Keep competitive battlecards current
Build an agent that watches your competitors' public pages and your won/lost deals, then keeps each battlecard accurate so reps walk into competitive deals with current facts, not last quarter's.
Tools you'll use
A competitive battlecard is a one-page reference that tells a sales rep how to win a deal against a specific competitor: that rival's pricing, positioning, strengths, weaknesses, and the traps to set in a live conversation. The whole point is to be right on the day a rep needs it. The problem is that battlecards go stale the moment a competitor changes a price, ships a feature, or shifts a message, and a quarterly PDF refresh almost guarantees the card is wrong when it matters.
This is worth fixing because competitive intelligence is not a nice-to-have: 61% of businesses say it has a direct impact on revenue, up from 52% the year before (Crayon, 2021 State of Competitive Intelligence Report, 1,000+ respondents). A battlecard is the moment that intelligence reaches the rep, so its accuracy on the day of the deal is where the payoff actually lands.
The work splits cleanly. One part is monitoring and drafting: watch a fixed list of competitor pages plus your own loss notes, flag what changed, and draft the update. The other part is judgment: deciding what is true, what is fair to claim, and what gets published. An agentic tool handles the first part on a weekly cadence; a human owns the second. You are not buying a competitive-intelligence platform here. You are building a small, owned routine that turns scattered public signals into a card your reps trust.
Moriva's take
This clears all three gates. Gate 1: keeping battlecards current is a recurring weekly job that quietly decides competitive deals, so it is real work, not a demo. Gate 2: with Claude Code your team owns the monitoring script and the prompts outright, so a single operator can run it, fix a broken page scrape, or add a new competitor without calling a vendor. Gate 3: it is measurable because you already track win rate against named competitors, so you can see whether fresher cards move the number. Keep a human approving every published change, since a wrong claim about a rival is a self-inflicted wound in front of a buyer.
How do you keep competitive battlecards current?
- 1
List your competitors and their watch pages
Pick the 3-8 competitors that actually show up in your deals. For each, write down the public URLs that change and matter: pricing page, product or features page, the homepage hero, key blog or release-notes pages, and their G2 or review page. This list is the spine of the whole system. Keep it in a plain file your team controls.
- 2
Stand up a weekly monitoring agent with Claude Code
Point Claude Code at that list and describe the job in plain English: fetch each page weekly, compare it to last week's saved copy, and write a short diff of what changed in pricing, packaging, positioning, or messaging. It builds a script and a stored snapshot folder that your team owns and can run on a schedule. No competitor-intel subscription required; you own the code and can extend it to new competitors yourself.
- 3
Feed in your own win/loss signal
The best battlecard content comes from your own deals. Export recent closed-lost and closed-won notes that mention a competitor from your CRM, and have the agent cluster the recurring themes: which objections come up, which traps work, where you actually lose. This keeps the card grounded in what reps see, not just what marketing wishes were true.
- 4
Draft the card updates, with sources attached
Have the agent turn the weekly diffs and deal themes into proposed edits for each battlecard: a changed price, a new feature to counter, a refreshed 'why we win' line. Require every claim to cite where it came from (the exact URL and date, or the deal it came from). A claim with no source does not get drafted. This is what makes the output checkable instead of confident-sounding guesswork.
- 5
Route every change through a human owner
A product marketer or sales-enablement owner reviews the proposed edits each week: approve, fix, or kill. This is non-negotiable because anything you say about a competitor can be quoted back to you by a buyer. Approval takes minutes when the diff is small and sourced; it is the difference between a card reps trust and one they quietly ignore.
- 6
Publish where reps already work
Push the approved card to wherever your reps already live: the CRM, a shared doc, or your enablement hub. The single biggest failure mode for battlecards is that nobody opens them, so do not create a new place to look. Update the existing one and date-stamp it so reps can see it is fresh.
- 7
Use Claude Cowork for the non-coding owner
If the person who owns battlecards does not code, Claude Cowork lets them do the research, synthesis, and drafting side directly: summarize a competitor's new pricing page, draft the 'how to sell against them' section, and tighten the language. Claude Code runs the scheduled monitoring underneath; Cowork is where a non-technical operator does the weekly judgment work.
