I have one big revenue goal for HobbyScool this holiday season and it's $150K to $200K between mid-November and mid-January. Most of that comes from one offer, the 2027 All Access Pass and I have 49 broadcasts already scheduled in Kit to sell it.
So I asked Claude a simple question. Based on everything I have scheduled, how much money is this going to bring in?
This post walks through how I had Claude estimate it from my real Kit stats, what the full forecast came out to and how I used it to find the moves that raise the total. The prompts I used and the scheduled task that checks my sales twice a week are at the end.
Check where every number comes from.
Claude gave me a clean table in about a minute, with sales counts for every price step and a total at the bottom.
Before I planned anything around it, I asked one more question: "What did you base these numbers on?"
Claude split its answer in two. My buyer tag counts, the sales already in and my prices came from my account. The conversion rates were its own assumptions.
A clean table looks the same whether the numbers came from your account or not, so ask this every time an AI tool hands you numbers. Then have it swap in your real data wherever it can.
"What did you base these numbers on? Split them into what came from my real data and what you assumed."
The question to ask every timePull the real numbers straight from Kit.
I have the Kit connector turned on in Claude, so Claude can read my Kit account directly. I had it pull what my list is doing right now and it had these in a couple of minutes:
| What it pulled | How | What it showed |
|---|---|---|
| List size | Recipients on the last Sunday newsletter sent only to that segment | About 48,000 people |
| Open rates | Stats on every broadcast from the last 90 days | About 50% on sales emails |
| Click rates | Same stats, sorted into sales emails, newsletters and buyer emails | 1% to 3% on list-wide sales emails, 8% to 21% on emails to buyers |
| Buyers so far | Purchase tag counts | How many already bought, and when |
| Best customers | Tags for people who bought 3+ and 5+ events | The size of the early access groups |
The gap between those two click rates matters a lot. Past buyers click my sales emails at several times the rate of the whole list, so they get their own line in the forecast.
The math is one line per email.
Once Claude had the real numbers, it rebuilt the forecast with one simple line for every sales email I have scheduled:
It ran that line for each email, then added the results up by price step. It also counted people who click more than one email only once, so the same buyer wasn't counted six times across six emails.
Every input got a label so I'd know what to trust:
From Kit list size, open rate, click rate, buyer counts and prices
Assumption how many clickers buy
That second one is the only number left that's a guess and there's a reason for it. Kit stops at the click. What happens on the checkout page shows up in my checkout, so that number gets real once sales start coming in.
Since so much rides on that one guess, I asked for two versions. The low case uses a lower buy rate and the expected case uses a higher one and the space between them shows how much that single assumption moves the total.
What the numbers came out to.
Here's the full forecast. It's built from a list of about 48,400, clicks of 1.5% (low) and 2.5% (expected) on list-wide sales emails, and 10% to 15% on emails to past buyers. The assumed buy rate is 2% to 3% of list clickers and 10% to 15% of buyer clickers. All figures are gross, before affiliate commissions.
2027 All Access Pass, by price step
The pass is sold in price steps: early access for past buyers, a Black Friday week price, then a higher price until it closes January 18.
| Price step | Emails | Low | Expected |
|---|---|---|---|
| Early access, $247 | 2 emails to about 1,000 past buyers and Club members | 15 sales, $3,705 | 34 sales, $8,398 |
| Black Friday week, $297 | 6 emails to 48,400 | 52 sales, $15,444 | 131 sales, $38,907 |
| December, $347 | No customer emails | 5 sales, $1,735 | 10 sales, $3,470 |
| January, $347 | 5 emails to 48,400 | 44 sales, $15,268 | 109 sales, $37,823 |
| Total | 116 sales, $36,152 | 284 sales, $88,598 |
Plus 18 passes already sold at $247 ($4,446) before the forecast window.
Black Friday week
Black Friday week has two offers running at once, the pass at $297 and a one-day lifetime deal on the HobbyScool Club.
| Item | Low | Expected |
|---|---|---|
| All Access at $297 | $15,444 | $38,907 |
| Club lifetime deal, $197, one day only | $3,431 (17 sales) | $8,582 (44 sales) |
| Week total | $18,875 | $47,489 |
This is the week with the widest range, $18,875 to $47,489, because it has the most emails going to the whole list.
The whole two months
| Item | Low | Expected |
|---|---|---|
| All Access Pass | $36,152 | $88,598 |
| Club Black Friday deal | $3,431 | $8,582 |
| Annual Pass, Hall of Fame, year-end clearance | $5,500 | $8,500 |
| December and January summits (set at $30K each) | $60,000 | $60,000 |
| Club renewals and ad revenue | $9,000 | $9,000 |
| Total | about $114,000 | about $174,700 |
The goal is $150K to $200K. The expected case lands inside it at about $174,700. The low case comes in about $36K short and most of that gap is Black Friday week.
So now I know which week to watch hardest and I know it before the week starts.
Find the moves that raise the total.
Because the forecast is built one email at a time, it also shows where the money comes from and where the gaps are. That's what tells you what to do next. Here's what mine showed.
Past buyers click at several times the rate of the whole list. My emails to buyers get 8% to 21% clicks and my list-wide sales emails get 1% to 3%, so every email to past buyers counts for a lot more per person.
