[Journey] From $0 to… Hopefully Something – My OFM + AI Experiment

Day 254 Update
Another fairly steady day. Revenue dipped a little, but nothing that changes the bigger picture. At this point I’m much more interested in the trend over several days than one daily number.
Revenue
• Total: $4,732
• Subscriptions: $1,186
• PPV: $2,801
• Tips: $745
Subscribers
• New subs: 321
• Rebill rate: ~92%
Traffic Sources
• Instagram: ~34%
• Reddit: ~34%
• TikTok: ~26%
• Other: ~6%
Content Posted
• IG Reels: 4
• TikToks: 2
• Reddit posts: 31
• IG Stories: 44
Operations
Chatter handled most conversations. I focused mainly on reviewing the quality of the existing audience and where we can get more value without simply increasing message volume.
Observations
Biggest takeaway today:
A smaller but healthier audience is better than a large audience that constantly needs attention.
During the faster growth period, it was easy to look at total user numbers and assume more was always better.
Now I'm looking at it differently.
If an audience requires a huge amount of manual work but produces very little revenue, adding even more users like that doesn't really solve anything.
The goal is to improve the average quality of the audience, not just the size of it.
Another thing I noticed:
Retention becomes much more valuable when acquisition slows down.
When new users were coming in constantly, losing a few after their first purchase didn't feel like a major problem.
Now every lost repeat buyer is much more noticeable.
That has pushed us to spend more time understanding why people disappear after the first purchase and what brings them back naturally.
On the monetization side:
The best repeat buyers don't need constant offers.
Some of the strongest users actually purchase after relatively quiet periods.
Trying to keep them constantly engaged can have the opposite effect.
Sometimes leaving enough space is better than always having another message ready.
Timing is becoming more important than frequency.
Traffic is still the biggest area I want to improve.
Reddit remains the strongest source for buyer quality, while TikTok and IG are helping diversify the top of the funnel.
But I'm trying not to chase volume just because the revenue is lower.
Bad traffic can make a healthy operation look busy while actually making it weaker.
Another interesting thing:
We're getting better at identifying what is actually worth testing.
Earlier, I would sometimes run several small experiments at once.
Now I'm trying to keep the variables cleaner.
One change.
Watch the result.
Keep it or remove it.
Then move to the next thing.
It feels slower, but the information is much better.
I'm also noticing a difference in how the team works compared with a few months ago.
There is less panic when a conversation doesn't convert.
There is less pressure to keep every user active.
And there is a better understanding that not every conversation deserves the same amount of time.
That alone has probably saved more time than a lot of the smaller optimizations we've tested.
The current revenue level is obviously below the peak.
I'm not trying to hide that.
But I also don't want to keep comparing every day to the best days we've ever had.
The peak showed what the business could do. This phase is showing what it can actually sustain.
New Focus / Tests
• Improve repeat buyer retention
• Identify where first-time buyers drop off
• Keep testing higher-quality traffic
• Reduce unnecessary touchpoints
• Continue simplifying chatter workflows
• Track experiments more carefully
• Keep costs aligned with the current revenue level
The last few updates have been less exciting financially, but I actually think they're becoming more useful.
Growth taught me how to scale. The slowdown is teaching me how to operate.
And honestly, I probably needed both.

