This one is for sales enablement leads, sales managers, and any AE who has opened a competitor battlecard mid-call and found it was last touched eight months ago.
What: Using Breeze Assistant to build a competitor battlecard from data you already have in HubSpot — deal notes, closed-lost reasons, call summaries — rather than researching one from scratch or hoping someone remembers to update the old one. This doesn't replace your playbook. It keeps the content inside it honest and current, grounded in what's actually happening in your deals rather than a research session from eighteen months ago.
Prompt of the week:
Every sales team builds a battlecard the same way. Someone spends an afternoon researching a competitor, drops their findings into a HubSpot playbook, and it's genuinely useful for about a quarter. Then the competitor ships a new feature, changes their pricing, or a rep loses a deal for a reason nobody wrote down. The battlecard stays live, but it stops matching reality. Nobody schedules "update the battlecard" as a task, so it just sits there being quietly wrong.
Most of that update is already sitting in the CRM. Every deal marked closed-lost with a competitor named in the reason, every note a rep leaves about an objection that came up on a call - that's raw material for exactly the intelligence a battlecard needs. Nobody's ever asked anything to go and collect it.
This prompt does that collecting. It pulls from your own deal history, your own lost reasons, your own reps' notes on what actually came up in the room, and turns that into the card. It won't tell you a competitor's roadmap. It will tell you, with evidence, which objections are real, which ones keep recurring, and which of your team's current talking points aren't landing anymore.
Prompt Structure
Paste this into Breeze Assistant and make sure CRM data access is enabled in your AI settings so Breeze can reference deal properties, notes, and closed-lost reasons:
Role: You are a sales enablement specialist who builds competitor
battlecards from real deal evidence rather than desk research. You
trust patterns in closed-lost reasons and rep notes over assumptions,
and you flag a claim as unverified rather than inventing detail
Task: Build (or refresh) a competitor battlecard for the competitor
named below, using only evidence from our own deal history: closed-
lost reasons, deal notes, and call summaries logged in HubSpot. Group
the evidence into recurring objections, name which ones we win
against and which we lose against, and flag anything that needs
human verification before it goes in front of a rep
Context:
- Competitor name: [COMPETITOR]
- HubSpot tier: [Sales Hub Professional/Enterprise, since deal
properties and playbooks depend on it]
- Where competitor mentions are captured today: [a "Competitor"
deal property / free-text in closed-lost reason / rep notes only]
- Time window to pull from: [last 6 / 12 months of deals]
- Existing battlecard, if one exists: [paste current content, or
say "none — building from scratch"]
Run the following:
1. DEAL EVIDENCE PULL
- Find every deal in the time window that named this competitor,
in the property, the closed-lost reason, or a note
- Split into deals we won and deals we lost against them
- For each, extract the specific reason given, not just "lost to
competitor"
2. OBJECTION PATTERNS
- Group the lost-deal reasons into recurring objection themes
(price, a specific feature, integration, support, contract
terms, incumbency)
- Rank themes by frequency and by deal value lost, not frequency
alone. One theme costing five enterprise deals matters more
than one costing ten small ones
- For each theme, note whether it's a real product gap or a
positioning/messaging gap we could address without changing
the product
3. WHAT'S WORKING
- From the deals we won against this competitor, extract the
reasons reps gave for why we won
- Turn these into the strongest current talking points, backed
by "we've won X deals citing this"
4. STALE OR UNSUPPORTED CLAIMS
- If an existing battlecard was provided, check each claim in it
against the deal evidence
- Flag any claim with no supporting deals in the time window as
"UNVERIFIED — no recent deal evidence, confirm before using"
- Flag anything contradicted by recent deals as "OUT OF DATE"
5. GAPS IN THE DATA
- If competitor mentions are only in free text rather than a
dedicated property, say so and recommend setting one up so this
audit gets easier and more reliable each time it's run
Constraints:
- Only state a claim if it's backed by a specific deal or note. If
evidence is thin, say "SIGNAL WEAK: only [N] deals, treat with
caution" rather than presenting it as settled
- Do not invent competitor product detail, pricing, or roadmap. This
prompt surfaces what our own deals show us, not external research
- Rank by deal value influenced, not just count of mentions
- Keep the output short enough for a rep to scan before a call, not
a research report
Output format:
### I. BATTLECARD SUMMARY
{3-sentence overview: how many deals this is built from, the
single biggest reason we lose to this competitor, the single
biggest reason we win}
### II. OBJECTIONS TO HANDLE (ranked by deal value)
| Objection | Frequency | Deal Value Lost | Real Gap or Messaging Gap | Suggested Response |
### III. WHY WE WIN
| Reason | Deals Won Citing This | Suggested Talking Point |
### IV. FLAGGED FOR REVIEW
| Existing Claim | Status | Why |
### V. DATA GAPS
{What to start logging so the next refresh is stronger}
Why this prompt works, and how to adapt it
Most battlecards get built once and trusted forever. A competitor's product, pricing, and your own team's win rate against them all move constantly, so a battlecard nobody re-checks against evidence just calcifies into folklore. This prompt treats the battlecard as something to keep testing against real deal data, rather than a document you write once and file away.
