Read signals over time · 11 min read · Updated 24 Sep 2026

Prompts to Analyse a Competitor Funding Round

A funding announcement is a press release, so almost everything in it is the company's own framing. The amount is self-reported and says nothing about revenue. What predicts behaviour is the sequence: how long since the last round, at what stage, and who led it. These prompts separate the claims from the record.

A competitor's funding announcement is a document they wrote

Nearly every competitor source is written for somebody else. A pricing page is for buyers, a job advert is for candidates, a filing is for regulators. A funding announcement is written to be re-reported, which makes it the only one produced specifically to travel.

That shapes what is in it. The amount, the investors and the customer names are checkable. The growth figures are chosen. The sentence about what the money is for is a statement of intent by people who have just been given the money to form one.

It also makes this the easiest competitor document for a model to hand back to you. Press release English is a register a language model writes fluently, so asking what a round means returns the company’s own frame with different verbs in it.

The amount is the least useful number in it

The headline figure is the one thing everybody repeats and the one that predicts least. It says nothing about revenue, nothing about profitability, and nothing about how long the money lasts. A company can raise a large round because it is winning or because it needs to.

What carries information is the sequence around it. How long since the last round. Whether the stage moved, repeated or went unnamed. Whether the lead changed, and whether a previous lead sat this one out. None of that is in the announcement, which is why the prompts below ask for the history as a separate input.

Prompts to read a competitor's funding round

Three passes, and the middle one has to be sent on its own. That is not a style preference.

The first read, once you have the announcement and whatever coverage exists.

[COMPETITOR] announced a [AMOUNT] [ROUND] round. Here is the announcement, the investor list and any coverage I have: [PASTE]. Tell me: - what they said the money is for, in their words - what the investor list suggests about the direction, and how confident that reading is - what this changes about how they will compete in the next four quarters, and what it does not change - which claims here are the company's own framing rather than independently reported Be specific about the last one. A funding announcement is a marketing document, and the amount raised says nothing about revenue.

As its own message afterwards, never as a bullet added to the prompt above.

Here is everything I have about [COMPETITOR]'s [ROUND] round: [PASTE]. Sort every factual statement in this material into three lists, and print all three: 1. stated by the company and not verifiable from anything here 2. stated by the company and confirmed by a named third party, with the party named 3. not stated anywhere, though a reader would assume it For list 1, quote the sentence. For list 3, name what is missing: revenue, valuation, the base figure behind any growth multiple, how much of the total is equity rather than debt, and burn. Do not summarise the round. I want the three lists and nothing else.

Once you have the earlier rounds, which is what makes this one mean anything.

Here is [COMPETITOR]'s funding history, oldest first, and here is the latest round: [PASTE]. Ignore the amounts for a moment and answer from the sequence alone: - how long since the previous round, and how does that compare with the gaps before it - did the stage progress, repeat, or go unnamed - did the lead investor change, and did any previous lead not participate - has the stated use of proceeds changed direction from the last time Then say which of those four you consider the strongest signal about the next year, and why. Where the history is too short to read, say so instead of reading it.

Asking a model to produce an analysis and then to say which parts it inferred returns a disclaimer rather than a separation. The same instruction sent afterwards, as its own message, returns a clean split. That behaviour was measured while building the displacement planning prompts, and it applies to every check worth running.

What a funding analysis needs beyond the announcement

Three inputs. The second is the one that turns an event into a signal, and the third decides whether the classification prompt has anything to put in its middle list.

The announcement in the company's own words

Where it comes from
Their own post rather than a news write-up of it, because the write-up is usually the post with the adjectives removed. Where a round becomes public record is covered under finding competitor funding.
What good looks like
The full text, including the customer names, the growth claims and the quote from the lead investor. The adjectives are evidence of positioning.
Without it
The analysis runs on a second-hand figure, and the wording that shows what the company wants believed has already been edited out.

The earlier rounds, with dates

Where it comes from
The funding databases and, where the jurisdiction requires it, the filings. What each database records and what it models is covered under reading a Crunchbase profile.
What good looks like
Date, stage, amount and lead for each previous round, oldest first, so the gaps between them are visible.
Without it
One round has nothing to be compared with, so the only thing left to talk about is how big the number is.

