The 100-day marketing challenge: rules, tracker and checkpoints
The 100-day marketing challenge is one distribution action, at a fixed volume, on every working day until you have logged a hundred sessions. A session is ten warm asks, or one published post plus its replies, or thirty-five to forty cold sends, or one capped ad test reviewed once a week. It costs about ninety minutes a morning, five mornings a week, which is roughly 150 hours across twenty weeks. You count sessions rather than calendar days, so a missed Tuesday moves the finish line by one day and costs nothing else.
That paragraph is the whole challenge. Everything below is the part that decides whether you finish it: what counts as one action in each of the four lanes, which lane goes on which morning, the four checkpoints and the numbers that read them, and what to do on the Wednesday after you have missed a week. Most challenge posts stop at the rules. The rules are the easy half.
The four lanes are Alex Hormozi's Core 4 from $100M Leads, and the hundred-day count is his Rule of 100. Neither idea is mine. What is mine is the compression: he assumes four protected hours every morning, and a solo founder who also writes the product has about ninety minutes. If the four lanes are new to you, read the Core 4 explained first, because the rest of this page assumes you know which one you are in on any given morning.
Step one, before session one
Do the arithmetic before you do anything else, because a target that needs volumes you will never produce turns a hundred honest mornings into a hundred mornings of falling behind. Divide what you want by the days you will actually work rather than the days on the calendar. The 100 in 100 goal calculator does exactly that division in about four minutes and tells you one of two things: that the target is smaller than you feared, or that it was fiction from the start. Both are cheap to learn on day one and expensive to learn on day fifty.
If the answer is fiction, change the target rather than the effort. A hundred sessions is a fixed budget of about 150 hours. You cannot buy more of it by wanting the number harder.
The rules
Eight of them. Every one exists because I watched a version of this fall over without it.
- One action, one lane, every working day. The morning contains one thing. If it contains a bit of posting and a bit of cold outreach, it contains neither, because both landed below the volume where a result is readable.
- The volume is decided before the session starts. Ten asks. Thirty-eight sends. One post. Write the number down the night before, so the session has an end you can recognise rather than a feeling you can argue with.
- Every session ends with something sent, published or spent. If nothing left your machine and reached a human who is not you, the morning was preparation. Preparation is useful and it is not a session.
- The message does not change inside a batch. Fifty cold sends of one unchanged message, then read the number. Fifteen sends, a rewrite, fifteen more is six weeks of work that teaches you nothing about any version.
- Count sessions, not calendar days. Session 47 is your forty-seventh, whenever it happens. This is the only rule that decides whether a hundred days survives an ordinary quarter.
- Rest is scheduled, not earned. Two days off every week, plus two named rest days a month. A rest day you planned is a decision. A rest day you hid is the first of three.
- Every lane gets a written exit condition before session one. Thirty posts. Two hundred cold sends to a list you still believe in. Five hundred euros and a stop date. Written while you have no ego in the result, so quiet weeks read as expected instead of as a verdict.
- The next action is written the night before. One line: the lane, the action, the number. Sixty seconds while the laptop is still open. Skip it and the morning opens with a question, and questions at 07:00 get answered by your inbox.
What one session is, in each lane
This is the table to copy out first. The middle two columns are the ones that settle arguments at 07:40 on a Thursday, when you have done some work and want to know whether it was a session.
| Lane | The daily minimum | Done means | Does not count | Per unit |
|---|---|---|---|---|
| Warm outreach | Ten asks to people who would recognise your name, or ten follow-ups plus eight new asks. | Ten messages sent. Not eight sent and two sitting in drafts. | Adding names to the list. Looking someone up and deciding to message them tomorrow. | 4 to 6 minutes each |
| Posting in public | One thing published, plus twenty minutes underneath it answering everything. | Live at a URL somebody else can open, and every reply answered. | A draft. An outline. A post you scheduled and never checked the comments on. | 45 to 90 minutes |
| Cold outreach | Thirty-five to forty sends off a list that already exists, all with the same message. | Thirty-five in the sent folder, message unchanged from the first to the last. | Building the list. Checking deliverability. Warming a domain. Rewriting the subject line. | 2 to 4 minutes each |
| Paid ads | No daily action at all. One capped test, and ten minutes on a Friday reading it. | The cap and the stop date were both written before the first euro left. | Opening the dashboard. Watching the spend tick. Changing a bid on a Tuesday. | 10 minutes a week |
Paid is deliberately the odd one out. There is no daily action in paid ads worth ninety minutes of a founder's attention, and checking a dashboard is not distribution. Set the cap, set the stop date, write both down before the first euro leaves, and look once a week. We spent 258 euros on ads and got 21 signups and zero customers, and the useful part of that was the stop date, not the money.
