River
Y CombinatorBacked by Y Combinator

Free template

Competitive Battlecard Template for Sales

A one-page battlecard with a source behind every claim, discovery questions to replace rumors, and sheets that rank rivals by lost revenue.

Free download · No account needed

What you get

Pack of 8 files: 5 documents, 3 spreadsheets

Competitive Battlecard Template for Sales
  • Claim Provenance StandardDocument
  • Which Competitors Get a CardDocument
  • BattlecardDocument
  • Trap QuestionsDocument
  • Landmine GuidanceDocument
  • Competitor RegisterSpreadsheet
  • Win Rate by CompetitorSpreadsheet
  • Claim Verification LogSpreadsheet
RiverCompetitive Battlecard Template for Sales

Claim Verification Log

Sixty-one inherited claims, audited against their sources

Larkhill Compliance  ·  three inherited cards  ·  every source opened and read, not just loaded

Claim typeClaimsWhat it can supportShelf life
Their own published words22A direct quote, attributed to them90 days pricing, 180 days docs
A public record9A dated fact the prospect can checkAt the printed expiry, or never
Aggregate review pattern7A claim about reviews, never about the product180 days, and re-count
Our own measurement5A number, with its method and date180 days or our next release
A single review quote1One customer's opinion, dated, never generalized12 months on relevance
Hearsay from a call11Nothing. This is not a claimn/a
Source no longer supports it6Nothing. Retired the same dayExpired 6 to 13 months ago
61 claims6144 verified, 17 undefendable (27.9%)

Nine of the seventeen sat on one card

The Northgate card. The competitor carrying 52.4 percent of Larkhill's lost contract value on 26.5 percent of the mentions. The card that gets opened most gets edited most casually.

The sentence that cost deals for eleven months

“They have a 25-seat minimum at $45 per user.” True when it was written. Their pricing page now shows three tiers from $29 with no minimum. A twelve-seat prospect told they were priced out went and looked, and found an entry tier below ours.

Five verified claims were left off every card. True, sourced, and answering nothing a prospect asks. Verified is not the same as worth saying.

Page 1 of 1266 words

How to build a competitive battlecard

A competitive battlecard is a one-page brief a sales rep reads when a buyer names a rival. A good one explains how that competitor wins and how to respond, admits what they do well and suggests a few questions to ask. Every claim on it shows where it came from, the exact words that support it and when someone last checked.

A card is only as good as its weakest line. Once typed up, a rumor from an old call looks just like a quote from the rival's pricing page, and reps repeat both.

A man in a suit talks on his phone and writes in a notebook beside an open laptop
Note what buyers tell you, then check it before it goes on a card. Photo: Vitaly Gariev on Unsplash

Choose which competitors get a card

Write cards for the rivals that cost you the most, not the ones everyone talks about. Export at least a year of lost deals from your CRM, with the competitor and contract value on each. Rank each rival by how many losses name them and by lost value, the total contract value of those losses.

The two lists often disagree. A rival winning many small deals feels like the main threat, while one winning fewer large deals takes more of the money.

The same lost deals, ranked two ways

CompetitorLossesAverage lost dealLost valueYour win rate
Rival A96$14,200$1,363,20039%
Rival B71$38,600$2,740,60020%
Rival C44$11,400$501,60056%

Illustrative figures. Rival B gets the first card: it costs most, and you win least against it.

Before trusting either list, count lost deals with no competitor named. A high share has two likely causes. Either the field isn't required in your CRM, which is a quick fix, or the buyer chose to do nothing, which no card will solve.

Keep the list short, because every card's sources need re-checking each quarter. Maintain cards for the two or three rivals behind most of your lost value, and give the rest one shared page of objection answers.

Sort every claim by where it came from

Gather the evidence first. A competitor analysis of their pricing, docs, public records and reviews collects most of it. For each claim, store the source, the exact supporting words, the date you read them and an expiry date.

Six kinds of claim and what each supports

Kind of claimWhat it supportsHow it sounds
Their own wordsA direct quote"Their pricing page lists no seat minimum."
Public recordA dated fact the buyer can check"The registry shows their certificate expiring in March."
Your own measurementA number with its method and date"Across our last 34 rollouts, median go-live was 6.2 weeks."
Review patternA claim about reviews, not the product"17 of 41 reviews this year mention onboarding."
Single reviewOne customer's dated opinion"One reviewer wrote in May that setup took a quarter."
Rumor from a callNothing yetA question for the buyer (see below)

Review patterns are the easiest to overstate: a count of reviews describes reviews, not the product. "Their customers hate onboarding" is a claim the count can't support.

