Oscar

I automate businesses with AI
Steal my process 👉

[Oscar @Elixify_00]

Every business owner I've watched buy AI automation bought the same
first thing. Not a chatbot. Not an "AI agent."

They bought their evenings back.

The pitch that closes isn't "AI will transform your operations." It's:

"Right now, how many calls go to voicemail in a week? What's a job
worth to you? Okay — that's $9,000 a month. I can catch most of it
for a few hundred."

Three questions. Their numbers, not yours. The math does the selling.

WHO ACTUALLY PAYS FOR THIS

- Trades with crews in the field: HVAC, plumbing, roofing, landscaping
- Anything appointment-based: salons, dental, vets, med spas
- Anyone whose average job is over $200, because the math gets
obvious fast

WHO DOESN'T

- Retail storefronts with someone at the counter all day
- Businesses under ~$150k/year — the pain isn't sharp enough yet
- Anyone whose leads all come from one referral partner

Sell to the guy in the attic who can't answer his phone. Not to the
guy with a receptionist.

Free ebook breaks the whole build down. Follow + repost.

[Oscar @Elixify_00]

One automation. 49 seconds to fire. $6,800 a month.

A mobile dog groomer in Charlotte set up a single text that sends
automatically when a call goes unanswered:

"Hey, sorry we missed you — we're with a dog right now. What are you
looking to book? I can get you on the schedule right here."

That's it. That's the whole build. One tool, one message, about 20
minutes in GoHighLevel or Textedly.

Reply rate: 44%. Booking rate on those replies: 61%.

The people calling you already chose you. They just won't wait.

Follow me + repost and I'll send the free ebook over.

[Oscar @Elixify_00]

Every business owner I've watched buy AI automation bought the same
first thing. Not a chatbot. Not an "AI agent."

They bought their evenings back.

The pitch that closes isn't "AI will transform your operations." It's:

"Right now, how many calls go to voicemail in a week? What's a job
worth to you? Okay — that's $9,000 a month. I can catch most of it
for a few hundred."

Three questions. Their numbers, not yours. The math does the selling.

WHO ACTUALLY PAYS FOR THIS

- Trades with crews in the field: HVAC, plumbing, roofing, landscaping
- Anything appointment-based: salons, dental, vets, med spas
- Anyone whose average job is over $200, because the math gets
obvious fast

WHO DOESN'T

- Retail storefronts with someone at the counter all day
- Businesses under ~$150k/year — the pain isn't sharp enough yet
- Anyone whose leads all come from one referral partner

Sell to the guy in the attic who can't answer his phone. Not to the
guy with a receptionist.

Free ebook breaks the whole build down. Follow + repost.

[Oscar @Elixify_00]

A dental practice in Boise made $38,000 from 1,100 people who already
knew them. They spent $0 on ads to do it.

The list was patients who hadn't been in for 14+ months. It had been
sitting in their software the whole time, untouched, because nobody
had time to call 1,100 people.

The campaign:

Text 1 — "It's been a while since your last cleaning. Want me to find
you a time?"
Text 2 (4 days later, only to non-responders) — one line about the
new hygienist and Saturday hours.
Text 3 (9 days later) — a specific offer with a real deadline.

Results: 137 replies, 94 booked, average first-visit value $405.

Total cost: the texting tool, about $60 for the month.

Every business with more than two years of history has this list.
Almost nobody runs it, because it feels like admitting you let people
drift away.

Run it anyway.

Free ebook has the exact three-message sequence. Follow + repost and
it's yours.

[Oscar @Elixify_00]

One automation. 49 seconds to fire. $6,800 a month.

A mobile dog groomer in Charlotte set up a single text that sends
automatically when a call goes unanswered:

"Hey, sorry we missed you — we're with a dog right now. What are you
looking to book? I can get you on the schedule right here."

That's it. That's the whole build. One tool, one message, about 20
minutes in GoHighLevel or Textedly.

Reply rate: 44%. Booking rate on those replies: 61%.

The people calling you already chose you. They just won't wait.

Follow me + repost and I'll send the free ebook over.

[Oscar @Elixify_00]

A dental practice in Boise made $38,000 from 1,100 people who already
knew them. They spent $0 on ads to do it.

The list was patients who hadn't been in for 14+ months. It had been
sitting in their software the whole time, untouched, because nobody
had time to call 1,100 people.

The campaign:

Text 1 — "It's been a while since your last cleaning. Want me to find
you a time?"
Text 2 (4 days later, only to non-responders) — one line about the
new hygienist and Saturday hours.
Text 3 (9 days later) — a specific offer with a real deadline.