The wording of the ask is a separate problem from the volume, and it is the one that stalls people for three days before session one. Take a first version off the shelf and change it later: warm outreach templates for Monday, cold email templates for Wednesday, and what to post when nobody is reading for Tuesday.
The week
Five working mornings, one lane each, two days off. The point of rotating is not balance for its own sake. It is that on day one you do not yet know which lane is yours, and the four resolve on completely different clocks. Warm answers the same afternoon. Cold needs about fifty sends before its reply rate stops being noise. Posting needs roughly thirty posts across six to twelve weeks. Pick one lane on day one and there is a decent chance you picked the slowest one and will quit it two weeks before it answers.
| Morning | Lane | The action | The number that counts as done |
|---|---|---|---|
| Monday | Warm | Ten asks to people who already know you. | 10 sent |
| Tuesday | Posting | Publish one thing, then answer every comment for twenty minutes. | 1 post, 20 minutes of replies |
| Wednesday | Cold | Send one unchanged message to a list you built earlier. | 35 to 40 sends |
| Thursday | Warm | Follow up Monday's silence, then ask eight new names. | 10 follow-ups, 8 new asks |
| Friday | Whichever answered | Repeat this week's best lane at its own volume. Then ten minutes on the ad test. | That lane's number, unchanged |
| Saturday, Sunday | Rest | Nothing. | 0, and it is part of the plan |
The week changes once, around session 30, and it changes for a boring reason. Most solo founders have between 40 and 300 people worth contacting, so ten asks a Monday plus eight a Thursday empties the warm list inside six or seven weeks. That is not a failure, it is warm outreach doing the only thing it does. When the names run out, Monday becomes a second cold morning or a second posting morning, whichever the session 30 checkpoint pointed at.
If you genuinely cannot guess which lane deserves the extra mornings later, the which channel quiz asks seven questions about what you sell and who you could already name, and gives you a split across the four. It takes less time than the session you are about to run. The longer argument between the two slowest lanes is in cold outreach versus content marketing.
Rest days, and the counter that does not lie
Here is where most hundred-day challenges are quietly designed to fail. They count calendar days. You do nineteen good days, a customer's account breaks on the twentieth, and a counter somewhere resets to zero. Now tomorrow is day one of a hundred rather than day twenty of a hundred, and starting a hundred-day plan is a much heavier thing to do on a Wednesday morning than continuing one.
Count sessions. Session 20 is your twentieth session, whenever it lands. A missed day moves the finish line by a day and costs nothing else. This is not softness, it is what the evidence supports: in the Lally habit-formation study, missing a single opportunity did not materially affect the process. The full argument, with all six papers, is in why founders quit marketing, and the shorter version of why sessions beat dates is in the Rule of 100 in a two-hour day.
One support fire erases nineteen days of evidence, and the run usually ends here rather than at the missed day itself.
The finish line moves by one and every other number holds. A rest day you named is a decision rather than a slip.
Two days off every week, and two named rest days a month on top. Book them in advance. A rest day you planned on a Sunday costs the plan one session. A rest day you took because you could not face the block costs you the next three.
After a missed day, and after a missed week
This is the section every challenge post skips, and it is the one you will actually need, probably in week five.
One missed day
Write it in the tracker as a missed day. Do not do two sessions tomorrow. Doubling up sounds conscientious and it halves the volume of both, which breaks the one rule the whole thing rests on. The session you skipped has not happened yet. Run it tomorrow at full volume and let the finish line move.
Two or three missed days
Run the floor version tomorrow instead of the full block, and run it early. The floor exists so that the day after a bad patch is cheap enough that you actually do it. One warm ask is five minutes. Three cold sends off the list you already have is eight. Answering the replies sitting in your inbox is ten, and it produces more conversations than most full sessions.
