Budget questions often get vague answers because the AI doesn't understand your specific situation—your expenses, your income variability, your local costs. Learning to feed it the details it needs transforms it from a generic advisor into something that can actually speak to your numbers.
There's a frustrating moment many people hit with AI: they ask a question and get a generic, vague answer that doesn't help. The problem usually isn't the AI—it's the question. Prompt engineering is just a fancy term for "asking better questions so you get better answers."
Think of it like calling customer service. If you say "my bill is wrong," they'll ask clarifying questions. If you call and say "my internet bill jumped $30 this month even though I have the same plan as last month, and I need to understand what charges are new," you get a much faster, more useful answer. Same concept with AI.
Generic AI answers about budgeting won't work for your life. You're not a "standard household." You have specific childcare costs, maybe child support coming in or going out, inconsistent income if you do freelance work, school calendars that affect your expenses, and a hundred other variables. The more context you give AI, the more useful its answer becomes.
A vague prompt: "How should I budget?"
A better prompt: "I'm a single parent earning $4,500/month with one child. My fixed expenses are $2,800 (rent, utilities, insurance). Childcare is $600/month during school year, $900 during summer. I want to build an emergency fund but I'm currently saving nothing. What percentage of my remaining income should go to emergency fund vs. living cushion?"
The second prompt gives AI the context it needs to give you a real answer.
Be specific about your situation. Include numbers, timelines, and constraints. Instead of "I have debt," say "I have $8,000 in credit card debt at 19% APR and $15,000 in student loans at 4.5%, and I can put $300/month toward debt."
State your actual goal, not a vague one. "Help me with my budget" is vague. "I want to know if I can afford a $200/month tutoring program for my daughter without cutting food budget" is clear.
Ask for a step-by-step approach, not a general answer. Say "walk me through how to decide" instead of "should I do X."
Give constraints upfront. "I can't work more hours because of childcare limitations," or "I get paid weekly but expenses are monthly, so cash flow is complicated."
Good prompts lead to answers that are actually useful for your life. Instead of generic advice, you get prioritized steps. Instead of "consider an emergency fund," you get "with your income and expenses, a $2,000 emergency fund is reachable in 8 months if you redirect the $150/month you're currently spending on unused apps."
You also get fewer irrelevant suggestions. AI stops giving advice about expense categories that don't apply to you because you've already told it what your situation actually is.
People think prompt engineering is a special skill only tech people have. It's not. It's literally just being specific and giving context. You already know how to do this—you do it every day when explaining your situation to friends, family, or professionals.
Try this: Take one financial question you've been thinking about. Write it down the vague way (how you might ask it casually). Then rewrite it with: your actual income, your main expenses, what you've already tried, and your specific goal. Paste both versions into ChatGPT and compare the answers. You'll immediately see the difference specificity makes.
Peri can explain this concept, give practical examples, help you decide whether it applies to your situation, or recommend a journey if appropriate.
Explore related journeys or tell Peri what you're working through.