Skip to main content
Competitive landscapes move fast. New pricing pages, product announcements, G2 reviews, hiring signals — by the time someone notices, it’s already in the next deal cycle. Hindsight’s AI Agent gives you a way to monitor competitors systematically. Not a news digest. A research agent you configure, pointed at exactly the sources you care about, running on a schedule and pushing findings to your team.

​
How It Works

The AI Agent has a set of tools it can use to research competitors from the web in real time:
  • Web Search — live search for news, product updates, announcements, and competitive signals
  • Scrape URL — pulls and analyzes the full content of a specific page (pricing pages, docs, changelogs, blog posts)
  • Map URL (experimental) — discovers and indexes subpages from a root URL, useful for finding everything relevant on a competitor’s site
  • Browse Tool (experimental) — full browser access for sources that require it, like LinkedIn
These tools let the agent go beyond keyword matching. It reads pages, extracts signals, and synthesizes findings — the same way a researcher would, just automated and on cadence. You can invoke these tools directly in a prompt, or let the agent decide which ones to use based on your instructions.

​
The Default Skill

When you set up competitor monitoring, Hindsight creates a default skill as a starting point. It runs a general web search for recent activity on your tracked competitors. The default is intentionally generic. It works, but it might not find the signals that matter most to your team, or for a specific competitor. The intent is for you to customize it. Customize the skill and tell the agent exactly where to look. Examples:
  • Point it at a competitor’s changelog URL to track product releases
  • Give it a G2 search query to surface new reviews
  • Direct it to a competitor’s hiring page to catch team expansion signals
  • Set it to scrape a pricing page weekly so you catch changes before your reps do
The more specific your instructions, the more useful the output. A skill that says “research Competitor X” will return noise. A skill that says “scrape competitor-x.com/changelog and summarize new features added this month” will return signal.

​
Customizing a Skill

Once you find an approach that returns useful findings, save it. On the Skills page, add a Hindsight-created skill to your library, then edit your copy. Give it a name, edit the prompt, and save it as a reusable starting point. Build one skill per competitor, or one per signal type (e.g., “pricing changes,” “product releases,” “hiring signals”). Start from the same base skill for each competitor, then tailor the target, sources, and output format while keeping the procedure consistent. To create or customize a skill:
  1. Go to Skills in Hindsight
  2. Add a relevant Hindsight-created skill, or click Create skill to start fresh
  3. Edit the prompt with the sources and output format you need
  4. Run it in Hindsight to test
  5. Save when you’re happy with the output

​
Running on a Cadence with Workflows

A skill you run once is useful. A workflow that runs every Monday at 8am and posts findings to Slack is a competitive monitoring program. Use Workflows to automate your prompts:
  • Schedule recurring runs — daily, weekly, or on a custom cadence
  • Deliver to Slack — push summaries to a channel like #competitive-intel or DM a specific person
  • Save to your intel library — automatically update competitor profiles in Intel when new information is found
  • Connect to external tools — use Zapier to route outputs to your CRM or docs
This is how you keep battlecards current without a person owning it manually. The workflow runs, finds something new, and either surfaces it to your team or updates the record directly.

​
Why This Approach Works

Most competitive monitoring tools give you a firehose: keyword alerts, news feeds, review notifications. You still have to read everything and decide what matters. Hindsight’s AI Agent is different in a few ways: It reads, not just crawls. The Scrape URL and Browse tools extract and analyze content, not just check if a page changed. You get a summary of what’s new, not a raw diff. You control the sources. Generic web search is a starting point. The real value is pointing the agent at the exact pages you care about: a competitor’s pricing page, their docs changelog, their LinkedIn jobs board. You get the signal, not the noise. It connects to your deal data. Once new intel is saved to your library, the AI assistant can reference it when answering rep questions mid-deal. Competitive monitoring feeds directly into competitive enablement. It’s flexible. One team might want a weekly Slack digest of all competitor activity. Another might want per-competitor workflows running on different schedules with different sources. Skills and Workflows support both.
Start with one skill for your top competitor. Get the output right. Customize copies for others. Add Workflows once the prompt is dialed in.
Related: Setting Up Your Knowledge Base | Skills and Workflows | AI Agent Tools