A missed week
Resume at the next number. If your last session was 34, the next one is 35, and it happens on the next working morning rather than on a symbolic Monday. Run three floor sessions in a row first so the return is cheap, then pick the week structure back up. Do not attempt to make up five missed sessions in a weekend. Catch-up weeks fail almost every time, and the failure gets read as evidence that the plan was wrong.
Then ask one question about the week you lost, and answer it honestly. If a customer emergency took it, nothing needs fixing and the plan already handled it. If the block was there and you did not open it, the block had a decision sitting at the front of it, and that is what you fix before session 35. One line the night before, naming the lane, the action and the number.
Two weeks or more
Keep the count, restart the setup. Your sessions are still yours, and nothing about the first thirty-four of them stopped being true. But treat the next ten as a fresh session 10 checkpoint: the only question that matters is whether the block happens eight times out of ten, and you should change nothing about which lane you are running until it does.
The four checkpoints
Four moments where you stop and read the numbers. Each one has exactly one question and one threshold, so the decision does not depend on how the last fortnight felt. The volumes in the second column are what the default week produces by that point, and they are there so you can tell whether a lane is even readable yet.
| Checkpoint | The question | What you have by then | Keep going if | Change something if |
|---|---|---|---|---|
| Session 10 End of week 2 | Did the block actually happen? | About 76 cold sends, 36 warm asks, 2 posts. None of it is readable yet. | Eight of the first ten sessions ran at full volume. | Fewer than six ran. Move the hour earlier and change nothing else. The problem is the cue, not the channel. |
| Session 30 End of week 6 | Is any lane producing replies from people who are not already friends? | About 230 to 300 cold sends, 108 warm asks and 60 follow-ups, 6 posts. | Cold is replying at 5% or better in total, or three real conversations are open in any lane. | Cold is under 2% total replies after 200 sends. Change the list, not the message. A collapsed reply rate is nearly always a targeting problem wearing a copy costume. |
| Session 60 End of week 12 | Has anybody paid? | About 600 to 800 cold sends, the warm list probably exhausted, 12 posts. | Three or more people have paid, or eight conversations are live with people who asked the price. | Zero customers from 400 cold sends and 150 warm asks. Move the Monday and Wednesday volume into whichever lane produced the conversations, and keep one posting morning running underneath it. |
| Session 100 End of week 20 | Which single lane earned the next quarter? | About 1,400 cold sends, up to 360 warm asks, 22 to 25 posts, one finished ad test. | The lane with the lowest cost per customer measured in sessions. That lane gets three mornings a week from here. | Drop any lane that cost more than 25 sessions per customer, with one exception: posting is allowed to have produced nothing by session 100, because 22 posts is below the 30 it takes to read. |
Session 10 is the one people want to skip and the one that matters most. Nothing about any channel is readable after two weeks. Seventy-six cold sends is not a test, two posts is not a test, and a founder who changes lanes at session 10 will change lanes again at session 20. The only legitimate change at that point is the hour the block starts.
Session 100 has one exception written into it, and it is worth reading twice. Posting is allowed to have produced nothing. At one post a week you finish the challenge with 22 to 25 posts, which is below the 30 that make the lane readable, so a zero there is a measurement problem rather than a verdict. Either run two posting mornings a week from session 30, or accept that the content verdict lands sometime after day 100 and keep publishing anyway.
The tracker
One row a week, twenty rows, nine columns. It fits on a notebook page turned sideways and it fits in any spreadsheet. I have tried more elaborate versions and they all get abandoned around week four, because a tracker that takes longer than the five-minute slot at the end of a session is a second job.
| Column | What you write in it | Why it is there |
|---|---|---|
| Week | 1 to 20. | Only there so the rows sort. |
| Sessions | How many of the five actually ran. | The honest number. Write 3 when it was 3. |
| Running count | Total sessions so far, out of 100. | The only number the challenge is scored on. |
| Warm asks | New asks plus follow-ups. | Tells you when the list is about to run out. |
| Posts | Published, not drafted. | Counts toward the 30 that make content readable. |
| Cold sends | Messages that left the machine. | The denominator of your reply rate. |
| Replies | Humans who wrote back, of any tone. | The earliest signal that survives small numbers. |
| Customers | People who paid. | The late signal. Usually zero for the first three weeks. |
| One line | What you would tell yourself next Monday. | In week 14 this column is the whole record. |
Filled in, the first three weeks of a real one look like this, bad week included. Copy the header row, then add seventeen more rows and leave them empty.
