Skip to main content
Battlecards go stale. Objection handling docs get written once and forgotten. The usual problem is not a lack of information — it’s that updating content requires someone to own it manually. Hindsight gives you a way to build competitive content from verified deal data and keep it current automatically. This guide covers how to create it, validate it, save it, and maintain it.

​
Step 1: Start with a Default Skill

The fastest way to create competitive content is to use a pre-built skill from Hindsight’s library. To find one:
  1. Go to Skills in Hindsight
  2. Open Hindsight-created
  3. Add a relevant competitive intelligence skill to your library
You’ll find starting skills for common use cases: battlecards, competitor comparisons, objection handling guides, win/loss breakdowns by competitor. Pick the one closest to what you need and run it. This gives you a first draft grounded in your actual deal data before you write a single word.

​
Step 2: Create Your Own Skill

Default skills are starting points. For content you’ll use regularly, build your own. Add a default skill to create a copy you can edit, or start from scratch with Create skill. A good competitive content skill is specific. It names the competitor, defines the deal set to draw from, and tells the agent what to produce. Vague prompts return vague content. Example: Generic
“Create a battlecard for Competitor X.”
Example: Specific
“Search closed-won deals in enterprise accounts where Competitor X was evaluated. Find the objections buyers raised and how reps responded in wins. Then search closed-lost deals and identify where we fell short. Combine into a battlecard with: top win themes, top loss themes, objection-handling talk tracks, and 3 things to avoid.”
The specific version uses Hindsight’s tools in sequence — deal search, within-deal search, synthesis — and produces content that reflects what actually happened, not what someone assumed. Build assets for the scenarios that matter. Hindsight’s model is designed for specificity. An objection handling guide for enterprise deals is different from one for midmarket. A battlecard for Competitor X in a technical buyer evaluation reads differently than one for a business stakeholder. Build for each scenario you actually face. Examples:
  • Competitor X — Enterprise Objection Handling
  • Competitor X — Midmarket Battlecard
  • Competitor Y vs. Competitor Z — Comparison (Technical Buyers)
  • Competitor X — Late-Stage Pricing Objections
The more precisely an asset matches a rep’s situation, the more they’ll use it. Add context to make skills reusable. Tell the Agent which competitor, segment, fields, or knowledge-base files to use, and describe the output you expect. This lets one skill guide a repeatable process without rewriting the procedure each time. Once you’re happy with the skill, save it. Give it a clear name and description so teammates know when to use it.

​
Step 3: Pass the Smell Test

Before you save content to your library, read it like a skeptic. Hindsight draws from verified deal data, but output is only as good as your prompt and your data coverage. Run through these checks: Does it reflect deals you recognize? Hindsight shows which deals and documents it used. If the sample is thin — fewer than 5 to 10 deals against a competitor — flag the content and note the sample size. Thin coverage means thin confidence. Is the win/loss framing balanced? A battlecard built only from closed-won deals will skew positive. Make sure your skill draws from both wins and losses. The loss data is where the real differentiation lives. Would a rep trust this? Read the talk tracks out loud. If they sound like marketing copy, rewrite them. Reps need language that works in a live conversation, not a press release. Is anything missing? Check the output against what you know. Use Web Search or Scrape URL in the AI chat to fill gaps with current information — recent product announcements, pricing changes, G2 reviews. Save anything useful to your intel library while you’re at it.

​
Step 4: Save to Your Library

Once content passes your review, save it. In the chat window, click Save to Library on any output. Tag it with the relevant competitor so the AI assistant can surface it when reps ask questions mid-deal. Title assets clearly. The name is how the AI finds and cites your content, and how reps know which one to use. A title like “Battlecard” is not enough. Be specific: Competitor X — Enterprise Battlecard, Competitor X — Midmarket Objection Handling, Competitor Y — Pricing FAQ. A few other things to keep in mind:
  • Format matters. Markdown is Hindsight’s preferred format. Clean structure improves how the AI retrieves and cites your content.
  • Add a “last updated” note at the top of each document. This tells your team how fresh the content is and flags when it’s due for a refresh.

​
Step 5: Keep Content Current with Workflows

Saving content once is not enough. Use Workflows to have Hindsight suggest updates to your saved assets automatically. How it works:
  1. Build a workflow that uses your competitive content skill on a schedule
  2. Add the Update Asset action at the end of the workflow
  3. Point it at the asset you want to keep current
When the workflow runs, Hindsight compares new deal data and intel against your existing content and suggests edits. It does not overwrite anything. You review each suggestion and confirm or reject it — so you stay in control of what goes into your library. This means your assets reflect your most recent deals. When three reps in a row lose to a competitor because of a new objection, Hindsight will surface it the next time the workflow runs, not six months from now. To set this up:
  1. Go to Workflows
  2. Create a new workflow and use Browse Skills to add your skill to its prompt
  3. Set a schedule (weekly works for most teams)
  4. Add the Update Asset action and select the target document
  5. Add a Slack notification so your team knows when suggestions are ready to review
You’ll get a Slack message when updates are waiting. Click through, review the proposed changes, and accept what’s useful. This works for every asset type. Run it against your enterprise battlecard with a filter on enterprise deals. Run a separate workflow for your midmarket version filtered to midmarket deals. Each asset stays relevant to the segment it was built for.
Related: Setting Up Your Knowledge Base | Monitoring Competitors | Skills and Workflows