How Tom works

A retainer starts with one goal for 90 days, a baseline for every number and a name against every step.

From Retainer kickoff

Write the 90-day goal in one sentence, with a table of how you will know on the last day. Give each platform one job, one KPI and one named owner. Log a baseline for every KPI from the month before the start, and leave targets blank until there is data.

From Retainer kickoff

The day a client signs, I write the plan that night. One goal for 90 days, one job per platform, and every checkpoint on a Thursday so it lands in the midweek meeting.

From Retainer kickoff

Every piece of work starts from what the content already shows. Scrape the last 30 posts, count everything, find what is three to four times the median, and only then say what to make.

From The creator audit

Write down the handle, the platforms, the purpose and who will read the audit. Scrape the last 30 posts with a capped run, and record the date and scope. Leave old pinned posts and anything under 24 hours old out of every median, and call out any boosted post.

From The creator audit

Before I say anything about someone's content, I scrape their last 30 and look at what's actually in front of me. I work out the median, then the median against their followers, then anything at three to four times that median, because that's an outlier worth studying.

From The creator audit

One booked studio, one person asking and a bank of questions written in advance. The first fake pod gave 34 usable clips in an hour, because the writing was done before anyone sat down.

From Recording day

Put every question skipped or never reached last time at the top of the new bank. Write new questions from the calls, packets and concept pages that have arrived since the last recording. Name the specifics in each question: the brand, the person, the number.

From Recording day

Recording day is the cheapest way I know to post every day. I book a studio, someone reads out questions I've written, and I answer them.

From Recording day

Watch the whole talk, map who said what, then cut one idea per clip at 40 to 75 seconds. Every beat carries its timecodes on every angle, so the editor knows exactly what goes in and in what order.

From The clip bank

Sync the camera angles on their audio and pick the master: the one that starts earliest, runs to the end and carries the questions clearly. Write the source page with every speaker and timestamp before choosing a single clip. Keep only moments that stand alone, carry one of Hormozi's seven and are clip shaped: a contrarian take, a number, tactical advice, a candid admission or a real question.

From The clip bank

When I get a talk back, I watch the whole thing and map who said what before I pick anything. Every clip is one idea, 40 to 75 seconds.

From The clip bank

Every session is researched before it starts, run to a timed agenda on a page the client can see, and recorded. The recording becomes the breakdown: a roadmap, the transcript, and the next step priced as a three-month sprint.

From The consulting session

Research their profiles before the call: median, outliers at 3 times the median or more, and every claim they make. Write the prep doc with three lines to land and Tom's lean on the next step. Build the run of show as timed blocks, each with its question and a done when line.

From The consulting session

Before the call I've read everything you've posted. I find your median and your outlier, and that's the first thing we look at.

From The consulting session

The first message is soft: who Tom is, one tip they can use today, then go. The one-hour call only comes up once they reply.

From Soft DM outreach

Collect the DMs where someone asked for 1:1 help, and get the account owner's approval first if the inbox is someone else's. Pull each person's followers, bio, offer and last 5 posts before writing anything. Tier them: strong if they named a problem and wanted help, medium if they gave no goal, not a lead with a reason Tom can overrule.

From Soft DM outreach

When someone's asked for help, the first message is soft. Your content came up on my feed, I think I remember your profile engaging with Sam Stoffel, I used to be his head of content.

From Soft DM outreach

Every client so far has come in through a paid hour and been offered the next step in writing within days.

From Pricing and the offer ladder

Send the session breakdown within 48 hours with the next step priced at the foot of it. Offer three options at most, recommend one, and credit the smallest against the biggest. Use the sprint when there is a team to align: six two week sprints and a session every fortnight.

From Pricing and the offer ladder

Every client I have right now started with one paid hour. I send a breakdown within a couple of days with the next step priced at the bottom.

From Pricing and the offer ladder

Tom has built a content function from zero five times, and at a certain level that is the job people hire him for. The order is the same each time: audit, move people, hire, install the system, hand the gates over.

From Building a content team from zero

Spend week one on the audit: the people, the costs, the tools, the output and what has been tried. Move people into the job their skill fits before you open a new role. Write the ad with the pay range and a bar for month 1, month 3 and month 6.

From Building a content team from zero

At a certain level people hire me to build the group of people who make the content. Week one is an audit: who is here, what it costs, what has been tried.

From Building a content team from zero

The middle of the funnel keeps an audience, and carousels and fake pods are how Tom fills it. Nothing becomes a carousel until the video behind it has already won.

From Carousels and the middle of the funnel

Make a carousel only from a video that beat the trailing 30 day median. Put the hook on slide one in bold, aimed at one person. Use slide two for the credential line that says why you get to say it.

From Carousels and the middle of the funnel

Top of the funnel brings people in. The middle keeps them, and that is carousels and fake pods.

