Sell AI Photos to Restaurants with Nanobanana PRO (Easy guide for new agencies)

This might just be the easiest way to

get a client for your AI creative agency

and start earning in the new year

because this market is massive. In the

US alone, the restaurant industry is

[music] projected to reach $1.5 trillion

in sales in 2025. And there is one thing

that every restaurant needs, great

photos of their food. But most

restaurants are small. They don't really

have the time nor the investment range

for professional food photography.

Because of this, their Instagram and

Google Maps listings usually look like

this, [music] this, and this. Even when

the food is genuinely great, the photos

fail to show how good it really is. And

so, this is where you can step in.

[music] Because with AI, you can fix

this problem for them in minutes by

using free tools without a studio or a

camera, where you turn unappetizing food

shots into visuals that sell. And you

build trust with your clients by showing

these restaurants how incredible their

food can look. Then you can offer this

as an AIdriven brand refresh and [music]

turn them into a long-term client. So,

in this video, I'll show you exactly how

to find clients for this service, how to

package it properly and professionally,

[music] and when you're ready to deliver

and scale, I'll also show how to

automate the whole thing using creative

AI and automation tools. So, if you're

looking for a simple AI service that you

can start selling right away without

getting too technical, then this video

is for you. Let's get started.

Hey, if you're new here, my name is Jay.

I spent a decade in creative and

marketing work and half a decade leading

data teams. I'm now operating our AI

creative agency and RoboNuggets our

education arm where we have several

hundred members all AI practitioners

across the globe and here our mission is

to make creating with AI easy to learn

for anyone with AI training and

certification that most people join for

but most members stay because of the

strong community we have built. Okay

friends, so what we're going to cover in

this lesson is two parts. First I'll

show you the easiest and free way to do

this using Nano Banana Pro and Gemini

Tree. But I'll also show you how you can

do things faster if you learn about

automation using a tool like N8N. And if

you stick till the end, I'll also show a

few bonuses for you to learn, which is

how you can potentially expand your

service from beyond just food photos.

And also another nugget which we'll save

near the end of the lesson. But firstly,

I'll show you how easy it is to do this.

Because if you go to Google and you just

type something like restaurants in my

area or a specific area like in this

case in Crows Nest, Sydney, and you can

see just by viewing the maps here,

there's already a lot of potential leads

that you can offer this to. But for you

to narrow it down, there's a few things

you can keep in mind to find your ideal

client. This local restaurant, for

example, what you want to do is to view

their menu. Because for customers trying

to find a place to eat, usually what's

the customer journey is the fact that

they open Google Maps, look at the

options around them, and browse their

menu to see if the food is appetizing or

not. And for this restaurant,

unfortunately, I don't think they have

invested much into their menu listings

here. So, this seafood pa mixa, I've

actually tried this, and this is really

good. But if you just view their photo

here, I don't think it does justice to

how good this dish actually was. And

this obviously isn't really the

restaurant's fault because usually they

don't have the time nor the required

investment range in order to hire a food

photography studio to upgrade this for

them. But with AI, it's actually easy to

upgrade these food photos. Now, it's

just that not a lot of people know about

it yet. And to show how common the

problem this is for restaurants, this is

another example where if you go to their

menu, it's the same thing where the food

shots here can definitely be optimized.

And the other thing is if you go to

their Instagram, it's the same story

where their food may taste great, but

these photos don't really do it much

justice. And so what we'll do now is to

message them on Instagram showing an

example of how we can turn their current

food photos into more appetizing shots

that will surely bring in more customers

for them. And the way you do this is

actually very simple right now. You can

just browse their menu for a photo that

you want to upgrade. Let's say this one.