You can have Claude find these for you. Give it your forecast and ask which emails and price steps bring in the most, where the biggest gaps are and what changes would raise the expected total. Have it show how much each change adds using the same math, so you can pick the ones worth your time.
Look at this forecast and show me which emails and price steps bring in the most and where the biggest gaps are. Suggest up to three changes that would raise the expected total, like adding an email, sending to a different segment or changing an offer, and show how much each one adds using the same math. Label which inputs in each suggestion come from my real data and which are assumed.
Turn the forecast into pace checkpoints.
A total at the bottom of a table doesn't help much in the middle of a launch. What helps is knowing how many buyers I need by a certain date, so I can tell whether I'm on pace while there's still time to do something about it.
So the forecast became four checkpoints. These count every All Access buyer, including the 18 who bought before the forecast window:
| All Access buyers by the end of | Low | Expected |
|---|---|---|
| Early access (Nov 22) | 33 | 52 |
| Black Friday week (Nov 30) | 85 | 183 |
| December (Dec 28) | 90 | 193 |
| Cart close (Jan 18) | 134 | 302 |
Being behind on day two of Black Friday week is fixable. Finding out in January isn't.
A scheduled task checks it twice a week.
I didn't want to remember to pull these numbers myself in the middle of the holidays, so I set up a Claude scheduled task that runs every Monday and Thursday until the cart closes. It runs on its own in the cloud and here's what it does each time:
- Reads my Kit purchase tags and email stats
- Counts buyers by price step and compares them to the checkpoints
- Projects the total from where things stand and shows the gap to my goal
- Suggests up to three moves for me to look at
- Saves a short report to my Claude project and sends a summary to my phone
It's read only. It never edits, schedules or sends an email. If it says I'm behind and suggests sending an extra email to past buyers, I decide whether that happens.
I have 49 broadcasts scheduled, each going to its own audience. A task that can change those on its own could send the wrong email to the wrong segment at the worst possible time. A task that reads and reports can't do any of that.
The prompts I used.
Here's every prompt, in order. Swap in your own offer, dates and goal. You'll need your email platform connected to Claude for prompts 3 and 5 to pull real numbers.
Look at everything I have scheduled in Kit for [offer] between [start date] and [end date]. Estimate gross revenue by price step and by email. My goal is [$X]. Show a low case and an expected case.
What did you base these numbers on? Split them into what came from my real data and what you assumed.
Pull my current list size, open rates and click rates on recent sales emails from Kit. Rebuild the forecast as list size x click rate x share of clickers who buy x price, for each email. Label every input as from Kit or assumed.
Look at this forecast and show me which emails and price steps bring in the most and where the biggest gaps are. Suggest up to three changes that would raise the expected total, like adding an email, sending to a different segment or changing an offer, and show how much each one adds using the same math. Label which inputs in each suggestion come from my real data and which are assumed.
Save this forecast to my project. Then create a scheduled task that runs twice a week until [close date]: read my Kit purchase tags and email stats, compare sales to the pace I need at each price step, project the total, show the gap to [$goal], and suggest up to three moves. Read only. Don't edit or send any emails.
What I'd do if I were you.
If you have a launch, a sale or a cart opening on the calendar, you can do this before it starts:
- Connect your email platform to Claude so it reads your numbers instead of guessing industry averages.
- Ask what every number is based on and have it label real data and assumptions.
- Run a low case and an expected case so you can see how much rides on the one number you can't know yet.
- Ask Claude for the moves that raise the total and pick the ones that add the most.
- Turn the forecast into checkpoints by date, then let a scheduled task watch them for you.
- Keep the tracker read only. You decide what changes.
For me, the forecast showed that Black Friday week decides whether I hit the goal and the tracker tells me twice a week whether I'm on pace.
Find the weak spot before your next launch.
The free Creator Business Scorecard takes two minutes and shows which part of your business needs work first.
Take the free scorecard →Frequently asked.
How do I forecast launch revenue from my email list?
Run one line for every sales email you have scheduled: people who get the email, times the click rate, times the share of clickers who buy, times the price. Add the lines up by price step. Take the list size and click rates from your email platform's recent stats instead of guessing.
Can Claude read my Kit stats?
Yes. With the Kit connector turned on, Claude can pull your broadcasts, recipient counts, open rates, click rates and tag counts, including purchase tags. It reads the numbers, so you don't have to export anything.
What number in a launch forecast is still a guess?
The share of clickers who buy. Your email platform stops at the click, so that number comes from your checkout once sales start. Run a low case and an expected case with two different buy rates so you can see how much rides on it.
How can Claude help me increase launch revenue?
Give it a forecast built one email at a time and ask which emails and price steps bring in the most, where the biggest gaps are and what changes would raise the expected total. Have it show how much each change adds with the same math so you can compare them.
How often should I check my sales against the forecast?
Twice a week works well during a launch. Turn the forecast into checkpoints, meaning how many buyers you need by the end of each price step and set up a scheduled task that compares your real sales to those checkpoints and reports back without changing anything.
Dr. Destini Copp helps digital product creators build sustainable, systems-based businesses through the Creator Growth Flywheel framework. She's the founder of Creator's MBA and HobbyScool, and has been teaching online business strategy for over a decade. Learn more →
