I gave ChatGPT my skills, time and income goal. Here’s the side hustle it picked

To find out, I gave ChatGPT my professional experience, interests, available time, starting budget and income target. I asked it to suggest realistic options, choose the one most suited to me and then build a four-week plan to test it.

I also explained the practical restrictions I would face while working on something alongside my job, although I have kept some personal details private.

I first gave ChatGPT my complete profile

I did not want AI to immediately throw a list of side-hustle ideas at me. I gave ChatGPT six things:

• Skills
• Professional experience
• Interests
• Available time: 11 hours a week
• Starting budget
• Income target

I also explained the practical restrictions I would face while working on something alongside my job.

Instead of asking it to suggest ideas immediately, I first asked it to analyse and understand my profile.

ChatGPT converted my details into a business profile, identified my professional experience as an advantage and calculated that my available time worked out to roughly 11 hours a week.

It also understood that I did not want generic “make money online” suggestions. Any idea had to fit my skills, time, budget and income target.

The five side hustles ChatGPT suggested

I then asked ChatGPT to suggest three to five side-hustle ideas based on my profile. I wanted it to explain how each could generate income, identify what a potential first customer would look like and point out the biggest obstacle I might face with each option.

ChatGPT shortlisted five options.

• Editorial/AI-assisted content service: Sell research, article writing, editing and SEO-ready content to founders, professionals and small businesses. The main challenge would be positioning myself as more valuable than a generic freelance writer.
• Faceless Reels creation service: Create faceless educational Reels for businesses and creators. ChatGPT identified production speed as the biggest challenge, given my limited weekly time.
• Book-cover and other author-required design services: Create ebook covers, paperback covers, series covers, mock-ups and promotional graphics. The main challenge would be building a credible portfolio before approaching customers.
• Digital Canva products and templates: Create reusable digital products such as templates and planning tools that customers can purchase online. ChatGPT said distribution could be a bigger challenge than creation.
• Fiction and author services: Offer services such as story development, editing, book descriptions and author launch material. The main challenge would be avoiding too many different offerings and losing a clear commercial position.

The useful part was that ChatGPT also identified the biggest obstacle for each.

For content services, it said the challenge would be positioning myself as more valuable than a generic freelance writer. For Reels, it identified production speed as the biggest challenge. For book covers, it identified the need for a credible portfolio. For digital products, it said distribution could be a bigger challenge than creation.

Why did AI finally choose?

I did not want to choose an idea simply because it sounded interesting, so I asked ChatGPT to select the option most likely to generate my first Rs 40,000 within one month. I also asked it to explain its reasoning and identify the three biggest reasons the idea could fail for someone in my situation.

It recommended editorial/AI-assisted content packages. Its reasoning was that this offered the “shortest distance between what you already know and something a stranger will pay for.”

Rather than simply selling myself as a writer, ChatGPT suggested positioning the service as: “I’ll take a topic/problem from you and turn it into publication-ready content.”

It also suggested different combinations of clients and monthly packages instead of relying on a large volume of low-priced assignments.

But ChatGPT did not present the income target as easy.

It identified three likely reasons the plan could fail – I could spend the entire month preparing instead of selling; limited time could make underpricing dangerous; and my broad range of skills could make it difficult to communicate one clear commercial offer.

“A customer should immediately understand one clear commercial offer,” ChatGPT said.

Could AI turn the idea into a four-week plan?

Once the idea was selected, I asked ChatGPT to turn it into a four-week execution plan, with specific weekly tasks, measurable milestones and a fallback strategy if I failed to meet those milestones.

It created a “4-Week Rs 40,000 Side-Hustle Sprint.”

Week 1 focused on becoming sellable rather than perfectly prepared. It suggested defining the target customer, creating packages, preparing three writing samples, building a simple portfolio, drafting an outreach message and creating a prospect list.

The target was at least three samples, more than 30 prospects and 15–20 personalised outreach messages. The expected income at this stage was modest: Rs 0– Rs 5,000, with a first paid trial as a possible outcome.

Week 2 focused on customer acquisition through personalised outreach, freelance projects and follow-ups.

By Week 3, the focus moved to creating a repeatable delivery system – from client brief and research to writing, editing, SEO checks and delivery.

Week 4 focused on converting one-off assignments into recurring packages and moving towards the income target.

ChatGPT also suggested tracking six numbers – prospects contacted, replies, conversations, paying customers, revenue and earnings per hour.

That made the idea feel less vague. Instead of simply trying to “make money”, I had specific numbers to track.

Where did ChatGPT tell me to find customers?

I then asked ChatGPT to go beyond generic advice and identify specific platforms, communities and tools I could use during the first two weeks to find potential customers. I also asked it to flag anything it was uncertain about rather than presenting outdated resources as fact.

ChatGPT recommended LinkedIn as the primary customer-acquisition channel. It also suggested freelance platforms for writing opportunities.

For basic infrastructure, it suggested Google Sheets for tracking prospects, Google Drive for the portfolio and delivery, and Canva Free for presentation.

But its most important advice was not to sit around waiting for job listings.

“Find founders/businesses with a content gap and approach them directly,” ChatGPT said.

How much did AI tell me to charge?

Finally, I asked ChatGPT to suggest how much I should charge for the service. I wanted it to give me a specific price or range, explain how it arrived at that figure and tell me how I could check whether the amount was realistic for the Indian market.

ChatGPT recommended Rs 3,500 for a 1,200–1,500-word researched article as a starting price.

The proposed package included research, writing, an SEO-friendly headline, meta description, suggested URL slug, basic search-intent optimisation, sources where appropriate, one revision and a 3-4 days turnaround.

It suggested increasing the price to Rs 4,500– Rs 5,500 after two or three successful projects and testimonials, while also considering monthly packages rather than relying entirely on individual assignments.

ChatGPT also advised checking current rates on platforms such as Upwork and comparing Indian freelance benchmarks before settling on a price.

What worked and where AI fell short

The biggest advantage was that ChatGPT did not leave me with one vague idea. It gave me several options, explained how each could make money, identified obstacles and then helped me choose one based on my circumstances.

It also saved time. The initial planning that could have taken days was reduced to a series of structured questions and decisions.

But this is where the limitations became clear.

Some platforms and resources ChatGPT suggested needed to be checked manually to see whether they were active, relevant and worth using. I could not simply assume that every recommendation was correct.

The same applies to its income projections. A side hustle can have a sensible business model on paper and still fail to find customers. AI can estimate what might work, but it cannot create demand.

That was particularly obvious with ideas such as a faceless YouTube channel. Building an audience could take months to start getting meaningful views or even years to build a substantial audience. Nobody can start earning instantly, so I would not blame ChatGPT for that.

The most useful part of the experiment was therefore not the side-hustle idea itself. It was the structure.

ChatGPT helped me turn my skills, time and financial target into a plan, identify potential customers, set milestones and think about pricing. But I still had to verify its recommendations, judge whether the opportunity was real and, ultimately, convince a human being to pay me.

AI can help turn an idea into a plan. But it cannot create demand, find customers or guarantee income.

Disclaimer: This is a first-person experiment conducted for informational purposes. Financial Express does not encourage or discourage the use of AI tools for building a side hustle and is not responsible for any decisions, financial outcomes or losses arising from the use of AI-generated recommendations discussed in this article. The ideas, income estimates, pricing and other suggestions generated by AI are not guarantees of success and should be independently evaluated before being acted upon. Financial Express does not recommend or endorse moonlighting. Employees should check their employment contracts and organisation’s policies before undertaking any paid work or business activity outside their primary employment.