And once you have this photo, you can

just go to Google Gemini here in this

URL. You can copy that image by doing a

right click, paste it here, and we can

just send this prompt where we're asking

Gemini to turn this photo into a

premium, professionally styled food shot

of the same dish. I have a couple more

details around the prompt here which you

can copy or adjust which this prompt I

also just uploaded in the link below so

that you can just copy it and what you

want to do is just set it to thinking

mode so that you can use Nano Banana Pro

and under the options here you want to

select create images and just go ahead

and send. Now I'm at the free version of

Gemini here so you can do this even if

you don't have the Gemini Plus plan and

also remember that Google Gemini has a

mobile app and so this is perfect if

you're comfortable to pitch on the spot

the next time you dine out. What you can

literally do is take a photo of their

food with your mobile phone. use this

prompt to turn it into a shot like this,

which now it just finished, and then

show and share it with the owner for

free to see if it's something that

they'd be interested in to upgrade their

whole menu. So, after a few seconds,

Nano Banana Pro, which is the model we

used in Gemini, was now able to turn

this photo into a much more professional

shot. And you can see it retains all of

the ingredients. And overall, it just

makes it look better and more appetizing

for the consumer. Do note that in

Gemini, there is a small watermark here,

but in part two, I'll show how to get

these images without the watermark

whenever you generate them. And just to

show you another example, I also sent

the same prompt for this paella image

and Nano Banana Pro was able to give us

this more professionally styled food

shot. And so now what you want to do is

to go to their Instagram and send them a

message. And we can just expand this so

that you can see. And you can just take

a screenshot of that image you generated

without including the watermark if you

like. And you also want to save the

original reference image. Attach both of

those to your message and send them a

message like this. which if you break

down this message, you essentially have

this hook at the top, which is important

because if you look at your DMs on

Instagram, it shows you the preview of

your very first line. So the goal of

that first line is for the receiver to

open your message. So you can use a

simple hook like misleading food shots

in your Google listing, but in the very

first sentence, we clarify that and say

that we saw their menu on Google Maps

and thought they were misleading. You

introduce yourself by saying you're a

local creative studio owner in the area

and say that you know that their food

looks better than what they have in the

photos online. You then give them some

free value by showing a before example

of how they can be enhanced and that

they can use it in their socials if

they'd like. And then the last sentence

here gives them an idea of what you're

offering by saying that you're currently

working with a small group of local

business owners to give them a social

media brand refresh. And if they want to

be part of that small group, then they

can respond to your message here. So

that's just a simple example where you

have a hook, you have an offer, and you

also introduce a bit of scarcity by

sharing the fact that you are working

with only a small group of local

business owners. And that's part one. a

very simple and easy way for you to

start offering services using your

knowledge of AI. But now the question

becomes once you're ready to deliver

that work or if you want to do things

faster and at scale, how can you now

automate that whole process? And so this

is one example of how we did exactly

that where you can see we automated the

sourcing of the leads that will offer

the service to automated the download of

our reference images which we

transformed into two images this time so

that we have more options. And if we

look at these images, you'll also notice

that they don't have the Gemini

watermark, which is perfect for us when

we're dealing with production work. So

you can imagine if you're doing a whole

makeover of their menu and let's say you

have 30 items, you can just load all of

them here, run your automation, and

you'll instantly get these images back,

which you can package and provide to

your client. And apart from that, you

can also use this automation for scaling

your Instagram DM outreach. Because

apart from creating the images, we can

also use this system in order to get

their Instagram URLs so that we can do

that same process of reaching out to

them via socials. And that automation is

all powered by this system which we

built in N8N which for this next part

I'll show how it works so that you can

decide if you want to learn it as well.

But in case you're new, N8N is a no code

automation builder similar to Zapier or

mate.com. But what's great about N8N is

if you're watching this and you're part

of the community, you can just get this

template from our resources page here.

And when you download that and import

that file into N8N, it will load that

whole system for you from scratch. And

the way this automation works is quite

simple because we practically just have

two parts here where we first get the

restaurants where we can offer the

service to and then we create the images

for them. This purple section is more of

just a bonus if in case you want to

expand beyond images and create videos

for your client as well. And for us to

start, if you want to get multiple

restaurants that we can offer this

service to, we just click execute

workflow here in the very first row.

That will open up a form where we can

just declare how many listings or

restaurants we want. In this case, we'll

do 10. Let's say we want to get two

images per listing. And the location we

want to search in is Sydney, Australia.

So once we click submit, what it's now

going to do is to run that automation

for us that will scrape for restaurants

within Google Maps in that location that

we asked for. And when that has finished

processing, it now stores all of those

listings into this tool called Air

Table, which was the tool we were

viewing earlier that functions quite

similar to Google Sheets, except that

you can load media, photos, and videos

into the individual cells so that you

can preview them. So, it's much easier

to work with it. And they also have a

really generous feed here. So, you can

use this tool as you're getting started.

But you can see when our get restaurants

automation ran in Air Table, it gave us

the name of the restaurant, some

information about that business such as

their address, their website, and

importantly, their Instagram handle. And

we also got photos of some of their

dishes straight from Google Maps, which

we are about to turn into AI enhanced

photos all at the same time. And by the

way, this Air Table template, if you're

watching this and you're part of the

community, remember that you can just

access it in this link so that you can

duplicate it and it's much easier for

you. But just to give you an overview,

the way we got those listings is through

this tool called Appify, which if you're

completely new, Appify is basically a

collection of scrapers where you can see

here there's options on Tik Tok, there's

options on Instagram. The specific one

we use for this automation is this

Google maps scraper where you can

basically provide inputs to it like what

we did here in the form around how many

listings we want to source and the

location from which we will source that.

And whenever this node runs in N8N that

will return all the listings for us to

use in our automation. And so if we open

this node to inspect it if you're

curious in N8N you basically have the

inputs to this node here at the left.