Results: 137 replies, 94 booked, average first-visit value $405.

Total cost: the texting tool, about $60 for the month.

Every business with more than two years of history has this list.
Almost nobody runs it, because it feels like admitting you let people
drift away.

Run it anyway.

Free ebook has the exact three-message sequence. Follow + repost and
it's yours.

[Oscar @Elixify_00]

Never let AI handle these five things. I don't care how good the tool
is.

1. An angry customer. Anger needs a human voice within the first
sentence or it escalates.
2. A real emergency — burst pipe, animal in pain, gas smell.
3. A custom quote with real variables. Guessing a number you can't
honor is worse than saying "let me check."
4. A discount request on a big job. That's a negotiation.
5. Anything the system doesn't actually understand.

A good setup says "let me get someone for you" and means it. A bad one
guesses and embarrasses you in front of a customer you spent $200 to
acquire.

The line is simple. If the answer is identical every single time,
automate it. If it requires judgment that could damage trust when it
goes wrong, route it to a person.

That's the whole rule.

Free ebook covers the escalation scripts. Follow + repost.

[Oscar @Elixify_00]

This conversation booked a $310 job at 11:48 on a Sunday night. No
human was awake.

Customer: hey do you guys do move out cleans
AI: We do! What size is the place — how many bed/bath?
Customer: 3 bed 2 bath
AI: Got it. Move-out cleans for a 3/2 run $290-340 depending on
condition. I have Thursday at 9 or Saturday at 1 — want one?
Customer: saturday. do you do inside the oven
AI: Yes, oven's included in move-out. You're booked for Saturday at
1. Can I grab the address?

Four replies. Under two minutes. Zero involvement from the owner,
who was asleep.

The thing people picture when they hear "AI receptionist" is a 2009
phone tree. This is a text thread that knows your prices and your
calendar.

Follow + repost and I'll send you the free ebook that walks through
building it.

[Oscar @Elixify_00]

Never let AI handle these five things. I don't care how good the tool
is.

1. An angry customer. Anger needs a human voice within the first
sentence or it escalates.
2. A real emergency — burst pipe, animal in pain, gas smell.
3. A custom quote with real variables. Guessing a number you can't
honor is worse than saying "let me check."
4. A discount request on a big job. That's a negotiation.
5. Anything the system doesn't actually understand.

A good setup says "let me get someone for you" and means it. A bad one
guesses and embarrasses you in front of a customer you spent $200 to
acquire.

The line is simple. If the answer is identical every single time,
automate it. If it requires judgment that could damage trust when it
goes wrong, route it to a person.

That's the whole rule.

Free ebook covers the escalation scripts. Follow + repost.

[Oscar @Elixify_00]

A 2-truck HVAC company in Mesa was losing $14,200 a month and had no
idea. Nobody stole anything.

They missed 41 calls in 30 days. Not ignored — the guys were in
attics.

31 of those callers never left a voicemail.
9 called a competitor within 6 minutes.
The average job for that company is $470.

41 missed × 60% who actually would've booked = ~25 jobs
25 × $470 = $11,750, plus two system replacements they never heard
about

They didn't need more ads. They needed someone to pick up.

Go check your phone's missed-call log for last month. Count them.
Multiply by your average job. That number is the whole reason this
matters.

Free ebook has the full audit. Follow me and repost this and I'll
send it.

[Oscar @Elixify_00]

This conversation booked a $310 job at 11:48 on a Sunday night. No
human was awake.

Customer: hey do you guys do move out cleans
AI: We do! What size is the place — how many bed/bath?
Customer: 3 bed 2 bath
AI: Got it. Move-out cleans for a 3/2 run $290-340 depending on
condition. I have Thursday at 9 or Saturday at 1 — want one?
Customer: saturday. do you do inside the oven
AI: Yes, oven's included in move-out. You're booked for Saturday at
1. Can I grab the address?

Four replies. Under two minutes. Zero involvement from the owner,
who was asleep.

The thing people picture when they hear "AI receptionist" is a 2009
phone tree. This is a text thread that knows your prices and your
calendar.

Follow + repost and I'll send you the free ebook that walks through
building it.

[Oscar @Elixify_00]

A 2-truck HVAC company in Mesa was losing $14,200 a month and had no
idea. Nobody stole anything.

They missed 41 calls in 30 days. Not ignored — the guys were in
attics.