A few things worth knowing about how it's put together:
Deal value counts more than mention count. A weak objection that's come up in ten small deals matters less than a strong one that's cost you three enterprise accounts, so the prompt weighs them accordingly rather than treating every mention as equal.
Real gaps and messaging gaps get sorted separately. Plenty of "we lost on integrations" turns out to be "we never clearly explained our integration story" once you look closely. That distinction matters, because a genuine gap needs a product conversation and a messaging gap just needs a better slide, and mixing the two up wastes effort on the wrong fix.
Weak evidence gets called weak. Three deals isn't a pattern, and the prompt says so rather than dressing up a thin sample as a confident trend. A rep who repeats an overclaimed "fact" and gets pushed back on it mid-call loses more credibility than the battlecard was worth in the first place.
An existing battlecard gets audited, not just supplemented. If you already have one, the prompt checks each of its claims against recent deals and flags what's gone stale. That's often more useful than any new content it adds, because the thing quietly costing you deals is usually an outdated line reps are still repeating with confidence.
Adapting it for your portal
Don't have a dedicated "Competitor" property on deals? The prompt will flag this, but you can get ahead of it: "Before building the battlecard, tell me exactly how to set up a Competitor property and where in the deal process reps should fill it in, so future refreshes don't rely on hunting through free-text notes." Worth doing once, since it makes every future refresh faster and more reliable.
Multiple competitors in play? Run it once per competitor rather than trying to cover them all in one pass, or the ranking logic gets muddled: "Focus only on [Competitor B] this time. I'll run this again separately for the others."
Want it to feed straight into a Playbook? Ask for the output shaped for the delivery mechanism you already use: "Format Section II and III as content ready to paste into a HubSpot Playbook, matching the structure of our existing competitor playbooks." That keeps this as the intelligence-gathering step and your existing playbook as the format reps actually see mid-call.
Sales team small, not much deal history yet? Say so upfront: "We've only got a handful of deals against this competitor so far. Build what you can from that, and tell me what to start logging consistently so this is genuinely useful in two quarters' time." You'll get a thin but honest first pass, plus the instrumentation to make the next one stronger.
Worried about reps logging inconsistent notes? Ask for a nudge alongside the output: "Also tell me the three things reps should log after every competitive deal, win or lose, so the next refresh has better raw material to work with."
Beyond the prompt
What happens to the output matters as much as the prompt itself. A battlecard that lives in a doc nobody opens does no better than the stale one you started with.
Put the refreshed content where reps already look. For most teams that's the existing HubSpot Playbook, not a new document competing for attention. Update the playbook rather than starting a second one somewhere else.
Treat the "flagged for review" section as a task rather than a footnote. An unverified or out-of-date claim sitting in a live battlecard does more damage than an old one everyone already knows to double-check, because reps trust it by default. Get a human — ideally whoever owns the relationship with that segment — to clear each flag before the refresh goes live.
Set a standing cadence, the same way you would for a content audit. Quarterly usually works for a fast-moving competitor, twice a year for a stable one. Run it regularly enough and the battlecard never gets the chance to drift out of date again.
Competitive intel goes stale fast, and a battlecard nobody revisits eventually does more harm than good. Building it from your own deal evidence, and checking it against that evidence on a schedule, is what keeps it worth trusting on a call.