You have the configuration here in the

middle and you have the output of this

node here at the right. If you're

interested about how we configure this

we have several other tutorials in our

platform of how we set it up. But here

you can see it returned to us those 10

restaurants along with all of the

information and images that we loaded

into Air Table. But before we loaded it

into Air Table, you notice we have an

analyze image node here which if we open

that basically the only instructions

that we're asking this node to do is to

identify if the image that we got is an

image of a single dish of food. Because

remember in Google Maps sometimes the

images that you can get there are images

of the restaurant themselves, which is

not really what we're after. And so this

node just automatically skips these

images so that we can focus on images of

actual food dishes. And when that's done

and they're log in air tableable, we

have a clean list of images that we can

now transform. And so to do that, what

you can do is run this second automation

where if we execute that stream, it's

going to get all of those rows that we

want to create for and create images for

them by batch also using Nano Banana

Pro. And when that is done, we now have

our images loaded back into Air Table,

which we can now view here. So these

turned out pretty good. And just to show

a few examples, we turned this image

into this more professional shot. We

turned this to this. And so just overall

the quality and the detail of this image

is just much better than what we started

with. And again, if you want to learn

how to set up these nodes from scratch,

we have several deep dives in the past

of how we set this up. But just to show

you the prompts we used, if we open

this, the prompt we're using for the

very first image is basically the same

as what we manually did earlier, which

is this piece of text that you can see

in this area. Now, for the second image,

since we want slightly different

options, but of the same dish, what

we're actually doing is feeding the very

first image into this node. And if we

open this, you can see the prompt that

we are giving it is also quite simple.

We're basically instructing Nano Banana

Pro to give a different camera angle

shot of the reference image. And just to

guide it further, we're asking it to

choose between a direct overhead shot, a

close-up detail shot, or a low angle

shot, which you can see in these images.

And so now with these images, you're

actually done already for this specific

service. But when you start working with

clients, usually it's much easier to

scale your services from the same client

because you've already built a

relationship with them instead of having

to onboard several new clients each

time. And so what you can do as an

extended service is to actually turn

these images into videos which to

illustrate that we also have this

automation where if we execute this

workflow is going to create those videos

for us on autopilot as well. And when

those videos have finished processing

that would now give you the videos that

we showed earlier in the intro. And

basically to get those videos what we're

doing is to set the first image as a

starting frame and the second image as

the ending frame using the cling 2.5

video model. This is outside the scope

of this specific tutorial though. So let

me know if you want me to do a breakdown

of that as well. So that's pretty much

the whole workflow in full. But just as

a few final bonus nuggets, one of the

questions that for sure you will face is

how much you will want to price this

service. And one quick technique that

you can do is that in Google, you can

actually just ask it how much is a food

photography package in your location.

And it will do some research for you on

how much the rates are. And so in the

US, these services cost somewhere along

$750 to $2,000 for only 5 to 15 images,

which obviously depending on the

business that you're pitching to, you

will likely need to adjust this

depending on their purchasing power.

Because remember, since you're doing

this via AI, if you have the skill and

you can create these images in just one

shot, your cost to produce these images

are practically only 14 if you're doing

it via the automation that we showed

earlier. Or you can even do it for free

if you're doing it manually through the

earlier manual method that we showed.

So, even if you price each image as only

$10, you still net a profit of around

98% which is just insane margins all

because you know how to use these AI

tools. And the reason why this is such a

good opportunity and skill to learn is

because this doesn't stop at just

creating for restaurants once you learn

this skill, you can apply the same

technique into any business so that you

can upgrade their branding and their

social media presence as well. And the

last point that I wanted to mention is

that as you go through this creative

process and as you set up your own AI

creative agency or studio, unfortunately

at this point since AI is early, you

will still encounter a bit of push back

around the use of AI for enhancing

images. But one mental model that you

can use to support your work is that

when you're transforming an actual image

of an actual dish, if you prompt it so

that you can request an accurate

representation of that food, but just

enhanced, you actually start to realize

that sometimes doing it this way is more

faithful to what customers will expect

versus previous methods where even if

they're not done with AI, how the final

product is represented doesn't really

provide an accurate representation of

what the customer will receive. And so

at the end of the day, AI is just a tool

and it entirely depends on the decisions

of the agency and the business of how to

represent the product, whether it's done

in a studio or whether it's done with

AI. And you will likely encounter these

questions too. So I wanted to give you a

sense of a framework of how you can

answer them. I also wanted to give kudos

to Hamza who inspired the build for this

lesson. He has more content around the

same thread of offering AI services for

small businesses. if you're interested

in that. And as mentioned, all the

templates and resources we discuss in

this lesson, you can just find them in

the Roboas community, which you can find

in the link below, which if you're

passionate about creating with AI and

want to earn from that passion, then I

encourage you to check it out. You get a

lot of exclusive AI training courses

here, including training certification

when you complete them. You can connect

with a community of 1,000 plus AI

practitioners who share paid

opportunities and potential

partnerships. And recently, we also

teamed up with 570 plus of the top AI

and SAS tools to give massive savings

and free credits for our annual members,

which actually already cover the cost of

membership. And finally, we run prized

learning events, too, like this one in

December with a $3,000 prize pool. So,

check that out to see if that's for you,

just in the link down below. And also,

if you found value in this video, then

consider subscribing because that helps

us a lot to make more educational

content like this. That's it for this

one. I'll see you guys next time.

[music] Thank you.