In the US, naming a rival isn't the risk. The FTC's policy on comparative advertising, in place since 1979, encourages naming competitors1, and its small-business guide says advertisers need proof before an ad runs2. Hold every line on a card to that bar. The risk is a false claim: the federal Lanham Act lets a rival sue over commercial advertising or promotion that misrepresents its goods or services3.

When to involve a lawyer

Have a lawyer review any comparative claim before it moves from a card into ads, comparison pages or decks for many buyers. This guide isn't legal advice.

Write the battlecard for a 30-second read

Reps often open a card moments before a call, so the top must work on its own. Keep it to one page, built only from verified claims, in this order.

  1. Header

    The rival, the claim count, the next check date and why this rival earned a card.

  2. A 10-second read

    Two or three sentences on how they win and what to do about it.

  3. Claims grouped by source

    Their own words first, then public records, your measurements and review patterns.

  4. What they do well

    At least one genuine strength. A card that only attacks gets read once, then distrusted.

  5. What came off, and why

    Removed claims with reasons, so nobody repeats last quarter's card from memory.

  6. The one thing not to say

    The claim most likely to be both wrong and checked, usually about pricing.

ExampleThe top and bottom of a cardAn illustrative example written for this guide

The 10-second read: Rival B wins large accounts on procurement comfort, not product. Don't attack the product. Ask about their rollout timeline and seat pricing, both published on their own site.

The one thing not to say: "They have a 25-seat minimum." That was true last year. Their pricing page now shows no minimum, and a 12-seat buyer who checks will find a tier priced below ours.

  • Strategy first. It tells the rep what to do, not what the rival sells.
  • A warning with a reason. It names the stale claim and what a buyer would find.

Turn rumors into questions

Most unsourced lines on an old card started as a real remark on a real call. Don't delete or repeat them. Rewrite each as a question the buyer can answer from their own paperwork or talks with the rival, so you assert nothing and learn something about this deal.

Rumors rewritten as questions

TopicThe rumorThe question that replaces it
Timeline"Their rollout really takes nine months.""What did they quote for go-live, and is that date in the contract?"
Support"Their support is terrible.""After rollout, who owns your account: a named person or a queue?"
Contract"They lock you into three years.""What's the renewal term, and what happens to your data if you leave?"

Drop what no question can fix, like comments about named people or guesses about a rival's finances and plans. Two reps hearing the same rumor is still one rumor.

Ask two per call, chosen from what the buyer has said, and space them out. In 2017, Gong, which sells software that records and analyzes sales calls, published an analysis of 519,000 recorded discovery calls. Top sellers spread their questions evenly through the call4 instead of front-loading them. Add yours to your discovery call questions and log the answers: one that recurs is worth researching.

Never confirm a rumor

If a buyer repeats something you can't verify, say you don't know, ask where they heard it and write it down. Agreeing makes it your claim.

Give every claim a shelf life

Claims go stale at different speeds, so give each kind its own expiry, not one review date for the whole card.

  • Pricing: 90 days
  • Product documentation: 180 days
  • Certifications and registry entries: their printed expiry date
  • Review patterns: 180 days, then recount
  • Your own measurements: 180 days, or your next release if sooner
  • Court filings and other dated events: no expiry

Once a quarter, reopen every source and compare it with the words you stored. If nothing changed, push the expiry forward. If it's reworded but still true, update the quote. If it no longer holds, pull it from the card that day but keep its row in your log, so nobody re-adds it.

Save a dated copy as you go: the Internet Archive's Save Page Now gives you a permanent link5 that works after the original page changes or disappears. Then re-run the competitor ranking from a fresh export. Your own releases can expire claims too, so make re-checking the cards part of each product launch plan.

Before you share the card

  • One competitor, on one page, chosen by lost value
  • Every claim has a source, its exact words and an expiry date
  • Review counts describe reviews, not the product
  • At least one genuine strength is on it
  • Rumors appear as questions, not claims

For a faster start, open the template above with Edit with AI: River ranks your rivals from your lost deals and audits your existing cards. To draft one card from public research, use the sales battlecard generator.

