(1/10) People were curious to how Deep Research does with a more quantitative request

So I uploaded DoorDash’s S-1 and '20-'24 10-Ks, then asked it to analyze and then write an investment thesis

Below is my process and the full report. Judge for yourself how it did... (2/10) As a reminder, I asked ChatGPT to build me a prompt for Deep Research to do Deep Research on Deep Research prompting. It read all the blogs and literature on best practices and gave me a thorough report.

Then I asked for this to be turned into a prompt template for Deep Research. I've added it below. This routinely creates 3-5 page prompts that are generating 60-100 page, very thorough reports. I’ll be using this later:

Please build a prompt using the following guidelines:

Define the Objective:

- Clearly state the main research question or task.
- Specify the desired outcome (e.g., detailed analysis, comparison, recommendations).
Gather Context and Background:

- Include all relevant background information, definitions, and data.
- Specify any boundaries (e.g., scope, timeframes, geographic limits).
Use Specific and Clear Language:

- Provide precise wording and define key terms.
- Avoid vague or ambiguous language.
Provide Step-by-Step Guidance:

- Break the task into sequential steps or sub-tasks.
- Organize instructions using bullet points or numbered lists.
Specify the Desired Output Format:

- Describe how the final answer should be organized (e.g., report format, headings, bullet points, citations).
Include any specific formatting requirements.
Balance Detail with Flexibility:

- Offer sufficient detail to guide the response while allowing room for creative elaboration.
- Avoid over-constraining the prompt to enable exploration of relevant nuances.
Incorporate Iterative Refinement:

- Build in a process to test the prompt and refine it based on initial outputs.
- Allow for follow-up instructions to adjust or expand the response as needed.
Apply Proven Techniques:

- Use methods such as chain-of-thought prompting (e.g., “think step by step”) for complex tasks.
- Encourage the AI to break down problems into intermediate reasoning steps.
Set a Role or Perspective:

- Assign a specific role (e.g., “act as a market analyst” or “assume the perspective of a historian”) to tailor the tone and depth of the analysis.
Avoid Overloading the Prompt:

- Focus on one primary objective or break multiple questions into separate parts.
- Prevent overwhelming the prompt with too many distinct questions.
Request Justification and References:

- Instruct the AI to support its claims with evidence or to reference sources where possible.
- Enhance the credibility and verifiability of the response.
Review and Edit Thoroughly:

- Ensure the final prompt is clear, logically organized, and complete.
- Remove any ambiguous or redundant instructions.

People were curious, so here's how I'm using Deep Research. I'll walk through the prompting and then an example:

1. First, I used O1 Pro to build me a prompt for Deep Research to do Deep Research on Deep Research prompting. It read all the blogs and literature on best practices and gave me a thorough report.

2. Then I asked for this to be turned into a prompt template for Deep Research. I've added it below. This routinely creates 3-5 page prompts that are generating 60-100 page, very thorough reports

3. Now when I use O1 Pro to write prompts, I'll write all my thoughts out and ask it to turn it into a prompt using the best practices below:
______
Please build a prompt using the following guidelines:

Define the Objective:
- Clearly state the main research question or task.
- Specify the desired outcome (e.g., detailed analysis, comparison, recommendations).

Gather Context and Background:
- Include all relevant background information, definitions, and data.
- Specify any boundaries (e.g., scope, timeframes, geographic limits).

Use Specific and Clear Language:
- Provide precise wording and define key terms.
- Avoid vague or ambiguous language.

Provide Step-by-Step Guidance:
- Break the task into sequential steps or sub-tasks.
- Organize instructions using bullet points or numbered lists.

Specify the Desired Output Format:
- Describe how the final answer should be organized (e.g., report format, headings, bullet points, citations).
Include any specific formatting requirements.

Balance Detail with Flexibility:
- Offer sufficient detail to guide the response while allowing room for creative elaboration.
- Avoid over-constraining the prompt to enable exploration of relevant nuances.

Incorporate Iterative Refinement:
- Build in a process to test the prompt and refine it based on initial outputs.
- Allow for follow-up instructions to adjust or expand the response as needed.

Apply Proven Techniques:
- Use methods such as chain-of-thought prompting (e.g., “think step by step”) for complex tasks.
- Encourage the AI to break down problems into intermediate reasoning steps.