Anything reported by somebody who is not them

Where it comes from
Trade coverage that adds a fact rather than rephrasing the release, an investor's own announcement, or a regulatory filing of the transaction.
What good looks like
At least one figure or detail that appears outside the company's own material, so list two of the classification prompt is not empty.
Without it
Every claim lands in the same bucket, and the output cannot tell a confirmed number from an aspiration.
Then run it
The history goes in before the announcement, so the new round reads as the latest entry in a sequence rather than as an event on its own.
Before the output leaves the building
Strike any sentence that would still be true if the amount were half or double. What survives is the part of the analysis that rested on something other than the headline.

Use their post rather than the trade coverage of it. Most coverage is the release with the adjectives taken out, and the adjectives are the evidence. How a company describes its own market in a moment of maximum attention is a positioning statement they chose carefully.

A competitor's own framing, returned as your analysis

We took one real Series B announcement, in the company’s own words, and asked the two questions people actually ask of a competitor raise.

Attio's published Series B announcement, read on 24 September 2026 against the post itself.

Claude Opus 5handed a CRM competitor's own Series B announcement in full and asked what the round means competitively

2026-09-24

A $52 million Series B led by GV, taking the company to $116 million raised, signals strong investor conviction in the AI-native CRM thesis. With 5,000 paying customers and 4x ARR growth this year, the company has demonstrated product-market fit in the segment it targets. The appointment of a former GitLab revenue chief to the board points to a deliberate shift from founder-led selling towards a repeatable enterprise motion.

Checked against Attio's Series B announcement

Every figure is quoted accurately, and every figure comes from the company. The growth claim is the one to look at. The announcement says the company is on track to 4x ARR growth this year, which is a projection about an unfinished year, made by the company, with no base figure anywhere. A multiple without a base is a statement about a ratio. The analysis repeats it as a demonstrated result.

False precision

Claude Opus 5given the same announcement and asked what the round changes about how the company will compete over the next year

2026-09-24

Expect four shifts. Sales headcount will grow, particularly in enterprise, following the board appointment. Marketing spend will increase to defend the AI-native CRM position against incumbents adding agent features. Product velocity will rise as engineering scales. And pricing will become more flexible, since a funded company can discount longer in a contested deal to win a reference account.

Checked against Attio's Series B announcement

Three of the four are reasonable and none of the four is in the document. The announcement says the money goes to engineering, global expansion and product, and says nothing about pricing, marketing or discounting. What is useful here is not the prediction but the list. Each of those four is something you can watch rather than forecast, and each leaves a public trace within two quarters.

Over-inference

A multiple with no base is a ratio

One phrase in the first answer is worth stopping on. The company said it was on track to quadruple annual recurring revenue this year. The analysis reported that as growth the company had demonstrated.

Three things changed on the way. A projection became a result. An unfinished year became a completed one. And a multiple with no starting figure became evidence of scale. Four times a number nobody has published is still a number nobody has published.

None of that makes the claim untrue, and the company is entitled to make it. The failure is in the reading. A classification pass fixes it in one line, because the sentence goes straight into the list of things stated by the company and confirmed by nobody.

The test that survives a funding announcement

Cross out every sentence in your analysis that would still be true if the amount were half or double what it was. Most of a first draft goes. What is left is the part that rested on the sequence, the investors or the stated direction, and that is the part worth sending to anybody.

What a competitor's raise changes in four quarters

A round changes what a competitor can afford. It does not change what they are good at, who their product suits, or how long their sales cycle runs. Most of the anxiety a funding announcement causes is about the second list rather than the first.

The useful reframing is to stop predicting and start watching. Everything money buys leaves a public trace within two quarters, and each trace has a source you can already read.