From Carousels and the middle of the funnel

Getting there takes 6 to 8 titles and 50 to 60 ideas, built from her own winners and the outliers in her niche and the niches next to it.

From Monthly ideation

Scrape all of the creator's long form and the last five to six months of short form, with every metric and a transcript per post. Summarise each post from its transcript and put it into a format bucket with that bucket's median views. Search her niche and the adjacent niches for outliers and save the ones worth adapting into the monthly ideas tab.

From Monthly ideation

To land on thirty good ones I need fifty to sixty, so I generate double and cut. It starts with data: I scraped her long form and her last five or six months of short form, bucketed every post and looked at what actually works.

From Monthly ideation

A format pitch sells a show that repeats. Tom writes the promise, the rules of an episode and the device that comes back every week, then backs it with a season plan, a call sheet and a budget the client can sign.

From Pitching a show format

Write down the client's brief, the constraints and what they asked for on the first call, word for word where you can. Write the one-line promise in about ten words and three to five working titles under it. Set the mechanic in numbers: topics per episode, people per episode, runtime and clips per episode.

From Pitching a show format

Every show I pitch starts with one line you could say in a lift. Then the rules of an episode in numbers, and one device that comes back every week until it becomes the brand.

From Pitching a show format

Tom co-led the first in-game shirt sponsorship in FIFA history and signed Mason Mount as Combat Gaming's first ambassador. Both came from the same habit: give the talent or the brand the experience first, then make the ask.

From Brand partnerships and talent

Build or run something talent wants to be part of: an event series, a team or a format the creator owns. Invite the talent to take part before any commercial ask. Make the ambassador ask after they have enjoyed the thing.

From Brand partnerships and talent

Mason Mount played three Combat Gaming events before I asked him anything, and he became the first ambassador. Same with brands.

From Brand partnerships and talent

The biggest videos Tom produced won on story and on planning that survived chaos. Ideas come out of a planning session days later, the shoot is built around one moment, and the first hour after upload decides the rest.

From Making a viral YouTube video

Book a regular planning session with the creator and arrive having watched what is performing in the niche and the genres next to it. Give each topic a fixed slot in the session and expect the answers to arrive in the days after it. Pick the idea that has a single moment the whole video hangs on, and write down what you will do if that moment never comes.

From Making a viral YouTube video

Our biggest video, the KKK one, took 90 minutes of planning because Niko rang me mid trip and said he had convinced the leader of the KKK he was from the BBC. Most of the others took weeks.

From Making a viral YouTube video

A good title and thumbnail can be the difference between 100,000 views and a million. Tom's method is to go in with a title, pre-test the variants, isolate what you change, and own a name in every title.

From Packaging, titles and thumbnails

Write a working title before you film so the shoot knows what outcome it is serving. Write the hook first too, then swap it if the shoot gives you a better spontaneous moment. After the edit, refine the title against what the video actually delivers.

From Packaging, titles and thumbnails

A good title and a good thumbnail can be the difference between 100,000 views and a million. So I go in with a title first, because it tells me what the video is for, and I refine it after the edit.

From Packaging, titles and thumbnails

Prep the questions, film the hour, cut one idea per clip, then repackage the winners as carousels and text posts. One Creator Hotline hour gave Tom two months of daily content.

From Long form to short form

Write the question bank and your answer spines before the studio day, drawn from the problems your audience already has. Film the session as a run of standalone answers, one question and one idea each. Pick moments that stand alone with a sentence of setup at most, and that carry conflict, a number, tactical advice or a candid admission.

From Long form to short form

I had 60 minutes in a podcast studio for Creator Hotline and came out with two months of content, posting every day. That only happened because it was prepped.

From Long form to short form

Every proposal starts with where the client is and what they said on the call, in their own words. The fee comes at the end, as at most three options with one highlighted.

From Proposals that win

Audit their channels from a fresh scrape and link every video you cite with the date it was pulled. Open with an at a glance table of where they are, and keep the price out of it. Choose the three to six past clients closest to their problem and check every figure against Numbers.

From Proposals that win

When I come off a call I build the proposal that night, while what they said is still fresh. It opens on their numbers and their words, then their channels as they are today, then what I have done for people with the same problem.

From Proposals that win

A founder brand is sold on what the audience gets from the founder's experience. Pick one position with an enemy, repeat a handful of messages daily, and build the team that runs it.

From The founder's personal brand

Scrape the founder's last 30 posts and group them by format with the median views for each. List every offer the content has to sell, and note the gap between them. Write one line on who the founder is, who they are for and what they are against, and get it agreed in writing.

From The founder's personal brand

A founder's audience cares about what it gets from the founder's experience. So I start with the last 30 posts and find what is already working, which is usually opinion said to camera.