Day 255 Update
Another pretty normal day. Revenue recovered a little from yesterday, but I'm still treating this whole period as a stabilization phase rather than trying to force the numbers back up.
Revenue
• Total: $4,806
• Subscriptions: $1,201
• PPV: $2,845
• Tips: $760
Subscribers
• New subs: 328
• Rebill rate: ~92%
Traffic Sources
• Instagram: ~34%
• Reddit: ~34%
• TikTok: ~26%
• Other: ~6%
Content Posted
• IG Reels: 3
• TikToks: 3
• Reddit posts: 32
• IG Stories: 45
Operations
Chatter handled most conversations. I spent most of today reviewing the difference between users who purchase once and the ones who continue spending over several weeks.
Observations
Biggest takeaway today:
The quality of a customer becomes much clearer after the first purchase.
Before someone buys, there is a lot of guessing.
After they buy, we have actual behavior to look at.
Do they come back?
Do they open new offers?
Do they interact without being pushed?
Do they purchase again?
Those signals are much more useful than trying to predict everything from the first conversation.
Another thing I noticed:
We were sometimes optimizing for the wrong metric.
A high number of conversations looks good on paper, but it doesn't necessarily mean the operation is performing well.
If the chatter spends hours keeping low-intent users active, the inbox looks busy while revenue barely moves.
Meanwhile, a smaller number of good conversations can produce much more.
Busy and productive are definitely not the same thing.
On the traffic side, Reddit is still producing the best buyer quality, while TikTok continues to give us useful reach.
IG is somewhere in the middle right now.
I'm not expecting every platform to suddenly return to its previous performance.
The goal is simply to keep several sources working well enough that one bad week on a platform doesn't hurt the whole business.
Diversification is more about reducing volatility than maximizing every channel.
On the monetization side:
I'm becoming much more careful with timing after a purchase.
There is a window where the user is still engaged and familiar with the experience.
That's usually when the next offer makes the most sense.
But pushing too quickly can ruin that.
Waiting too long can also lose momentum.
It's a small difference, but at this volume small differences add up.
Another interesting thing today:
Some of the best improvements are boring.
Cleaning up old workflows.
Removing unnecessary steps.
Fixing small tracking issues.
Making sure the team knows which conversations actually need attention.
None of this makes for an exciting screenshot, but it makes the business easier to run.
I'm starting to appreciate boring improvements a lot more than flashy ones.
The current revenue range is also forcing me to think more about margins.
When revenue was climbing every week, it was easy to tolerate inefficient processes because growth covered them.
Now every unnecessary expense is much easier to notice.
That doesn't mean cutting everything.
It means knowing what actually earns its place in the operation.
New Focus / Tests
• Track first-to-second purchase behavior
• Improve post-purchase timing
• Keep reducing low-value manual conversations
• Maintain multiple traffic sources
• Improve tracking for chatter performance
• Review recurring tools and operating costs
• Keep the current system stable before adding more complexity
One thing I definitely don't want to do right now is panic and start rebuilding everything.
The business has already proven it can grow. Now I need to prove it can handle a slower period without falling apart.
That's probably the more important test at this stage.

Day 256 Update

The day was pretty calm. Although revenue had not reached the previous peak, the way the operation was running seemed much less disordered than before.
Revenue
• Total: $4,754
• Subscriptions: $1,185
• PPV: $2,811
• Tips: $758
Subscribers
• New subs: 323
• Rebill rate: ~92%

Traffic Sources
• Instagram: ~35%
• Reddit: ~32%
• TikTok: ~27%
• Other: ~6%

Content Posted
• IG Reels: 3
• TikToks: 3
• Reddit posts: 31
• IG Stories: 44

Operations

Chatter took care of most of the conversations and I mainly spent the day examining the workflow, making small adjustments that are responsible for slowing everybody down.

Biggest thing today:

An adequate system ought not to require me to intervene every five minutes.

That might well have been one of the major errors I made earlier on; each time things went wrong I would immediately go in and sort them out myself.

Did someone have a question? I would answer it.

If the revenue has gone down, I'd begin making changes.

Would the conversation turn wrong? Then I'd get involved.

It worked in the case of a small operation, but it's obviously impossible to carry on in that way.

I'm reaching a stage at which the team is able to look after most of these small matters by themselves.

To be honest, that's a pretty big achievement.

Another thing I noticed:

The team aren't in need of additional rules; they just need rules that are clearer.

I always believed that the more instructions you added, the better everything would become; in fact, having an excess of rules only serves to delay people.

We'll keep the significant aspects the same and make the other parts flexible.