31 of those callers never left a voicemail.
9 called a competitor within 6 minutes.
The average job for that company is $470.

41 missed × 60% who actually would've booked = ~25 jobs
25 × $470 = $11,750, plus two system replacements they never heard
about

They didn't need more ads. They needed someone to pick up.

Go check your phone's missed-call log for last month. Count them.
Multiply by your average job. That number is the whole reason this
matters.

Free ebook has the full audit. Follow me and repost this and I'll
send it.

[Oscar @Elixify_00]

Stop wasting hours manually juggling software that should work seamlessly.
A dentist pays $1,900/month for a receptionist doing tasks automation could handle in minutes, without adding more tools.

Check these stats:
- 4 minutes checking an insurance card, 15 times/week
- 2 minutes hunting a phone number, 20 times/week
- 1.5 minutes confirming a hygienist's schedule, 10 times/week

That's nearly 2 hours weekly, over 200 hours a year, just spent moving data between systems.
And nobody complains, these small tasks get ignored because they seem minor.

The real fix isn't more software.
It's connecting what already exists into one smooth path, from first contact to follow-up, with zero manual data entry.

Start by tracking your repetitive tasks for a week.
Identify the most common one.
Map it out loud.
Use AI to turn that map into a simple automation plan.
Build it with tools like Make or Zapier.
Test for two weeks, then automate the next task.

Stop guessing.
Start watching.
And watch your efficiency skyrocket.

Like this post, and I'll send you a full guide for free.

[Axel @AxelscalesAI]

A med spa in Scottsdale was burning $15,390 a month on empty chairs.

Nobody stole those clients. They just didn't show up.

THE MATH

388 booked appointments last month.
81 no-shows or same day cancels. That's 21%.
Average service: $190.

81 x $190 = $15,390 of staffed, scheduled, paid for time with nobody in the chair.

Her injectors still got paid. Rent still got paid. The room sat there with the lights on.

WHY IT KEEPS HAPPENING

Owners treat no-shows like weather. Something that happens to you.

It's a reminder problem and a friction problem. Both fixable in an afternoon.

THE 3 CHANGES

1. Deposits. Stripe or Square, $40 held at booking and applied to the service. This one change cut her no-show rate roughly in half before a single automation went live.

2. A real reminder sequence, by text, not email. Three touches:
48 hours out, confirm with a one tap reschedule link
24 hours out, what to expect and where to park
2 hours out, short and warm

3. Automatic waitlist fill. When someone cancels, Acuity or https://t.co/GRF33k18ES texts the next three people on the waitlist. First reply takes the slot.

THE WAITLIST IS WHERE THE MONEY IS

Actual cancellation, 9:40am on a Wednesday. A 2pm opens up.

9:41am TO THE CLIENT WHO CANCELLED:
"No problem at all, Dana. Freeing that up now. Want Friday 11am or next Tuesday 2pm?"

9:43am DANA:
"friday works"

9:41am, same minute, TO THREE PEOPLE ON THE WAITLIST:
"Hi Renee, this is Kayla at ___. A 2pm opened up today for your lip filler. First to reply gets it. Want it?"

9:44am RENEE:
"YES please"

9:44am SYSTEM:
"Locked in, see you at 2. Skip the ice on the way, we've got some here."

9:44am TO THE OTHER TWO:
"That one just got taken, sorry! You're still first up for the next opening."

Four minutes. Nobody on staff touched it. The chair never went empty.

WHAT IT COSTS

Acuity runs $27/mo on the tier she needed. Quo for the shared business number and texting is another $20 or so per seat. Stripe takes its cut per transaction.

Call it $120 a month all in. If you want HIPAA coverage with a signed BAA, Textline handles that properly but starts around $149/mo on its own, so budget for it rather than being surprised by it.

WHERE SHE LANDED

No-show rate: 21% down to 7%.
27 empty slots instead of 81.
54 recovered appointments x $190 = $10,260 a month.

THE GUARDRAIL

Test your deposit wording on real clients before rolling it out.

Her first version: "a $40 deposit is required to hold your appointment." Cold. Three people pushed back in the first week and one mentioned it in a review.

New version: "we hold your spot with $40, and it comes right off your service."

Same money. Zero complaints since.

People don't mind paying. They mind feeling distrusted.

An empty chair costs you exactly what a full one does.

Like this post and I'll send you the ebook.

[Oscar @Elixify_00]

Stop wasting hours manually juggling software that should work seamlessly.
A dentist pays $1,900/month for a receptionist doing tasks automation could handle in minutes, without adding more tools.