From The founder's personal brand

Everyone owns a column, every crossing between boards is a named handoff, and the founders see the near-final film.

From The Venture Room Asana operating system

Name the four questions the team cannot answer today: what is out this week, where the month goes, what a shoot yielded, what is waiting on me. Open one Planning ticket per shoot, set its Format and make the shoot date the due date before Plan sign-off. Harvest within two days: log selects from the transcript, create one child cut per select and set the parent's Cuts field to the yield.

From The Venture Room Asana operating system

A small content team runs on one board. Every idea, every edit and every bit of feedback lives on a card, and every card has an owner and a column.

From The Venture Room Asana operating system

Tom builds AI into the places a content team already works.

From AI in the content workflow

Write the brief before any code, then build the tool as the implementation of that brief. Turn the ideas channel into tickets automatically, once each, with the author credited. Give editors one working surface over the editing board so review and verdicts happen in one place.

From AI in the content workflow

With Sam Stoffel and Ben Heath we built really great systems that used AI to find the shortcuts. Then the tools join it up.

From AI in the content workflow

Short form earns attention and the newsletter keeps it. It is the only audience Tom owns outright, and every offer on the ladder launches to it.

From Owned audience

Decide who the audience is before you build it, and call those people out in the free content every day. Make the newsletter signup the main event on the website, and point every profile at it. Send one email a week: the week's argument written down, plus what performed and what that says.

From Owned audience

Short form is the loud bit, the newsletter is the bit I own. I'll push people from the website into a weekly email, and that list is where every paid thing launches.

From Owned audience

Tom has not launched his own paid community yet.

From The paid community

Name the one audience the community is for and talk only to them in your content. Put the course inside the community and set modules to unlock by months subscribed, spread over two or three months. Give it a simple name that says what it does, with you as the authority behind it.

From The paid community

Courses feel oversaturated right now. Too many bad ones.

From The paid community

Five strategists who between them claim billions of views describe the same front end: ideas come from videos that beat their own channel's average, read side by side with the ones that did not, then cut hard. Galloway cuts a hundred ideas to three or four, Hoyos keeps one in a hundred, and Kallaway kills any idea without a contrarian take.

From Outlier mining and the idea funnel

Search the niche on YouTube with Shorts filtered out, sort by most viewed, and note what established channels typically get. List the channels of 10,000 to 250,000 in the niche, starting from one you know is good and its nearest neighbours. Pull their videos from the last two months (7 or 14 days in a news niche), drop anything under 2% engagement, and sort by outlier score.

From Outlier mining and the idea funnel

Every strategist I rate starts ideas the same way. Find videos that did three or four times their own channel's average, and study channels between 10,000 and 250,000, because on a huge channel fame carries the hit.

From Outlier mining and the idea funnel

Galloway, Losardo, Kallaway, Colin and Samir, Hormozi and Lim agree that the package and the opening decide a long form video. Losardo puts 40% of the effort into packaging and 30% into the intro. The thumbnail carries three things at most and three to five words, the title fires a different idea from the thumbnail, and the first ten seconds confirm both on mute.

From Packaging and the first ten seconds

Before the shoot, write the title and mock the thumbnail; if the package does not clear the bar for the channel, change the idea. Write down what the viewer already knows and what they want to know, and rewrite until the gap between them can only be closed by clicking. Keep the thumbnail to three elements at most and three to five words, simple and slightly exaggerated.

From Packaging and the first ten seconds

Kallaway says YouTube is a packaging platform, and he proved it: the same video went from 1,700 views to 1.2 million with a new thumbnail. Seb Losardo puts 40% of the effort into the title and thumbnail and 30% into the intro.

From Packaging and the first ten seconds

Jenny Hoyos built 8 million subscribers in three years on Shorts and says her framework has worked in every video since she wrote it: visual shock, immediately start, rising action, anticipation, lasting payoff. Around it sit her hook criteria, a pacing curve, retention floors, Shorts of one length, and a testing loop that changes one thing at a time.

From The Shorts story formula

Gather at least 100 ideas from outliers, your own day and other media, and keep the one whose hook is strongest. Make the first frame a visual you understand on mute, with the action already started. Open on a question, then say why it matters, so the viewer knows what they get by the end.

From The Shorts story formula

Jenny Hoyos says the hook is the idea, and if the hook is weak she does not make the video. Her framework is VIRAL.

From The Shorts story formula

Between them: book guests who pass three tests and outperform on other channels, let the package pick the guest, make the guest comfortable before anything else, build the episode as sections that pay off, and test it before release.

From A podcast built to repeat

Score every guest idea on three questions: is the host interested, is the guest the expert who speaks well, does the audience want it. Book only on three yeses. Search for outlier guests, people whose interviews outperform on other channels, and check timing and exclusivity. Write the title and thumbnail before booking, and let the package decide the guest.