How to use this template

Use the files as they are, or open them in River. With Edit with AI, River asks for your closed-lost deals and any battlecards or comparison pages you already have. It ranks your competitors by the contract value you lose to them. Then it checks every claim on your current cards against its source and tells you how many you couldn't defend today.

  1. Download it or open it in River

    Download the five Word documents and three CSV sheets free, or open the template in River to work on it with AI.

  2. Rank rivals by lost value

    List a year of lost deals with the competitor and contract value, then write cards for the rivals that cost the most.

  3. Audit claims, then write the card

    Check each claim against its source, or have River do it, then turn rumors into questions and build the card from verified claims.

  4. Re-check it every quarter

    Reopen each source, update quotes that moved, retire claims that no longer hold, and re-run the ranking.

What's in the template

Which competitor earns a card, the card itself, and the questions that replaced the claims nobody could sourceThe Competitor Register, a finished battlecard, and the Trap Questions converted from hearsay.
Competitive Battlecard Template for Sales
RiverRegister

Competitor Register

412 closed-lost opportunities over 18 months. A competitor is named on 268 of them.

CompetitorLossesAvg lost valueLost valueShareWin rateDecision
Northgate71$38,600$2,740,60052.4%20.2%Maintained card, 19 claims
Verityx96$14,200$1,363,20026.0%38.9%Maintained card, 14 claims
Palisade44$11,400$501,6009.6%56.4%Shared objection page
Long tail, 12 names28$12,300$344,4006.6%54.1%Shared objection page
Cormorant29$9,800$284,2005.4%58.6%Shared objection page
Named268$5,234,000100.0%2 maintained cards

The two rankings disagree, and the disagreement decides the work. By mentions it is Verityx, Northgate, Palisade. By lost contract value it is Northgate first, twice Verityx's figure, on 26 percent fewer mentions. Verityx wins in a high-volume low-value segment and generates channel noise every week. Northgate wins in the enterprise segment and generates the losses.

Northgate gets the first card because both measures agree. Highest lost value and the lowest win rate of the five at 20.2 percent, which together mean the deals are large, they are being lost, and nothing currently in the field is working.

144 of the 412 closed-lost records, 35.0 percent, name no competitor at all. Report that before any ranking: it is the confidence interval on the whole table, and it has two very different causes.

RiverThe card

Battlecard: Northgate

19 claims, all verified. Next re-fetch in 60 days. Grouped by source type, because the source decides the phrasing.

Read this first, in ten seconds

They win large accounts on procurement comfort, not on product. Do not attack the product. Ask about the timeline and the seat economics, both published on their own site, both worse the smaller the account.

GroupClaimHow to say it
Their wordsEntry tier $29 per user, no seat minimumQuote the pricing page directly
Their wordsStandard engagement is 16 weeks“Their own services page says 16 weeks”
Their wordsFindings export as CSV onlyQuote the documentation
Their wordsTrust page lists no report periodA claim about what the page says, not about the certification
Public recordRegistry entry expires in 9 monthsA dated fact; let the prospect check it
Reviews17 of 41 reviews in 12 months raise onboarding“Of the reviews published this year, this many raise onboarding”
Reviews9 of 41 praise the reporting templatesSay it. A card that only attacks gets read once
Our measurementOur median to first audit cycle is 6.2 weeks, n=34With the method and the date, next to their 16

Nine claims came off the previous version. Seven were call notes: a nine-month implementation, bad support, a three-year lock-in, a doubt about the certification, an inability to handle multi-site, a customer supposedly leaving, and a plan to move upmarket. Five became questions, two were dropped. The other two were their old pricing and the seat minimum.

The one thing not to say: that they have a 25-seat minimum. It was true, it is not, and a small prospect told they are priced out will go and look.

RiverQuestions

Trap Questions

Eleven unverifiable claims. Seven became these questions. Four were dropped, because no phrasing was worth asking.

The claim nobody could sourceThe question that replaced it
Their implementation really takes nine monthsWhat did they quote you for time to first completed audit cycle, and is that number in the contract?
Their support is badWho owns your account after implementation closes, and is that a named person or a queue?
They lock you into a three-year termWhat is the renewal term, and what happens to your data if you do not renew?
They are not really certifiedHave you asked for their current report period, or just the badge?
They cannot handle multi-siteWhich of your sites would go live first, and how many are on the older process?
They are about to raise pricesWhat term are you being offered, and is the rate held at renewal?
Their free tier is a trapWhat happens to your reports if you go past the monthly cap?

