Competitive research for SaaS
Competitor research that keeps what you saw apart from what you guess
ZeroTwo reads public competitor pages and helps you build a dated brief where every claim is labelled observed, inferred or validated. Then it drafts battlecards, headline options and a pricing hypothesis that stay inside that evidence.
Observed
What a page, changelog or ad says, quoted with its URL and capture date.
Inferred
What you think it means. Written as a hypothesis, with a confidence level.
Validated
Confirmed by something you hold: customer calls, lost-deal notes, product tests.
The claim ledger: one row per thing you saw
An illustrative ledger for a fictional vendor. The brackets mark where your real captures go. The last column is the point: it is the strongest wording the evidence supports.
| Claim | Observed, with source | Inferred | What would confirm it | Safe battlecard wording |
|---|---|---|---|---|
| Messaging shift | Homepage headline now leads with “[quoted phrase]”[URL], captured [date] | May be repositioning toward [segment]. Low confidence: one page. | Changelog, careers page, two recent lost-deal notes | “They now emphasise [phrase].” Not “they pivoted because of [reason].” |
| Pricing page | Lists [N] tiers, priced per [seat or usage unit][pricing URL], captured [date] | Packaging may target [segment]; enterprise may be quote-only. | A second capture on a later date; a prospect who has seen a quote | “List pricing is per [unit] as of [date].” No claim about discounts. |
| Absence of a capability | No page among [N] checked mentions [capability][list of pages searched], captured [date] | They may not offer it. Silence is weak evidence. | Their docs search, a demo question, a support reply | “We did not find [capability] on the pages we checked.” Not “nobody offers it.” |
Fictional vendor, no real captures. ZeroTwo publishes this page, so read it as a method, not a market report.
From a finding to a cautious battlecard line
A research finding becomes one line a rep can use and one test you can run. Neither contains an invented number.
Step 1
Finding
Competitor X's pricing page lists per-seat tiers and a “contact us” enterprise plan.
Observed. Fictional example.
Step 2
Battlecard line
“Their published pricing is per seat; ask the prospect what they were quoted for [N] users.”
Turns a page into a question for the sales call, not a claim.
Step 3
One pricing hypothesis
If we offer [plan change] to [segment], then [metric you already track] will move [direction] within [period]. Stop the test if [guardrail metric] drops.
You choose the metric, threshold and period. Nothing here is a result.
How a session runs
One research session, four moves. Most of the value is in step three, where you decide what each statement is allowed to say.
Scope the set
List the competitors you meet in deals and the three questions the brief must answer. Fewer, better-chosen competitors beat a long list.
Research and capture
Ask ZeroTwo to read each competitor's public pages and return quotes with URLs. Open every page you plan to cite and note the date you looked.
Label every claim
Sort each statement into observed, inferred or validated. Anything you cannot source goes to a to-check list, not the battlecard.
Draft the outputs
Homepage headline options, a battlecard per competitor, a one-page differentiation memo and a pricing hypothesis, each citing the rows it came from.
ZeroTwo tools used, and the plans that include them
Web search
On every plan
Finds and reads current public pages. Free uses Serper; paid plans use Exa.
Deep research
Plus and above
For the longer, multi-source pass across several competitors at once.
Projects
Limited on Free
Keep the brief, its sources and your notes together so the next pass starts from them.
Snapshot, not monitor, and where it stops
The brief is only as current as the day you captured it. Here is how to refresh it, and what it cannot see.
Refreshing the brief
Keep the brief in a Project. When you want it current, paste the previous version, give today's date and ask for a changes-only comparison, quoting both versions. For tracking that runs without you, see the competitor intelligence monitor workflow.
A good fit
- Reading public positioning, packaging and changelogs side by side
- Turning labelled evidence into battlecard drafts and discovery questions
- Writing a pricing hypothesis you can test on your own customers
What a public-page brief cannot tell you
- Private or negotiated pricing; list prices are not what deals close at
- Anything behind a login or inside a competitor's product
- Why a competitor changed course. That needs your own deal and customer evidence.
Prompts that make the model show its sources
Each opens in ZeroTwo with the prompt filled in. Add your competitor list or ledger before you run it.
Build the evidence ledger
For each competitor below, read their homepage, pricing page and changelog. Return a table with the URL, a verbatim quote, what it says (observed), what you infer with a confidence level, and what would confirm it. Do not state anything you cannot tie to a page.
Compare pricing pages
Compare the public pricing pages I list: tiers, price unit (seat, usage or flat), free entry point and what is quote-only. Quote each figure with its URL, and mark anything that is missing or unclear instead of estimating it.
Find the claims worth testing
From the evidence ledger below, list positioning angles that look under-used. For each, say what evidence supports it, what would contradict it, and how I could check it with customers before putting it on the homepage.
Draft a cautious battlecard
Write a battlecard for [competitor] using only the observed rows in the ledger below. Include their pitch, a discovery question for each weakness, our advantage where it is validated, and a list of claims I should not make yet.
Write a pricing hypothesis
Using the pricing comparison and our current plans below, propose one pricing test: what to change, which metric I already track should move, a guardrail metric, and the risk if it fails. Leave thresholds for me to set.
Compare against last quarter
Compare the competitor brief I paste (captured [date]) with their pages today. List what changed, quote both versions, and separate the change itself from any reason you suspect for it.
Questions before you rely on it
Does the brief update itself?
Not from this workflow. A brief is a dated snapshot from one research session. To refresh it, paste the previous brief and ask ZeroTwo to compare it with the pages today. For recurring monitoring, see the competitor intelligence monitor workflow.
Can ZeroTwo tell me what competitors charge?
It can read public pricing pages and quote them, but you should open each page before you cite a figure. List prices change, and many vendors price enterprise deals by quote, so a public page is a starting point, not what a deal closes at.
Can it prove that no competitor offers something?
No. A search can show you did not find a capability on the pages you checked. It cannot show the capability does not exist, so write the claim in exactly that narrower form.
Which ZeroTwo plan do I need?
Web search is available on every plan. Deep research is 150 a month on Plus (it uses credits) and included on Pro and above. Plus is $14.99/month and Pro is $29.99/month ($26.99/month billed annually); the Free plan starts with 100 free credits and adds 20 bonus credits every day you log in. The pricing page has the full comparison.
Is it safe to give battlecards straight to sales?
Treat them as drafts. Each claim should cite a page and a capture date, and anything comparative that you publish externally is worth a second look from whoever owns your marketing claims.
Keep it current and put it to use
- Competitor intelligence monitor
For ongoing tracking and alerts rather than a one-off brief.
- Competitive pricing monitor agent
The deeper, normalised view of competitor pricing over time.
- Customer discovery and ICP refinement
Customer language is the main way to validate what the brief only infers.
- Founder-led outbound and demo prep
Where the battlecards and discovery questions get used.
- Founder-led content production
Turn validated differences into comparison pages and posts.
- Deep research
How the multi-source research mode works before you point it at competitors.