n8n Tutorial: Competitor Intelligence Monitor — Daily AI Briefings on Autopilot Automation

Right now, one of your competitors just dropped their price by 15%. Another shipped a feature your prospects have been asking you for. A third launched a Black Friday promo that's pulling deals out of your pipeline. You'll find out about all of it eventually — from a lost deal, a churned account, or a sales rep forwarding you a screenshot two weeks too late. Competitor intelligence isn't hard because the information is hidden. It's public. It's hard because nobody has time to check twenty pricing pages, changelogs, and blogs every single morning. This n8n workflow does exactly that, and hands you a briefing before your first coffee.

The problem: manual monitoring doesn't scale, and neither does your attention

Most teams "monitor competitors" the way people "go to the gym in January." Someone bookmarks a few pricing pages, checks them sporadically, and the discipline evaporates within a month. The result is a blind spot that grows in direct proportion to how fast your market moves.

The failure modes are predictable:

• Latency. A competitor changes pricing on a Tuesday; you notice when a deal stalls three weeks later.
• Noise. Even when you do check, you're reading full pages to find the one line that changed.
• No memory. Nobody remembers what the pricing was last quarter, so you can't spot trends — only snapshots.
• No distribution. The one person who checks doesn't reliably tell the five people who need to know.

You don't need a $2,000/month competitive-intelligence SaaS to fix this. You need a scheduled job that scrapes the pages you care about, detects what actually changed, has an LLM summarize why it matters, and drops a briefing into the channel your team already reads.

The solution: a diff-and-summarize loop that runs while you sleep

The Competitor Intelligence Monitor is a scheduled n8n workflow built around one simple idea: store yesterday's snapshot, fetch today's, and only alert on the delta. That "only on delta" part is what separates a useful briefing from a daily spam email everyone mutes.

At a high level, the workflow does five things every morning:

1. Wakes up on a Schedule Trigger.
2. Fetches each competitor's pricing page, changelog, or blog via HTTP Request.
3. Compares the new content against the last stored version.
4. Summarizes the meaningful changes with an AI node — pricing moves, new features, promotions — and rates their strategic impact.
5. Delivers a clean briefing to Slack, email, or a database, then saves today's snapshot as the new baseline.

Because the AI only ever sees what changed, your token costs stay tiny and the briefing stays signal-dense. On a quiet day, it tells you "no material changes across 8 tracked competitors." On a loud day, it tells you exactly who moved and what it means.

Step-by-step: building it in n8n

Here's the node-by-node structure. If you're installing the ready-made template, this is what you're looking at under the hood — and what to configure.

1. Schedule Trigger

Add a Schedule Trigger node set to a Cron expression like 0 7 * * 1-5 — 7:00 AM, weekdays only. Set the timezone in the node so your briefing lands before standup, not at 2 AM.

2. Define your targets

Use a Set (or Code) node to hold an array of the pages you track. Keep it structured so the rest of the workflow can loop cleanly:

[{ "name": "Acme", "url": "https://acme.com/pricing", "type": "pricing" }, { "name": "Beta", "url": "https://beta.io/changelog", "type": "changelog" }]

Follow it with a Split In Batches (Loop Over Items) node so each competitor is processed individually and one broken URL doesn't sink the whole run.

3. HTTP Request — fetch the page

An HTTP Request node pulls each URL. Set Response Format to string and add a realistic User-Agent header so you're not served a bot-block page. Turn on Continue On Fail here so a single 403 doesn't kill the workflow — you want the other seven competitors to still report.

4. Extract and normalize

Add an HTML Extract node to strip navigation, scripts, and footers, pulling only the main content region (target the pricing table or changelog container by CSS selector). Then a small Code node to collapse whitespace. Normalizing matters: without it, a reordered ad or a rotating testimonial reads as a "change" and floods you with false positives.

5. Compare against the last snapshot

Read the previous version from storage — an n8n Data Table, a Google Sheets row, Postgres, or Redis all work. A Code node diffs old vs. new (a simple length + hash check to gate, then a line-level diff for the payload). If nothing changed, route the item to a no-op branch with an IF node. Only changed items proceed to the expensive AI step.

6. AI summarization

Feed the diff into an AI Agent or Basic LLM Chain node backed by an Anthropic Claude model (Claude Haiku 4.5 is fast and cheap for this; Sonnet if you want sharper strategic read). Prompt it tightly:

"You are a competitive analyst. Given the previous and current version of {{name}}'s {{type}} page, describe only what materially changed: pricing, tiers, features, or promotions. Rate strategic impact High/Medium/Low. Ignore cosmetic edits. If nothing material changed, reply 'NO CHANGE'."

Use a low temperature and cap output length. You want a crisp bullet, not an essay.

7. Aggregate and deliver

Collect all competitor results with a Merge or aggregate Code node into one briefing. Send it via the Slack node (Block Kit for a clean layout), the Gmail/SMTP node, or post to a database for a dashboard. Finish by writing today's normalized content back to storage as the new baseline — this write-back is the step people forget, and without it every day looks like a total change.

What you actually get out of it

The payoff isn't "a robot reads websites." It's a set of concrete operational wins:

• Speed to react. When a competitor cuts prices, your sales team has talking points that morning — not after losing three deals.
• A written record. Because every snapshot is stored, you can reconstruct a competitor's pricing history and spot patterns (do they always discount at quarter-end?).
• Zero recurring cost beyond tokens. No per-seat SaaS. The AI only touches deltas, so a full week of monitoring across a dozen competitors costs cents.
• Consistent distribution. The briefing reaches the whole team automatically, so intelligence stops living in one person's browser tabs.
• Focus. Quiet days say "nothing material changed," which is itself valuable — it frees you to stop worrying and ship your roadmap.

Common pitfalls (and how to dodge them)

Most implementations fail for the same handful of reasons. Handle these up front:

• Diffing raw HTML. Pages carry rotating banners, CSRF tokens, and timestamps that change on every load. Always normalize to text and target a specific content selector before comparing, or you'll drown in false alerts.
• Getting bot-blocked. Some pricing pages sit behind Cloudflare or render client-side. Set a proper User-Agent, and for JavaScript-heavy pages swap the HTTP Request node for a rendering service or the Playwright/Puppeteer approach. Keep Continue On Fail on so one hard target doesn't break the run.
• Forgetting the write-back. If you don't save today's snapshot as the new baseline, tomorrow diffs against a stale version and either re-alerts the same change forever or misses new ones. This is the single most common bug.
• Over-alerting. Don't send every micro-change. Let the AI gate on "material" and route Low-impact items to a weekly digest instead of the daily ping. A briefing people trust is one they still read in month three.
• No error visibility. Add an error-handling branch (or an n8n Error Trigger workflow) that pings you if a target has failed for several days straight — a silently broken monitor is worse than none, because you'll assume "no news is good news."
• Scraping things you shouldn't. Stick to public pages, respect robots.txt and rate limits, and space out requests. This is competitive intelligence, not a load test.

Get those right and you have a system that quietly compounds: every day it runs, your competitive picture gets sharper and your reaction time gets shorter — with no ongoing effort from anyone on the team.

PRONTO PARA USAR

Ja construimos isso pra voce

Nao comece do zero. O Competitor Intelligence Monitor — Daily AI Briefings on Autopilot e um workflow n8n pronto para instalar — conecta suas ferramentas em minutos, sem codigo.

Instalar por $59 →