# An AI FollowThatPage alternative: rules that read meaning, not lines > Keyword line filters decide by string match. A language-model monitor like AyeWatch decides whether a change matches what you asked for. URL: https://claudegpt.si/news/ai-followthatpage-alternative Published: 2026-10-24T18:25:40.047Z Updated: 2026-10-25T18:25:39.885Z Author: ClaudeGPT Si Newsroom Section: AI Tools Topics: FollowThatPage, AI monitoring, Website change monitoring Publisher: ClaudeGPT Si ## Key takeaways - FollowThatPage decides relevance with keyword line and block filters; AyeWatch reads changes against a sentence you write. - AyeWatch writes a short summary for each alert and can apply one rule to several pages, topics or social accounts. - AI assistants can manage AyeWatch monitors through its MCP server and REST API on paid plans. An AI FollowThatPage alternative is a useful way to see what language models change about a very old problem. Follow That Page has been telling people when web pages change for years, using an approach that predates large language models: fetch the page, turn it into text, compare it with the last copy, email the difference. AyeWatch does the same first three steps and replaces the last one with a question a model answers: does this change match what the user asked for? This piece looks at how the two approaches decide relevance, and where each one wins. ## The classic method: diff, then filter Follow That Page's robot downloads the HTML, converts it to UTF-8 plain text and compares it with the previous version. Anything added or removed goes into an email report. Raw diffs are noisy, so it offers two filters. A line filter drops every line containing a keyword, or keeps only the lines that contain one. If a blog shows "five comments" and that keeps changing, you filter on "comments". If you only care about a price, you keep lines containing "price". A block filter does the same with a stretch of text between two markers. This is fully deterministic. You can predict exactly what will trigger an email. It is also literal. The filter has no idea what "price" means; it only knows the string. If the site rewrites "Price" as "Now only", or moves the figure onto its own line, the filter either misses the change or lets in noise. The robot also does not run JavaScript, so content drawn by scripts may never reach the comparison. ## The model-based method: diff, then judge AyeWatch also notices that a page changed. The difference is what decides whether you hear about it. Instead of keyword filters, you write a rule in plain language: > "Alert me when the council publishes a new date for the planning hearing. Ignore agenda formatting and minutes from past meetings." When the page changes, AyeWatch reads the change against that rule and decides whether it qualifies. If it does, the alert is a short written summary of what changed, not a list of lines. If it does not, nothing is sent. Three practical effects follow from that design: 1. **Wording changes stop breaking rules.** "Hearing date" can move or be rephrased and the rule still applies, because it describes meaning rather than position or spelling. 2. **One rule can span several sources.** [AyeWatch](https://ayewatch.ai) lets a single monitor cover several URLs, and topic monitors extend the same idea to news, forums, blogs and government pages across more than a billion sources, plus X, Reddit, YouTube, LinkedIn, Instagram and TikTok. 3. **Alerts carry context.** A summary saves the step of reading a diff and working out what it means. ## What you give up Model judgment is not deterministic in the way a keyword filter is. A filter will always behave the same way on the same text; a rule interpreted by a model is closer to asking a careful reader. For most monitoring that is the point, but if you need a guarantee that every single character change produces an alert, a plain diff is the more literal tool. Follow That Page, for a quiet page where any edit matters, does exactly that and costs nothing on its basic account. ## Monitors as something agents can manage The other shift is who sets the monitors up. Follow That Page is a form and an email inbox. AyeWatch exposes its monitors through a REST API on every paid plan, with signed `topic.update` webhook events that carry the AI summary. It also runs an MCP server at `https://ayewatch.ai/api/mcp`, so an assistant connected over the Model Context Protocol can create, change or delete monitors on your behalf. A research agent could set a watch on a regulator's guidance page, and the summaries would come back as structured events. ## Delivery and cost Alerts appear in AyeWatch's feed and as push notifications on iOS and Android on every plan. Paid plans add email, a daily digest email, Slack, Discord and webhooks. Pro costs $9 a month with hourly monitoring; Pro+ at $19 adds ASAP mode. Each check spends monitoring credits, most at one credit, and the app estimates daily usage before you save, so running a rule more often costs more. A Free Preview offers three runs across up to three topics, no card required. ## The takeaway Follow That Page shows how far a simple, honest diff can go: it has served people for years with an email and a keyword filter. A language model adds the step those filters were always approximating, which is reading the change and deciding whether it matters. For pages where most edits are noise, that step is the whole value. AyeWatch's [comparison with Follow That Page](https://ayewatch.ai/vs/followthatpage) lays out the feature differences in a table. ## Frequently asked questions ### How is an AI page monitor different from FollowThatPage? FollowThatPage emails the lines that were added or removed, minus any you filtered. AyeWatch checks whether the change matches your rule and summarizes it. ### Can an AI agent create page monitors? Yes. AyeWatch's MCP server at https://ayewatch.ai/api/mcp, on paid plans, lets a connected assistant create, update or delete monitors. ## Sources - [AyeWatch plans and features](https://ayewatch.ai/), AyeWatch - [6 Best FollowThatPage Alternatives in 2026](https://ayewatch.ai/blog/followthatpage-alternative), AyeWatch - [AyeWatch vs FollowThatPage](https://ayewatch.ai/vs/followthatpage), AyeWatch - [Follow That Page FAQ](https://www.followthatpage.com/faq), Follow That Page