| Week | Sessions | Running | Warm | Posts | Cold | Replies | Customers | One line |
|---|---|---|---|---|---|---|---|---|
| 1 | 5 / 5 | 5 | 18 | 1 | 38 | 6 | 1 | Monday produced four replies before lunch. Cold produced two and both were rude. |
| 2 | 4 / 5 | 9 | 18 | 1 | 38 | 5 | 0 | Thursday went to a support fire. Logged as missed, not invented later. |
| 3 | 5 / 5 | 14 | 18 | 1 | 76 | 9 | 1 | Friday repeated Wednesday, because cold was the thing answering. |
The running count is the only column the challenge is scored on. Everything else is there so that the checkpoints have something to read. If you want the volume side sized against your own reply and close rates rather than the defaults here, the outreach volume estimator does that arithmetic directly.
What a hundred days actually produces
Now the honest part, because a challenge post that ends on encouragement is worth nothing. Here is where the hundred sessions go in a typical run, and what they buy at the rates this site uses everywhere else.
| Where the sessions go | Sessions | Volume | What that buys |
|---|---|---|---|
| Warm asks and follow-ups | 30 | about 250 asks and 150 follow-ups | 12 to 20 customers if you had 250 names, 3 to 5 if you had 60 |
| Posting in public | 24 | 24 posts and their comment threads | 1 to 3 customers, plus an archive that keeps working after day 100 |
| Cold outreach | 38 | about 1,400 sends | about 8 customers at 0.6% per send |
| Reply-only and floor days | 8 | nothing new sent | no customers directly, and the reason several of the above closed |
| Paid ads | 0 | one capped test of 250 to 500 euros | 0 to 3 customers, and more often a number than a customer |
| Total | 100 | 150 hours across about twenty weeks | roughly 20 to 35 customers with a warm list, 8 to 12 without one |
So the challenge does not produce a hundred customers, and anyone telling you it does has not done the division. Cold outreach alone would need about 16,000 sends to reach a hundred customers at 0.6%, and a hundred sessions produces about 1,400 of them. The first 100 customers playbook works through what actually gets you to a hundred, and it is a longer road than one quarter.
What a hundred days does produce is worth more than the customer count on its own. One lane you know is real, measured rather than guessed. A message that has been sent at a volume where the reply rate means something. An archive of 22 to 25 posts that keeps bringing people in for years after you stop writing them. Twenty or so customers who can be asked for referrals, which is the highest-converting ask there is. And a block that has survived a bad quarter, which is the thing that makes the second hundred days cost half as much as the first.
If you are starting with no list, no audience and no budget at all, the shape holds but the first six weeks are almost entirely Wednesday. Marketing a SaaS with no audience and no budget covers that version, and the channels that work for indie SaaS covers which lane to weight once the first replies land.
Start on the next Monday
The obvious date to start is 1 January, and it is one of the worse ones. Your first ten warm asks and your first forty cold sends would land in the quietest fortnight of the working year, against a wall of out-of-office replies, and you would read that silence as a signal about your offer. It is not a signal about your offer. It is a signal about the calendar.
A hundred sessions counted from any Monday finishes exactly as far away as a hundred counted from New Year, because you are counting sessions. Start on the next working Monday in whatever month you are reading this. The only date that matters is the one you write next to session one.
Your first three days
No plan, no tooling, no research. Tonight, do the first three items. Tomorrow, run session one.
- Tonight: run your target through the goal calculator and change it if the arithmetic says so.
- Tonight: put a ninety-minute block in the calendar for tomorrow morning, before the product.
- Tonight: write one line. Tomorrow's lane, tomorrow's action, tomorrow's number.
- Tonight: write the three exit conditions. Thirty posts, two hundred cold sends, and a euro cap with a stop date.
- Day one: list everyone who would recognise your name, then message the first ten of them.
- Day two: publish one thing about a problem you solved last week, and answer every comment.