Where a competitor's new money becomes visible, and how long it takes
What the money buysWhere it showsHow long
PeopleTheir job board, in the functions and territories that appear first.Four to eight weeks. The fastest signal there is.
DemandAd libraries, in the number of live variants and how long each runs.Weeks, and visible to anybody who looks.
Room to discountYour own lost deals, before it ever shows on a published page.One sales cycle, and only if somebody records the reason.
ProductTheir release notes, and later their packaging.Two to four quarters. Slower than everyone expects.

Three of those four have a prompt of their own. Reading a competitor’s open roles answers the first, their ad creative answers the second, and their changelog answers the fourth. The discounting one stays yours, because it never becomes public.

Watch competitor funding rounds automatically

A funding announcement is a single moment, and the analysis you just ran describes it. What the money does takes a year, and none of that year is in the document.

The mistake this encourages is treating the announcement as the event. It is the start of one. A competitor who raised in March and has hired nobody by September has told you something the announcement could not, and the only way to know is to have been watching the four traces above.

One thing stays invisible however closely you watch. Valuation, terms, how much of a total is equity rather than debt, and how long the money lasts are all private, and estimating them from the headline produces a figure that gets quoted as fact later.

Flares tracks competitor hiring, pricing, packaging and product changes and dates each one. The quarters after a raise then read as a pattern with dates on it, rather than as a thing somebody mentioned in a meeting.

What a competitor does after they raise

Flares watches the hiring, pricing and product changes where a funding round actually becomes visible.

Discover Flares

14-day free trial · 30-second setup

Reading a funding announcement FAQ

What does a competitor's funding round actually tell you?

Somebody with money agreed on terms, on a date. A round confirms a company can raise, names investors who now expect a return, and usually states a direction. What it does not report is revenue, profitability or how long the money lasts, and none of those can be inferred from the amount.

Does a bigger round mean a stronger competitor?

Size and strength are different things, and occasionally they point opposite ways. A large round at a modest valuation, or one raised soon after the last, can mean a company needed the money. The amount describes what was raised rather than how well the business is doing.

How should you read the investor list?

As a statement of intent about direction, and read the absences too. A new lead from outside the previous investor group suggests a different thesis. An earlier lead that did not participate this time is worth noticing, though it has several innocent explanations and none of them is published.

What changes after a competitor raises money?

Four things become observable within a couple of quarters: hiring, advertising, pricing latitude and how long they will stay in a contested deal. Product rarely moves that fast. Everything else people predict after a raise is a guess dressed as analysis.

Why does AI repeat a funding press release back to you?

Because a funding announcement is written to be re-reported, and a model trained on press coverage is fluent in exactly that register. Ask what a round means and the answer arrives in the company's own frame, with their growth claims presented as findings. The instruction that helps is the one that sorts claims by who made them.

Can you work out a competitor's revenue from their funding?

No, and the common shortcuts are worse than no estimate. Multiples of a round size, or of a reported valuation, produce a number with no error bar that then gets quoted as fact. A growth multiple like four times ARR, quoted without a base, is a statement about a ratio and not about revenue.

How long does a funding round stay relevant?

About four quarters for behaviour and longer for the obligations it created. The spending shows up quickly and then becomes normal. The expectations attached to the round last until the next one, which is why the gap between rounds tells you more than either round does.

Should a funding announcement change your battlecard?

Only if something in it changes what a buyer would experience. A new tier, a named market or a stated pricing intent belongs on a card; a raise on its own does not, and reps who hear about it from a customer need an answer anyway. The line that usually belongs there is the one about what it does not change, written for the competitor battlecard.

What if a competitor has not announced a round in years?

Silence has more than one reading and the innocent ones are common. A company can be profitable, be raising quietly, or have agreed terms it does not want published. Registries in some jurisdictions record a share issue whether or not anybody announced it, which is the only way to tell silence from nothing happening.

Do funding rounds predict a competitor moving upmarket?

A raise is the money for it rather than the evidence of it. The evidence arrives in job adverts within a quarter and in packaging within two, and reading it across sources is the subject of working out who a competitor sells to. A round makes the move affordable and says nothing about whether it was chosen.

Track the spending, not the announcement

Flares dates every competitor change, so the quarters after a raise read as a pattern rather than a rumour.

Discover Flares

14-day free trial · 30-second setup