Check these stats:
- 4 minutes checking an insurance card, 15 times/week
- 2 minutes hunting a phone number, 20 times/week
- 1.5 minutes confirming a hygienist's schedule, 10 times/week

That's nearly 2 hours weekly, over 200 hours a year, just spent moving data between systems.
And nobody complains, these small tasks get ignored because they seem minor.

The real fix isn't more software.
It's connecting what already exists into one smooth path, from first contact to follow-up, with zero manual data entry.

Start by tracking your repetitive tasks for a week.
Identify the most common one.
Map it out loud.
Use AI to turn that map into a simple automation plan.
Build it with tools like Make or Zapier.
Test for two weeks, then automate the next task.

Stop guessing.
Start watching.
And watch your efficiency skyrocket.

Like this post, and I'll send you a full guide for free.

[Oscar @Elixify_00]

Somebody answered your customer at 9:47 last night.

It wasn't you. You were asleep, which is normal, and correct, and exactly why it keeps happening.

Track it for one week. Every call, text and DM that lands after you close. Just a tally on your phone.

A flooring shop did this and counted 61 after-hours contacts in seven days. They had booked 2.

61 - 2 = 59 people who reached out and got nothing.
Even if only a third were serious: 20 jobs.
20 x $890 = $17,800.

In one week. From people who were already looking at their phone thinking about giving them money.

Fix: an auto-reply that asks one qualifying question and offers two specific time slots. Not "we'll get back to you." Two actual times. ManyChat for DMs, OpenPhone for text. Under an hour to set both up.

"We have openings this week" books at about half the rate of "I've got Thursday 10am or Friday 2pm, which works?"

Same system. Different sentence.

Run the week-long tally before you buy anything. The number will make the decision for you.

Like this and I'll send you the full 5 book guide to Ai automation.

[Oscar @Elixify_00]

He found out he lost the job standing in line at a gas station.

Guy ahead of him mentions the new deck. Says the crew did nice work. Names a competitor.

That was his estimate. He'd walked that backyard. He'd shaken the man's hand. He'd said "I'll get you numbers by Friday" and then Friday turned into a Tuesday eleven days later.

Nobody sends you a notification when you lose. You just find out in public, later, from someone who assumed you already knew.

WHAT THE DELAY ACTUALLY COST

He runs about 9 estimates a week.
His quotes went out in 4-6 days on average.
Close rate at that speed: 22%.

He moved quotes to same-day, automated.
Close rate went to 38% over the next quarter.

9 estimates x 4 weeks = 36 a month.
36 x 22% = 7.9 jobs.
36 x 38% = 13.6 jobs.
Difference: 5.7 jobs a month.
5.7 x $3,200 average = $18,240 a month.

Same leads. Same crew. Same prices. He just stopped being the slow one.

HOW HE DID IT, START TO FINISH

He records a 2-minute voice memo in the truck before he pulls out of the driveway. Measurements, materials, anything the homeowner said they cared about.

That memo goes into Otter or a Claude project with a saved prompt: "Turn this voice memo into a line-item estimate using my pricing sheet. Flag anything I didn't specify."

He reviews it, fixes the two things that are wrong, and it goes out through Jobber with e-signature attached.

Total time per estimate: 11 minutes, down from about 40. Homeowner has it before dinner.

THE GUARDRAIL

He reviews every single one before it sends. Every one.

A wrong measurement in a fast quote is worse than a slow quote, because now you either eat the difference or call a customer to raise the price after they already said yes. Neither ends well.

AI writes it. You sign it.

The customer who picks someone else is never angry at you. That's what makes it so easy to keep losing.

Like this and I'll send you the 5 book bundle to Ai automation.

[Oscar @Elixify_00]

Somebody answered your customer at 9:47 last night.

It wasn't you. You were asleep, which is normal, and correct, and exactly why it keeps happening.

Track it for one week. Every call, text and DM that lands after you close. Just a tally on your phone.

A flooring shop did this and counted 61 after-hours contacts in seven days. They had booked 2.

61 - 2 = 59 people who reached out and got nothing.
Even if only a third were serious: 20 jobs.
20 x $890 = $17,800.

In one week. From people who were already looking at their phone thinking about giving them money.

Fix: an auto-reply that asks one qualifying question and offers two specific time slots. Not "we'll get back to you." Two actual times. ManyChat for DMs, OpenPhone for text. Under an hour to set both up.

"We have openings this week" books at about half the rate of "I've got Thursday 10am or Friday 2pm, which works?"