A question does three things an assertion cannot. It carries no exposure, because nothing is being stated about the competitor. The answer is specific to this deal rather than to a rumor about the market. And a prospect who says the thing out loud has convinced themselves, which no card has ever managed.

The four that were dropped. A large customer leaving, which arrived twice fourteen months apart from two reps and is therefore one rumor rather than corroboration. Their engineering headcount. A plan to move upmarket. Half their team leaving. Nothing about a named individual is ever load-bearing in a deal.

Ask two, not seven. Seven questions on a card becomes an interrogation in the hands of somebody reading it as a script.

  • BattlecardA finished one-page card for one rival, with claims grouped by source and the one thing not to say.
  • Trap QuestionsDiscovery questions that replace unsourced claims, each shown beside the rumor it replaced.
  • Landmine GuidanceFive things that never go on a card, three that lose deals anyway, and what to say instead.
  • Claim Provenance StandardThe six kinds of claim, how each may be phrased, and how long each one stays valid.
  • Which Competitors Get a CardHow to rank rivals by lost contract value, read the blank-field rate and decide how many cards to keep.
  • Claim Verification Log sheetEvery claim with its type, source, exact supporting words, dates, verdict and where it ended up.
  • Competitor Register sheetEach rival's losses, lost contract value, share of the total and card decision, plus the blank-field rate.
  • Win Rate by Competitor sheetWins, losses and average deal sizes for each rival, with a note on how much a card could change.

Common questions

What is a competitive battlecard?

A competitive battlecard is a one-page sheet that tells a sales rep how to handle one rival in a live deal. Product marketing or sales enablement usually writes it. A good one covers how that rival wins, how to respond, what it does well and which questions to ask, with a source and check date on every claim.

What should never go on a battlecard?

Anything you can't source. That rules out a rival's roadmap unless they published it, guesses about their finances or stability, capability claims with no documentation, and comments about named people. Never turn a review count into a verdict on the product. Rumors from calls become questions for the buyer instead.

How many battlecards should you maintain?

Probably fewer than you have. Two or three current cards for the rivals that cost you the most beat a shelf of stale ones, and everyone else can share one page of objection answers. Each card means re-checking its sources every quarter, and a rep can't tell a stale card from a current one by looking.

How often should you update a battlecard?

Re-check every source at least once a quarter, and right away when a rival changes its pricing or launches something major. Better still, give each claim its own expiry. The template gives pricing claims 90 days and lets a certification run to its printed expiry date. Put the next check date on the card's header.

Is it legal to name competitors on a battlecard?

In the US, yes, as long as what you say is true and you can prove it. The FTC's policy encourages naming competitors in comparative advertising and judges those ads by the same standard as any other. The risk is a false statement about a rival's product, so have a lawyer review any claim before it appears in ads or public comparisons.

Is this competitive battlecard template free?

Yes. The five Word documents and three CSV sheets download free, with no signup. They come filled in for a made-up software company, so you can study a finished card, its claim log and the competitor ranking before replacing them with your own. Edit with AI opens the same files in River.

What does Edit with AI do?

It opens the template in a free River workspace. River asks for your lost deals, with the competitor and contract value on each, and any competitive material you already have. It ranks rivals by what they cost, then reports how many existing claims you couldn't defend today. No export yet? It starts with the audit.

Sources

  1. eCFR. “16 CFR 14.15: In regard to comparative advertising.” ecfr.gov (opens in a new tab)

  2. FTC. “Advertising FAQ's: A Guide for Small Business.” ftc.gov (opens in a new tab)

  3. Cornell Law School. “15 U.S. Code § 1125 — False designations of origin, false descriptions, and dilution forbidden.” law.cornell.edu (opens in a new tab)

  4. Gong. “Mastering discovery calls to close deals effectively.” gong.io (opens in a new tab)

  5. Internet Archive. “Save Pages in the Wayback Machine.” help.archive.org (opens in a new tab)

Written by

Chandler Supple

Co-founder & CEO, River

Chandler has spent the last six years working in AI, building tools for millions of users and Fortune 100 companies. In 2025, he co-founded River to build a new kind of operating system for AI agents.

Reviewed by

Brian Joseff, Co-founder, River