For investors and for founders raising

Know the company
before the first call.

From the ngrok memo

A diligence question

Critical

What is ngrok's NRR and monthly churn rate across the 30,000 paid customer base? Please provide cohort retention curves by signup quarter for the last 8 quarters.

Why it gets asked

The deck discloses zero retention data. At $1M ARR on 30,000 paid customers, the implied ARPU of about $2.78 a month is only defensible if paid customers expand over time. NRR above 120% would be the strongest possible signal that ngrok edge is already working.

Check it against the whole memo

Upload a deck. You get a verdict, four graded axes, the flags with their slide numbers, and the questions the deck has to answer. Figures are cited back to the deck page or the web page they came from. About ten minutes.

Run your deck Both memos in full

Two published memos

Same four axes. Opposite answers.

Passing on Theranos is easy in hindsight. The harder test is whether the same bands find anything on a company that worked. Both memos are public and whole, so you can check the call rather than take it.

ngrok

Invest

Developer tools / Infrastructure · Growth

Ngrok is a bootstrapped developer standard with genuine product-market fit and a credible enterprise upsell thesis, but the monetization gap is the bet.

Unicorn opportunityB+ 7/10
Capital efficiencyA 9/10
Team velocityA 8/10
Moat durabilityB+ 7/10

5 flags, none criticalWhole memo

Theranos

Strong Pass

Healthcare / Diagnostics · Series B

The case-for-investing has not been made on any dimension that survives a single diligence call.

Unicorn opportunityC 3/10
Capital efficiencyF 2/10
Team velocityC 3/10
Moat durabilityF 2/10

10 flags, 4 criticalWhole memo

How a grade is set

One scale for every deck, so a grade means the same thing on both memos above.

F 1-2: Unfundable on this axis, no matter how the story is told.
C 3: A significant gap, at the hard end of fixable.
C+ 4: A real gap, but one that capital and time can close.
B 5-6: The venture median. A typical company in this category scores here.
B+ 7: Above the median, with one named, defensible advantage.
A 8-9: Strong, backed by evidence that holds up under diligence.
A+ 10: Best in cohort, and rare by design.

What a deck leaves out

Most decks lose the room on something small.

Six gaps from the two memos above. The shortest is first, because almost every deck has it.

ngrok
No founding date, headquarters location, or team size is disclosed in the deck.
TheranosSlide 3
Deck presents $120-$300M revenue projection for next 1.5 years (slide 3) but provides no current revenue or ARR baseline, making the growth rate and achievability impossible to assess.
ngrokSlides 3-4
Slide 3 and 4 show traction metrics (5.4M accounts, 4,000 daily sign-ups, 30,000 paid customers, $1M ARR) but no growth rates, cohort retention, or unit economics (LTV, CAC payback) are provided.
TheranosSlide 3
Slide 3 states 6 deals with 5 companies in validation phase worth $6-12M, but also states existing deals will generate $120M-$1.5Bn in revenue; these two statements are unreconciled.
ngrokSlide 7
Slide 7 states 'raising $50m to chase a $73b TAM' but no valuation, lead investor, or use-of-funds breakdown is disclosed.
TheranosSlide 21
Slide 21 lists Series A and Series B investors with detailed fund profiles, but does not disclose the valuation or share price for either round, making it impossible to assess dilution or valuation progression.

At the end of every memo

The questions the deck has to answer.

Anyone can ask for cohort data. Only something that has opened the deck can say which figure a real answer would blow up.

01Theranos
Reconcile the revenue figures on slide 3: the deck states 6 deals with 5 companies in validation phase worth $6-12M and simultaneously states existing deals will generate $120M-$1.5Bn in revenue. Are these figures from different time periods, different deal stages, or different definitions of 'existing deals'?
A Series B investor cannot model returns without knowing whether current revenue is $0, $1M, or $10M. The unreconciled contradiction is either a presentation error or a material misrepresentation. Either way, it must be resolved before any term sheet discussion.
02ngrok
What is the ARPU breakdown by customer segment, free, paid individual, paid team, and enterprise? How many of the 30,000 paid customers are on enterprise contracts, and what is the average ACV for Databricks, Zendesk, Klaviyo, and Copado?
The enterprise ACV data is the single number that most directly validates or invalidates the ngrok edge upsell thesis. If even 100 enterprise customers are paying $10K-$50K ACV, the blended ARPU picture changes materially.
03Theranos
What is the derivation of the $50M average revenue per Phase IV trial? Provide the calculation: (assumed patients per trial) x (monitoring periods per trial) x ($7,500 per patient per period) = $50M, and cite at least one reference trial where this revenue has been or is contractually committed to be generated.
The $50M per-trial figure is the load-bearing unit-economic assumption in the deck's market sizing and revenue projections. If it is a theoretical maximum rather than a realized or contracted figure, the $39B TAM and $120-$300M projection both collapse.

Start to finish

How the memo gets made.

Upload the deck
A PDF or a slide file. Nothing to set up first.
It gets taken apart
Claims are pulled out and checked against public sources. Contradictions get named.
The memo arrives
Every line quoted above is from a finished one.

Ethan

Founder, sent.bio

“Your product helped bring a lot of clarity on what the best path forward is. Its brutal honesty helped do that.”

Your data is safe

Every deck you upload is sensitive. Here is exactly what happens to it.

Encrypted

Your deck is encrypted.

It stays encrypted while it sits on our servers and while it moves between your browser and Verdict. The exact standards are on the security page.

Private to you

Only your account can open it.

Your files belong to your account, and Verdict looks them up by that account every time it reads them. Nothing in the database can be reached straight from a browser.

Never trained on

Your deck never trains an AI.

That covers the memo it produces and any feedback you write, not just the deck itself. Anthropic runs the AI behind Verdict and its business terms forbid training on anything a customer sends.

Yours to delete

Delete it and it is really gone.

Delete a memo from the report page and the deck file leaves storage in the same step. Anything you keep is deleted for you after 90 days, or 24 hours from when you uploaded it if you marked the deck sensitive, whether or not the memo finished.

Read the full security posture, with retention tiers, sub-processors, and what’s still in progress.

Frequently asked

The questions everyone asks first.

01How is this different from ChatGPT, Claude, or Gemini directly?

A climate-tech deck I ran claimed 450 terawatts of waste heat on slide 3. Terawatts measure power, not energy, so the number was physically impossible, and nobody who built the deck caught it. Verdict caught it on the first run. A chatbot is trying to hand you an answer you'll like, where Verdict is looking for the thing that kills the deal.

02What about hallucinations?

Every number links back to where it came from, the exact deck page or the web source, so you can check any claim in one click. Verdict also writes two separate verdicts, one on the company itself and one on whether it fits a particular fund's thesis, so it has to commit to each of them instead of giving you a single vague answer.

03Are you SOC 2 certified?

Not yet, and I'll be straight with you about why: I'm a solo founder and I'm early. If your team needs a signed data-handling agreement before you upload anything, email info@useverdict.io and I'll sign one.

04Can I share a memo with someone else?

Every memo has a read-only share link, and the person you send it to does not need an account, the same as sending a Google Doc. Paid plans also export a PDF that carries a confidential mark and lists every source at the back, so it can be passed on without editing it first.

05I'm a founder. What do I get out of running my own deck?

The memo a partner is going to write about you, before you walk into the room. Verdict sizes your TAM against real sources and names the competitors you left off the slide. It writes down the objection that kills the deal, and that one is usually the hardest to hear. The Coaching toggle turns each hit into a fix you can make before the meeting rather than during it.

Run a deck and see what comes back.

Both memos are published in full, exactly as the product wrote them.