Same system. Different sentence.

Run the week-long tally before you buy anything. The number will make the decision for you.

Like this and I'll send you the full 5 book guide to Ai automation.

[Oscar @Elixify_00]

There are 412 people in your phone who already told you they wanted to buy from you.

You quoted them. They said "let me think about it." You meant to follow up. You never did.

That list is not dead. It is the warmest money you will ever touch, sitting in a folder you haven't opened since spring.

Let me show you what it's worth.

THE MATH ON YOUR OWN GRAVEYARD

Pull your last 18 months of quotes. Say 412 of them never closed.
Industry pattern on cold quote revival is 6-9% when the follow-up is actually good. Use 7%.
412 x 0.07 = 28 jobs.
28 x $1,450 average ticket = $40,600.

From people who already knew your name. Already had your number. Already wanted it.

You didn't need a single new ad to earn that.

A cabinet installer I know ran this on a list of 190 old quotes. He booked 11. $23,800 in six weeks. His total ad spend that quarter was zero.

WHY YOU DIDN'T FOLLOW UP

Not laziness. Follow-up is a background task. It never yells at you the way a broken truck yells at you. So it gets pushed to "later," and later quietly becomes never.

You cannot fix that with willpower. You fix it with a system that runs whether you remember or not.

BUILD IT IN ONE AFTERNOON

1. Export every unclosed quote from the last 18 months into a spreadsheet. Name, phone, service, quote amount, date. 30 minutes.
2. Tag them into three buckets: quoted under 90 days, quoted 3-12 months, quoted 12+ months. Each gets a different message.
3. Drop the sheet into Claude or ChatGPT with this: "Write a 3-message text sequence for each bucket. Reference the exact service and quote amount. Sound like a contractor texting from his truck, not a marketing email. Under 30 words each."
4. Load into GoHighLevel, Podium or a Google Sheet + https://t.co/IGacJv1v3H setup. Space the messages 4 days apart. 45 minutes.
5. Send 40 a day, not 412 at once. You want to handle the replies like a human.

THE THING THAT BREAKS IT

Build the exit ramp first. The second someone replies, every remaining automated message for that person has to stop. Nothing destroys trust faster than a "just checking in!" text arriving 90 minutes after a real conversation.

Test it on yourself before it touches a customer. Reply mid-sequence and confirm the rest actually dies.

Also: no blasts. If a message could have been sent to all 412 people unchanged, it will read like it was.

The worst part of losing a customer this way is that it costs you nothing today. That's why it keeps happening.

Like this and I'll send you the follow-up book with all 14 sequences written out.

Not free.

[Oscar @Elixify_00]

He found out he lost the job standing in line at a gas station.

Guy ahead of him mentions the new deck. Says the crew did nice work. Names a competitor.

That was his estimate. He'd walked that backyard. He'd shaken the man's hand. He'd said "I'll get you numbers by Friday" and then Friday turned into a Tuesday eleven days later.

Nobody sends you a notification when you lose. You just find out in public, later, from someone who assumed you already knew.

WHAT THE DELAY ACTUALLY COST

He runs about 9 estimates a week.
His quotes went out in 4-6 days on average.
Close rate at that speed: 22%.

He moved quotes to same-day, automated.
Close rate went to 38% over the next quarter.

9 estimates x 4 weeks = 36 a month.
36 x 22% = 7.9 jobs.
36 x 38% = 13.6 jobs.
Difference: 5.7 jobs a month.
5.7 x $3,200 average = $18,240 a month.

Same leads. Same crew. Same prices. He just stopped being the slow one.

HOW HE DID IT, START TO FINISH

He records a 2-minute voice memo in the truck before he pulls out of the driveway. Measurements, materials, anything the homeowner said they cared about.

That memo goes into Otter or a Claude project with a saved prompt: "Turn this voice memo into a line-item estimate using my pricing sheet. Flag anything I didn't specify."

He reviews it, fixes the two things that are wrong, and it goes out through Jobber with e-signature attached.

Total time per estimate: 11 minutes, down from about 40. Homeowner has it before dinner.

THE GUARDRAIL

He reviews every single one before it sends. Every one.

A wrong measurement in a fast quote is worse than a slow quote, because now you either eat the difference or call a customer to raise the price after they already said yes. Neither ends well.

AI writes it. You sign it.

The customer who picks someone else is never angry at you. That's what makes it so easy to keep losing.

Like this and I'll send you the 5 book bundle to Ai automation.