Compound
Open the app
About Compound

Keeping up with tech isn't the same as getting sharper.

Compound is a short daily edition of what happened in venture and tech, built so you do something with it. You read the day, then commit to a call on it. Over time your record starts to show you how you actually think. Five minutes, most days.

Open Compound → Free · no account · works on your phone
One edition a day ~5 min to read or listen Your calls stay on your device
The idea

Reading is the easy half.

The investors you admire formed views before the answers were in, then lived with how those views aged. That loop is what trains judgment, and no amount of reading substitutes for it. Compound gives you the loop.

A day in Compound

The whole loop takes about five minutes.

Five steps, and you can see all of them a screen from here.

1

Read the day

One edition, cut down from the serious sources of the venture world to the few stories that matter. There is no feed to scroll. When you reach the end, you're done.

2

See it through The Lens

Each story is reasoned through the way a seasoned investor would frame it, including what they'd be skeptical of and where the second-order effects hide. The Lens stops at the live tension and never gives a verdict, so the call stays yours, not an echo. Then you commit: pick a side, set how sure you are, and say why in your own words.

Compound
Deals · Delivery M&A · 1 newsletter
Uber's bid for Delivery Hero is a bet on owning food delivery, and a looming regulatory fight.
The Lens
The strategic rationale hinges on scale unlocking profit in a notoriously hard sector. The core tension is whether the combined entity can find real synergies fast enough to justify the valuation and outrun the regulators.
Make the call
Will the acquisition close by Q2 2027?
Resolves: Q2 2027
YesNo
How sure are you? Likely
LeanLikelyStrong
☆ Conviction call
Your reasoning optional
Regulators will grumble, but the divestitures are pre-baked and the scale logic is real.
3

Make the call

Once you commit, the call is locked and dated. Your reasoning is kept word for word, so the version of you that made it can't be quietly rewritten later. They wait here until the news settles them.

Compound
Calls
14 calls · tap any to see your reasoning
Awaiting the verdict
Does DeepSeek price its IPO above $60B?Pending
Will Anthropic's G7 access be restored by Q4?Pending
Resolved
Does Apple raise hardware prices this cycle?Right
Will open-source pass 40% of new deploys?Somewhat
Does OpenAI ship a consumer device in 2026?MissedReflect →
4

Grade the thinking, not the result

When the news settles a call, a reflective card walks you through what happened and asks whether your reasoning was sound given what you knew. A sound call that got unlucky is still a sound call.

Compound
1 · Your call

You said Yes at Likely confidence.

"Regulators will grumble, but the divestitures are pre-baked and the scale logic is real."

2 · How it resolved

The EU opened a Phase 2 review; the deal slipped past the window.

3 · Why it happened

The divestitures answered overlap. The regulator cared about data consolidation.

4 · The verdict that counts

Sound given what you knew then, not whether it worked out.

SoundMixedFlawed
Net · Somewhat — it missed, but you rated the reasoning mixed.
5 · One lesson
Read what the regulator protects, not what the company offers.
5

Watch it compound

Your calls add up to a judgment profile. It shows whether the calls you marked likely land that often in practice, whether being right was earned or lucky, and it keeps every lesson you write down along the way.

Compound
Judgment Profile
38
calls made
9
conviction
24
resolved
Tells · what your calls say about you
When you say Strong you land 71% of the time (47-87% given 17 calls). Strong should mean about 88%, so even the generous end of that range is short. You're overconfident where you're surest.
Calibration · does your confidence mean anything?
80-99%71%47-87% · vs 88% · 17
65-79%69%42-87% · vs 72% · 13
50-64%2 more resolved to read this
Process · was your reasoning sound?
9Earnedright · sound
4Luckyright · not sound
6Unluckywrong · sound
3Genuine misswrong · not sound

Screens show sample data. Yours starts empty and fills in as you make calls.

To be clear

What Compound isn't.

Not another newsletter

There is no archive to fall behind on and no inbox guilt. One edition a day, and the news is mostly raw material for the practice.

Not predictions or picks

Compound never tells you the answer. That would defeat the whole thing. It sharpens the question and hands you the call.

Not a course

There is nothing to memorize. The frameworks stay invisible in the daily flow, and they surface only if you go looking for them in The Method inside the app.

The method

Built on ideas that are older than any newsletter.

Good forecasters' 70%s happen about 70% of the time. That skill is called calibration, and it's the centerpiece of your judgment profile.

Process over result is where learning lives, so Compound grades the decision you controlled, not the outcome you didn't. It also expects you to change your mind when facts change, not when the clock runs down. Do that daily, and it compounds.

Why this exists

I made Compound because I wanted to get better at venture judgment, and reading more wasn't doing it. This is the practice I wanted to exist. It works better with company.

— Sami

Start today

Today's edition is waiting.

Open Compound →
On iPhone: open in Safari, then Share → Add to Home Screen to keep it as an app.