Set a Role or Perspective:
- Assign a specific role (e.g., “act as a market analyst” or “assume the perspective of a historian”) to tailor the tone and depth of the analysis.

Avoid Overloading the Prompt:
- Focus on one primary objective or break multiple questions into separate parts.
- Prevent overwhelming the prompt with too many distinct questions.

Request Justification and References:
- Instruct the AI to support its claims with evidence or to reference sources where possible.
- Enhance the credibility and verifiability of the response.

Review and Edit Thoroughly:
- Ensure the final prompt is clear, logically organized, and complete.
- Remove any ambiguous or redundant instructions. So here's how it works with an example. I did this in 5 minutes. I'd always be way more structured in my context, inputting more about my hypothesis, more context etc. I just did this for fun for you all

Prompt:

Use the best practices provided below and the intial context I shared to create a deep research prompt on the following topic:

Context:
I am an investor who wants to better understand how durable DoorDash’s business is. My hypothesis is that they have a three sided network between drivers and riders and restaurants that would be incredibly hard to replicate. Additionally, they built it when interest rates were low so it would be hard to create a competitor today. I need you to make sure you deeply research a few things, at least, though you will find more things that are important -
- doordash’s business model
- how takeout is a part of the restaurant business model, and the relationship restaurants have with delivery networks. Advantages, risks, etc
- the trend of food away from home consumption in America, how it has changed in the last decade and where it might go
- Doordash’s competitors and the history of their competitive space

I need the final report to be as comprehensive and thorough as possible. It should be soundly rooted in business strategy, academic research, and data-driven. But it also needs to use industry blogs and other sources, too. Even reviews are ok.

It's unfair to say Zoom fumbled the bag

Simply put: COVID pulled forward their future, eating their runway. It alerted competition to the opportunity. Now Zoom is fighting for their life

A bit more about why this has been so tough for the companies who should have won:

https://twitter.com/FracSlap/status/1798137169258721508

In normal times, growth businesses are deliberate about expansion. They spend up to a decade building a strong core product w/ sticky user base and go-to-market muscles like sales and customer success

They use their strong core to fund their next oppty

COVID fucked that all up

1. I’ve been lucky to work with some incredible people throughout my career

I’ve randomly posted the great advice they’ve shared, but thought it was worth consolidating them all

Maybe some folks will find these helpful. I certainly have… 2. On Success

“Beyond a certain point, a “successful” career is a selfish pursuit. The sooner you embrace this, the better your relationships will be

It’s ok for your career to be about making you happy. Pretending it’s for other people is the dangerous game”

Some random observations from our implementation of generative AI for scaled customer support

Who knows, maybe people will find it interesting: 1. Ticket deflection opportunity is real:

Pretty immediately AI is able deflect 20% of support queries. It’s pretty shocking when you dig in on a case-by-case basis how many questions are “known” vs “unknown” + bugs

40% deflection seems reasonably attainable

Appreciate everyone sharing their favorite reads from 2023

Paying it forward, here are 10 books I really enjoyed reading this year. A mix of books I read for the first time and some re-reads, too.

Hope everyone finds something they might like, and here’s to 2024… 1. I re-read Christensen’s Innovators Solution. The predecessor gets all the hype, but IMHO this one is better

The stuff in here about integration vs modularity should be required reading for every operator and investor

(1/8) Yesterday I said it was wild how many talented people struggle to sell themselves during interviews

Lots of people asked how to do it

I've interviewed 100's (if not 1,000's) of ppl, and done a bunch interviewing myself, so here are 5 simple elements of selling yourself (2/8) Some notes before we begin:

This won't cover basic advice like Be Confident or Look Your Interviewer in the Eye

It also won't cover the skills to do the job in question

While some will say the suggestions below are obvious, doing them puts you ahead of 90% of candidates

Random shit I’m hearing/seeing from tech talent market

- Companies turning up heat via reviews/higher standards. Rise to occasion or leave
- Attrition is down. Managers must fire low performers, replace in open mkt
- Hard to get hired. Thousands of apps per role - Growth, Eng roles opening up. Ruthless prioritization in Customer Success, Marketing, IT outsource/contract where possible
- Backfills not on autopilot. Must re-justify role
- De-layer your org. Minimize middle management, increase span of control

(1/8) Good mini thread from @GavinSBaker on why CFO is so important in software. Thought I’d add my thoughts:

At the root of it, software (especially SaaS) is about the financialization of customer cohorts

Here’s what that entails… (2/8) The first thing to understand about SaaS economics

The cash flow from each customer has a trough - When you add more customers the trough becomes deeper but the velocity out of the trough faster

So there’s an incentive to grow quickly

The most important skill to develop in your career is “creating legibility”

Legibility: The quality of being clear enough to see

Everyone’s too busy; too mired in their own problems to give a shit about yours

You must make the complex simple. Your idea needs to be legible This is the #1 thing new leaders lack. They have intense conviction, but can’t:

1. Succinctly lay out the facts
2. Get people to care
3. Close the deal
4. Drive results

The missing ingredient is legibility. They can see it. They can’t make others see it

Having career convos in the new year. One consistent theme:

Early-career folks are often mad that their careers don't have more "liquidity"

They want a direct and immediate connection btw output and compensation/title. Essentially, they want to be "priced" in real time... Early on, I got 3 great pieces of advice, which I pass on to them:

1. You can force re-pricing by changing companies
2. You don't have unlimited pricing events
3. But, appreciation isn't linear. Changing during a plateau can ruin appreciation unrealized, just on the horizon...

Whenever ppl ask for career advice re: joining a startup, I always say the same thing:

“Mentally price that option package they gave you at 0. Do you still want to do it?

Well, Foursquare is letting employee options expire There so much mythology around “hitting the lottery” with an IPO. But employees want a risk-free call option - they don’t like it when the option goes to zero.

And I’ve seen it. A lot. Through dilution. Through acquisition. Through “not worth it to exercise” pricing

The real indicator of “Late Stage Capitalism” is that everything is now a subscription

Tractors, the heated seats in your car, - shit that has 0 business being a subscription

When you don’t own the physical goods you purchased outright - that’s late stage capitalism It’s maximum rent seeking - companies are trying to juice profits and margins by attaching subscriptions to hardware/physical goods

Make no mistake…EVERY SINGLE EXECUTIVE is trying to figure out how to do this

Microsoft has been in the news recently for their partnership with OpenAI and the roll-out of chat in Bing

But many don’t know the true story of how Bing got it’s weird and wacky name 🤪

The name actually comes from The Godfather 🤯

Time for some Microsoft folklore 🧵 It turns out James Caan is ultimately responsible for why Microsoft called their search engine Bing!

Here’s how it breaks down. In The Godfather, Caan, playing hothead Sonny Corleone, ad-libs the line “Bada Bing!” while talking to his brother Michael

Layoffs have come to tech

It's easy to say that COVID pulled forward demand, businesses staffed up to meet it, and now they're pulling back

But something deeper happened

COVID pulled forward their future. They rushed forward to meet it, but now they're fighting for survival In normal times, growth businesses are deliberate about expansion. They spend up to a decade building a strong core product w/ sticky user base and go-to-market muscles like sales and customer success

They use their strong core to fund their next oppty

COVID fucked that all up

One benefit of working for companies for decades at a time is you experience the true fickleness of Wall Street

A few quarters of top-line growth. Then margin expansion. Actually back to growth

Incentives btwn founders and investors are less aligned than we pretend they are You don’t have to squint that hard to see where super-voting shares and shit like that comes from

In fact, we have Twitter. You can literally watch the fund managers change priorities in real time. You just have to scroll back on their timeline

What you'll actually find is the higher up you go the more the Senior Leaders ask the operational questions

"yah yah yah sounds nice how are you going to get it done?"

McKinsey consultants only talk this way because they're not accountable to the outcomes If you want to grow your career, learn to sell AND execute:

- Why this is a big fucking deal (very good outcomes/catastrophic if we ignore it)
- Here's the plan
- Here's how we will measure success
- Here are the blockers/resources I need
- Here's how I'll keep you informed

(1/12) Every software company wants to go from single point solution to multi-product suite

Survivor bias blinds companies + investors to the execution risk. They pencil our crazy expectations

Don't be blinded by the TAM. Here's how Act 2 can derail, and even kill, a business (2/12) Going multi-product too early

The goal is to stack S-Curves, so when core product matures, you have a new leg of growth

Essentially, the core business must "fund" the new venture. But companies launch Prod 2 before their core is stable, and end up with 2 bad products