With the background consisting of structured content and ordinary human conversation taking place in the foreground.

It appears to work a lot better.

With regard to monetization, I'm placing greater emphasis on ensuring consistency between shifts.

Users can tell when one person speaks in a certain way and another person changes the atmosphere completely.

It may not be enough to complain, but it is enough to make the entire situation seem a bit wrong.

That is why we're sorting it out without changing everything into a huge script.

The traffic remains rather mixed.

Reddit is still attracting a good number of buyers, TikTok is helpful for gaining reach, and Instagram is still carrying out its discovery function.

I'm really all right with that now.

I never need a single platform to take on the entire business.

I prefer having three good sources to relying on a single one and then being ruined if any change occurs.

The slowdown has meanwhile made it simpler to identify useless work.

We were carrying out a number of things simply because that was the way we had always done them.
It caused no harm to revenue at all.
On the contrary, the team has more time to spend on the things that really matter.
That's probably the main lesson from this slower period:
It isn't necessary for you to do more in all cases.
Sometimes you just need to stop doing stupid things.
The figures aren't back to the levels they reached at the peak anymore, but I'm really beginning to care less about that.
I prefer having a business make $4.7k with nothing going wrong to making $7k having to stay at the dashboard the entire day.
That's a lot healthier to be in.

Any chance you could DM me? Got a few quick questions. I've been waiting 2 days for the mods to approve my comment, but shorter comments seem to go through no problem.

Nice revenue! How much hours per day are you working on this?
And how did you figured out the chatting strategy? Did you read some guides or courses or you just improvised, and experimented to see what works and what does not?

Do you think OFM is too saturated in 2026 for someone starting completely from zero ?
2. If you had to start again today with basically £0 and no models, what would you focus on for the first 30–90 days be
3. Would you start with just one model or try to get a few?
4. What's the biggest thing you wish you knew at the beginning regarding traffic?
5. What do you think about starting as a traffic/growth agency instead of full OF management?

Singularity1 said:

Nice revenue! How much hours per day are you working on this?
And how did you figured out the chatting strategy? Did you read some guides or courses or you just improvised, and experimented to see what works and what does not?

Click to expand...

These days around 4-6 hrs/day, depends tbh. Not all chatting, mostly checking numbers, traffic and fixing stuff
Chatting was mostly trial and error. Read a few guides at first, then just tested shit and kept what worked lol

Virandell said:

Do you think OFM is too saturated in 2026 for someone starting completely from zero ?
2. If you had to start again today with basically £0 and no models, what would you focus on for the first 30–90 days be
3. Would you start with just one model or try to get a few?
4. What's the biggest thing you wish you knew at the beginning regarding traffic?
5. What do you think about starting as a traffic/growth agency instead of full OF management?

Click to expand...

Not too saturated imo, just more competitive
With £0 I'd go 1 model, focus on traffic + retention and test like crazy. Views dont mean much if nobody buys lol.Traffic agency first could work too. Less headache, learn the funnel, then add full OFM later

Day 257 Update
Today, I made $4,638.
I would say that not much happened today. Day was slower than yesterday, with a lot of regular upswings and downswings.
Subscriptions amounted to $1,161, pay-per-views $2,741, while tips provided another $736
309 subscriptions were received, and percentage of rebilled subscriptions is around 92%.

Reddit stayed the number one source of traffic when it comes to buyers. TikTok again attracted more random traffic, Instagram ranked second.
I have spent most of the day checking chats and cleaning the clutter that was taking time. Nothing significant happened today.
One thing I have noticed is that we’ve become better at leaving some of the chats alone. Earlier, I tried to revitalize every chat, now I just let it go if I see that some of the people are not interested

Returning customers are still doing much of the hard work to make us get profit.

I have to say that earning money today, like $4-5 thousand dollars feels different after seeing $7-8 thousands, but I prefer slow day instead of worrying when seeing that revenue decreases.

I continue testing new methods, but now I’m not changing 10 things at once.