- Day three: thirty-five cold sends to strangers who match your customer, same message on all of them.
- Every day: five minutes at the end. The count, the replies, and one line for tomorrow.
- Every week: one row in the tracker, written honestly, including the weeks that were bad.
Session one is tomorrow or it is not happening. That is the only part of this page that has a deadline.
Frequently asked questions
- What is the 100 day marketing challenge?
- It is one distribution action, at a fixed volume, on every working day until you have logged a hundred sessions. One session is ten warm asks, or one published post plus twenty minutes of replies, or thirty-five to forty cold sends, or a capped ad test reviewed once a week. The block runs about ninety minutes and happens before the rest of the day starts, so a hundred sessions at five a week is twenty weeks and roughly 150 hours. You count sessions rather than calendar dates, which means a missed Tuesday moves the finish line by one day and does not reset anything.
- What counts as one distribution action?
- Something that left your machine and reached a person who is not you. Ten warm messages sent. One post live at a URL a stranger can open, with its comments answered. Thirty-five cold emails in the sent folder with the same words in all of them. Money spent against a cap you wrote down first. What does not count: building a list without sending to it, drafting a post you did not publish, checking a dashboard, warming a domain, choosing a CRM, or reading about marketing. All of that is preparation, and preparation is not a session.
- Can I do a 90 day marketing plan instead of 100 days?
- Yes, and it changes almost nothing, because both are counted in sessions rather than dates. Ninety sessions at five a week is eighteen weeks against twenty. The only checkpoint that moves is the last one, and the honest caution is the same either way: at one post a week you finish with 18 to 20 posts, which is below the 30 that a content lane needs before silence means anything. If ninety is the number you will actually finish, run ninety and keep posting afterwards.
- What happens if I miss a day of the challenge?
- Nothing happens to the count. The session you did not run has not happened yet, so the finish line moves by one day and every other number stays where it was. There is evidence for treating it this way rather than as a failure: in the Lally study of habit formation, missing a single opportunity did not materially affect the process. Do not double up tomorrow either. Two sessions in one morning halves the volume of each, which is the exact thing the rules exist to prevent.
- What do I do if I miss a whole week?
- Resume at the next number, not at one. If your last session was 34, the next one is 35. Make the return cheap by running three floor sessions in a row first: one warm ask, three cold sends, one reply answered. Then pick the week structure back up on the following Monday. Do not try to make up fifteen missed sessions, because catch-up weeks fail and the failure reads as proof the plan was wrong. Once you are back, ask what took the week. If it was an emergency, nothing needs fixing. If the block had a decision sitting at the front of it, fix that before session 36.
- How many hours a day does the 100 day challenge take?
- Ninety minutes, five mornings a week. Seventy minutes on the action, fifteen answering yesterday, five to write the count and name tomorrow. That is seven and a half hours a week and about 150 hours over twenty weeks. Below sixty minutes the volumes drop under the point where the result tells you anything, and above ninety it starts eating the product. On a genuinely bad day, run the floor version, which is between four and ten minutes and still counts as a session.
- Do I have to start on 1 January?
- No, and the January start is the most common way this gets abandoned in week two. A hundred sessions counted from any Monday finishes exactly as far away as a hundred counted from New Year. Starting in the first week of January does have one real cost: the people you are messaging are out of office, so your first ten warm asks and your first forty cold sends land in the quietest fortnight of the year and produce a reply rate that reads like a verdict. Start on the next Monday, whichever month that is.
- What results should I expect after 100 days of marketing?
- At these volumes, roughly 20 to 35 customers if you began with a warm list of a couple of hundred names, and closer to 8 to 12 if you began with nobody. Not a hundred. Cold outreach alone would need about 16,000 sends for a hundred customers at 0.6%, and a hundred sessions produces about 1,400. What the hundred days does reliably produce is one lane you know is real, a message tested at a volume where the number means something, an archive of 22 to 25 posts that keeps working after you stop, and a habit that has survived a bad quarter.
Source
- Lally, P., van Jaarsveld, C. H. M., Potts, H. W. W. and Wardle, J. (2010). How are habits formed: modelling habit formation in the real world. European Journal of Social Psychology, 40(6), 998-1009. onlinelibrary.wiley.com/doi/abs/10.1002/ejsp.674