs a response within 48 hours, worth sharing or amplifying, and no action needed. For anything needing a response, draft a short, calm first line I can build on, not a full reply.”

Can AI Monitor the Internet for You, Continuously, on Its Own?

No, not reliably, and not without being asked. A general AI chatbot doesn’t run in the background, checking the web every hour and flagging what it finds. It answers when prompted, and even with web search switched on, it searches once, for that one request, rather than maintaining a standing watch the way a dedicated monitoring tool does.

That gap is exactly what free alerts and saved searches in the workflow above are there to cover. Treat the chatbot as the analyst reading and sorting what those feeds already caught, while the alerts and saved searches stay the ones doing the actual catching.

Tool Comparison: DIY Stack vs Brand24 vs Mention vs Isentia

The DIY stack above (Google Alerts, saved social searches, a weekly chatbot digest) costs nothing beyond time and covers most of what a solo business or very small team needs. Its limits show up fast with volume: it misses mentions behind paywalls, it won’t catch broadcast or print coverage at all, and reading everything by hand stops being a 30-minute job once the business gets genuinely busy in the media.

Brand24 is the first serious step up, built for social listening and brand mentions across the open web, with sentiment scoring and basic reporting. Current published pricing runs from US$199 a month for the Individual tier (billed annually; US$249 billed month-to-month) up to US$599 a month for the Business tier, with Enterprise starting from US$1,499 a month for higher volumes and custom needs, according to brand24.com’s pricing page, correct at the time of writing. That’s a genuine cost increase from where tools like this sat a few years ago, and worth checking directly before budgeting, since plans shift.

Mention’s pricing page now shows a single Company Plan with demo and trial buttons, according to mention.com/en/pricing, checked 23 October 2026. That page doesn’t show a confirmed price, so get a direct quote before budgeting.

Isentia sits at the top of this comparison, an established media monitoring provider covering TV, radio, print, online, and social with a dedicated account manager. It doesn’t publish pricing at all; a quote comes from a sales conversation, according to isentia.com, correct at the time of writing. It suits a business with real broadcast exposure and a team to act on what it finds. For almost everyone else reading this, it’s serious overkill.

What This Actually Costs

The free version costs nothing beyond the hour of setup and the weekly digest time: Google Alerts, native social search, and a chatbot subscription the business may already have, such as Claude Pro at US$20 a month billed monthly, according to claude.com’s pricing page, correct at the time of writing.

The first paid step up, Brand24’s Individual tier, runs US$199 a month billed annually. That’s worth it once manual checking is eating more than an hour or two a week, or once the business needs to prove ROI on PR activity to someone else in the room.

Enterprise monitoring through Isentia or Mention’s current offering suits organisations with a comms team and a genuine broadcast or crisis-monitoring need. A solo operator checking in once a week is unlikely to be that organisation. If a quote comes back and the number makes the DIY stack look perfectly reasonable again, that’s usually the right read.

What Not to Do With Media Monitoring

Don’t set it up once and never check it. An alert that’s been quietly filtering into an unread folder for six months is worse than no alert, because it creates the false comfort of thinking something is watching when nothing actually is.

Don’t respond to a negative review or comment within minutes of seeing it. A fast response reads as defensive. A same-day, considered one reads as a business that’s paying attention and taking it seriously.

Don’t treat every mention as equal. A passing reference in a listicle and a dedicated feature both show up in an alert the same way. Spend the triage time the workflow above describes before deciding what’s worth a reaction.

Don’t rely on a single AI chatbot query as proof of what’s being said about the business across the internet. It answers from what it can access in that moment, not a complete record, and a different chatbot or a repeat query can turn up something different.

Don’t skip checking AI-generated answers just because they feel less tangible than a news article. A customer who asks ChatGPT for a recommendation in your category and gets a competitor’s name instead of yours never tells you that happened. It just quietly didn’t become a lead.

What AI Can’t Do Here

AI can’t crawl the web continuously and alert you the moment something new appears. That’s a dedicated monitoring tool’s job, and no amount of clever prompting turns a chatbot into one.

AI can’t access paywalled coverage, most broadcast transcripts, or anything behind a login, which means a chatbot-based check misses a meaningful slice of exactly the coverage worth knowing about.

AI can’t judge tone with full reliability. It can flag that a review reads as negative, but it can miss sarcasm, local idiom, or a complaint that’s really about something else entirely. A person still needs to read anything flagged as sensitive before deciding how to respond.

And AI can’t decide which mentions matter enough to act on. It can sort and summarise, but the judgement call on whether to reply, pitch a follow-up, or let something go still sits with someone who knows the business and the relationship behind the mention.