- 8
Measure and tune
Track win rate against each named competitor before and after, plus how often reps actually open the card. Review the trend quarterly. If a competitor's page never changes, drop it from the weekly watch; if a new rival keeps appearing in losses, add them. The system is yours to prune.
What could go wrong (and how to handle it)
A wrong claim about a competitor reaches a buyer and damages your credibility or invites a legal complaint.
Require a source URL and date on every claim, and keep a human approving all published changes. Stick to the competitor's own public statements and your verified deal experience; never assert a rival's internal plans or unverified rumors.
Battlecards get built and updated but reps never open them, so the effort produces nothing.
Publish into the tools reps already use, keep each card to one page, date-stamp it, and track open rates. If adoption is low, the problem is usefulness or placement, not freshness.
Scraping competitor pages breaks when their site changes, silently leaving cards stale.
Have the agent flag when a watched page fails to load or returns nothing, and surface that as an alert in the weekly review. Because your team owns the script, an operator can fix the broken page the same day.
The agent over-collects, pulling in pages or data you have no business storing.
Limit the watch list to public pages a buyer could also see. Do not feed in anything obtained through a competitor's gated trial, NDA, or a customer's confidential information.
Win/loss notes from the CRM contain customer names or sensitive deal details that leak into a shared card.
Strip identifying details before drafting, and have the human reviewer confirm no customer is named in the published card. Keep raw notes in the CRM, not in the battlecard.
The card drifts into marketing spin reps don't believe, so they stop trusting it.
Ground every 'why we win' line in a real deal theme, keep the tone factual, and let reps flag claims that don't hold up in the field. A trusted card beats an impressive one.
Prompts to get started
FAQ
Isn't this just a competitive-intelligence platform I could buy?
Platforms like Crayon or Klue are good, and if you want a managed product, buy one. The difference here is ownership: with Claude Code your team builds and owns the monitoring routine and prompts outright, so you can add a competitor, fix a broken scrape, or change the card format yourself in an afternoon, with no per-seat fee or vendor ticket. Start owned and small; graduate to a platform if scale demands it.
How is this different from just asking an AI 'what's true about my competitor'?
A general model will answer confidently from stale or made-up knowledge. This system only works from sources you point it at, today, with the URL and date attached to every claim. That sourcing discipline is the whole value, because it's what makes the output checkable before a rep repeats it to a buyer.
Can the agent just publish updates automatically?
It can, but you shouldn't let it. Anything you say about a competitor can be quoted back to you in a deal, so a human owner should approve every published change. The good news is that approval takes minutes when each proposed edit is small and comes with a source.
Is it legal and ethical to monitor competitors this way?
Watching public pages a buyer could also see, and learning from your own deals, is standard competitive practice. Keep it to public information and your own data. Don't use a competitor's gated trial under false pretenses, breach an NDA, or repurpose a customer's confidential information.
How do we know it's actually working?
You already have the metric: win rate against named competitors. Track it before and after, alongside how often reps open the cards. If fresher, sourced cards move the win rate or reps start citing them in deals, it's working. If not, the problem is usually placement or trust, not the technology.
Sources
- 61% of businesses say competitive intelligence has a direct impact on revenue, up from 52% the prior year (1,000+ respondents). — Crayon, 2021 State of Competitive Intelligence Report
More from Sales
Prep renewals and spot upsell openings
Build a repeatable agent that pulls usage, support, and account data into a per-customer renewal brief flagging churn risk and concrete upsell openings, 90 days ahead of each renewal date.
Qualify and route inbound leads
Build an agent that reads every inbound lead, scores it against your ideal-customer rules, enriches it, and routes it to the right rep in seconds instead of hours. Your team owns the rules and the code.
CRM hygiene and weekly pipeline summaries
Build an agent that reads your CRM every week, flags the records that are wrong or stale, and writes a plain-English pipeline summary for the team meeting. Your reps stop doing data cleanup by hand, and your manager walks into the review already knowing which deals are at risk.
Want help shipping this?
We'll build it with your team on your real work — and leave you owning it, not renting it.