From A podcast built to repeat

Diary of a CEO books a guest only on three yeses: Steven is interested, the guest is the expert, the audience wants it. Two is a no.

From A podcast built to repeat

Ralston, Kallaway, Hormozi and Lim build personal brands for people who sell something, and they agree on the order: decide the outcome, find the contrarian belief, narrow to the buyer, turn their painful problems into ideas, make every piece for a stranger, and grade on emails, followers and who reaches out. When Hormozi narrowed, his views per video fell and his revenue per thousand views rose 68%.

From A personal brand that converts

Answer the four brand journey questions, backwards from the outcome you want. Write two things to be known for and two to be known against. Fill two columns: what your industry says that you disagree with, and what you believe instead. Test each take in content and keep the one that gets DMs.

From A personal brand that converts

Caleb Ralston says you can have millions of followers and make no money. So start with the outcome.

From A personal brand that converts

An event puts the talent, the audience and the proof in one room for a day, and someone else has paid for the room. Plan it content first, run four or five stations that each give one repeatable format, and the footage feeds short form, carousels, email and ads for weeks.

From The event as a studio content day

A week out, write the station list for the event: queue, stage, whiteboard wall, hot seat, testimonial room, with the format each one feeds. Book a videographer to follow the talent all day, D-Rock style, with licence to stop them and say: say that again to camera. Kit the stage with one or two static cameras and one roaming for faces and hands, and put a second mic on the talent.

From The event as a studio content day

I treat every event as a studio day. The talent is there, the audience is there, the proof is there, and someone else has paid for the room.

From The event as a studio content day

Month one is the audit, the budgets, the hires and the tools, then test shoots. By month three it is execution.

From Month one challenges assumptions

Challenging assumptions: the why, the landscape, the team, the hires, the budgets and the tools. Then the first test shoots.

From Month one challenges assumptions

The team, clear budgets, a cadence of booked shoots and a clear strategy: the North Star, the why, the enemy and the audience.

From Month one challenges assumptions

Execution, week by week and day by day.

From Month one challenges assumptions

The skill someone already has is worth more than the job title they arrived with. Move people before you hire.

From Move the people you have before you hire

Move the people you have into the jobs their skills fit.

From Move the people you have before you hire

One. Three inherited roles changed first.

From Move the people you have before you hire

The skill someone already has is more useful than the job title they arrived with.

From Move the people you have before you hire

Rehire the whole team, find three formats, build a system that ships three a day. The formats outlived Tom leaving.

From Sam Stoffel: rebuild the team, then find the formats

The team. He rehired it: a script writer and producer, a content producer and creative director, and freelance specialists.

From Sam Stoffel: rebuild the team, then find the formats

Selfie videos, podcasts and reviewing personal brands, output three times a day.

From Sam Stoffel: rebuild the team, then find the formats

0 to 500,000 followers in 10 months, and formats that are still running after Tom left.

From Sam Stoffel: rebuild the team, then find the formats

Why are you investing, what does success look like, what is the budget and runway, and who are you outside the business.

From The intake question is why, and what it is costing them

Why. Why are they investing in a personal brand, and what does success look like to them?

From The intake question is why, and what it is costing them

Because the register of the content follows from who they really are.

From The intake question is why, and what it is costing them

Budget, runway and the KPIs to go after.

From The intake question is why, and what it is costing them

The founder with no cheaper path to their goal stays. The one who finds an easier route takes it. Ask before you start.

From They quit when there is an easier route to the same revenue

The founder with no cheaper path to the goal, like Sam Stoffel, who was building an audience to launch a product to.

From They quit when there is an easier route to the same revenue

The founder who finds an easier route to the same revenue.

From They quit when there is an easier route to the same revenue

Before the work starts. It is a qualification question.

From They quit when there is an easier route to the same revenue

Founders do not need to be full time. They need to accept that this is work with a purpose, and it needs their time and patience.

From Turn down anyone unclear about their commitment

Being unclear about their commitment to the project.

From Turn down anyone unclear about their commitment

No. Tom joins to make the best use of their time.

From Turn down anyone unclear about their commitment

Accepting that it is work with a purpose, and that it needs their time, patience and input.

From Turn down anyone unclear about their commitment

Immerse, audit the team and the costs, then sit with the founder on why. A daily podcast clip and a daily selfie video run from the start.

From Week one is an audit, and two formats are never up for testing

An audit of the people, the costs, the tools, the output and what has been tried, then a sit down with the founder on why.

From Week one is an audit, and two formats are never up for testing

The why, the enemy, the moat and then the formats, which are found by testing every week.

From Week one is an audit, and two formats are never up for testing

A daily podcast clip and a daily selfie video.

From Week one is an audit, and two formats are never up for testing