Spiffy MCP gives your AI secure access to your Spiffy account, including customers, orders, subscriptions, payments, products, affiliates, webhooks, and the same reporting and analytics available in your Spiffy dashboard.
Instead of building reports or analyzing raw data yourself, you can simply ask things like:
- Why was revenue higher this week than last week?
- Which products have the highest lifetime value?
- Which marketing channels generate the most revenue?
To use the prompts in this article, you’ll need to connect your AI app to Spiffy.
The prompts below are grouped by the job you’re trying to do. Start with a broad question if you’re exploring, or jump to the specific prompts when you know exactly what you want to find.
Use your AI with Spiffy alone, or connect it to your other tools for deeper analysis into why the numbers moved, not just what they did.
This is not meant to be an exhaustive list. It’s intended to give you some ideas into what’s possible. You can use any prompt as written or customize it to fit your business and the tools you have connected.
- Not Sure What to Ask?
- Know How the Business is Doing
- Find Out Where Growth Comes From
- Protect Revenue You Already Have
- Run the Deep Numbers
- Look Something Up
- Take Action
- Go Beyond Spiffy
- Writing Prompts That Work
Not Sure What to Ask?
You don’t need to know which report to run or which metric to look at. Tell your AI what you’re trying to understand, and let it figure out where to look.
What are the most useful questions I can ask you about my Spiffy data?
Give me a business health check for the last 90 days and tell me what I should pay attention to.
Look for anything unusual in my Spiffy data over the last 30 days.
Find the biggest opportunities to improve revenue right now, based on the last 90 days.
If you were my business analyst, what would you investigate first? Use the last quarter.
Find three things in my Spiffy data from the last month that I should know about.
Pro Tip: Giving the AI a time frame, like “the last 30 days” or “last quarter”, makes your prompt much sharper. Without one it has to guess what you’re looking for.
Know How the Business is Doing
Start with the numbers that tell you how the business is performing. Use these prompts to spot trends, compare periods, and understand what’s moving your revenue, customers, and subscriptions.
Check headline numbers:
Show revenue, orders and average order value for last month, compared with the month before.
Watch trends:
Show revenue and orders by week for the last 90 days. Point out any week that broke the pattern.
Track recurring revenue:
Show MRR, new subscriptions and churned subscriptions by month for the last six months. Tell me whether growth is coming from new subscriptions or from keeping existing ones.
See what refunds and failures are costing:
Show refunds, refund count and failed payments by month for the last six months, next to revenue for the same months.
Compare any two periods:
Compare revenue, orders, customers and average order value for this quarter against last quarter, and tell me which of those moved the most.
Find Out Where Growth Comes From
Revenue can come from very different sources, and not all of them are equally valuable. Use these prompts to see which channels, campaigns, and referring sites are actually driving customers and revenue.
Find the channels that actually make money:
Break down revenue, orders and average order value by channel for the last 30 days. Rank by revenue and tell me which channel has the best average order value.
Find your best campaigns:
Break down revenue and orders by UTM campaign for the last 90 days. Show the top ten and flag any campaign with high orders but low average order value.
Find where checkouts leak:
Show conversion rate and orders by channel for the last 30 days. Which channels bring people who don’t convert?
Compare first touch with last touch:
Break down revenue by channel for the last 90 days using last-touch attribution, then again using first-touch, and tell me which channels look better at discovering customers versus closing them.
Check customer quality by source:
Break down revenue per customer and customers by UTM source for the last 90 days. Which sources bring the highest-value customers, not just the most?
Protect Revenue You Already Have
Growth isn’t just about finding new customers. It’s also about keeping the revenue you already have. Use these prompts to spot churn, failed payments, refunds, and other signs that revenue may be at risk.
Spot churn trends:
Show new subscriptions and churned subscriptions by month for the last twelve months. Is churn getting worse, and when did it turn?
Find failing payments:
List customers with a failed payment in the last 30 days, what they were paying for, and how much is at stake.
Find at-risk payment plans:
List active payment plans with a failed or overdue payment, sorted by the amount still outstanding.
Find win-back candidates:
Find customers whose subscriptions were cancelled in the last 90 days, with what they were paying and how long they’d been subscribed. Rank them by what they were worth.
Run the Deep Numbers
When you need to go beyond the headline numbers, Spiffy’s reports can do the heavy lifting. Use these prompts to dig into product performance, customer value, affiliates, checkouts, and payment plans.
See which products carry the business:
Run a product sales report for the last 90 days and summarize which products drive the most revenue and which sell the most units.
Check whether customers are worth more over time:
Run a lifetime value report for the last twelve months and tell me whether newer cohorts are worth more or less than older ones.
Grade your affiliates:
Run an affiliate performance report for the last quarter and show me the top ten affiliates by revenue, with their commissions.
Check payment plan collection:
Run a payment plan performance report for the last six months and tell me how much is outstanding and how much is being collected on schedule.
Re-run something you already set up:
Run my [report name] report again for this month and tell me what changed versus last month’s run.
Look Something Up
Sometimes you don’t need a report, you just need an answer. Use these prompts to quickly find a customer, order, subscription, promo, affiliate, webhook, or other specific piece of information.
Investigate a customer:
Find the customer with email jane@example.com and summarize their orders, subscriptions and any refunds.
Check a specific order:
Show me order 10432. What was bought, what was paid, and current status.
Check a promo code:
Show me the SPRING20 promo code and the orders that used it, with total revenue and discount given.
Check an affiliate’s commission:
Show me the affiliate [name], their recent attributed orders and their unpaid commission balance.
Ask a product question mid-task:
How does Spiffy handle failed subscription payments and dunning?
Take Action
Your AI can do more than find and analyze information, it can make changes in Spiffy too. Use these prompts to handle routine tasks.
Actions are limited to your Spiffy user role permissions. Review each confirmation carefully before approving a change.
Refund order 10432 in full. The customer was double charged.
Retry the failed payment on [customer name]’s subscription.
Cancel the subscription for jane@example.com at the end of the current billing period.
Create a 20% off promo code called SPRING20 that expires at the end of next month.
Add a note to this customer’s record summarizing what we just found.
Go Beyond Spiffy
Spiffy can answer many of your business questions on its own, but your AI gets even more useful when it can connect the dots across your other tools. Use these prompts to combine Spiffy’s commercial data with context from your CRM, support, email, analytics, and other systems.
Use this prompt formula: get the breakdown from Spiffy → get the context from another tool → compare both → explain the pattern.
Find out if your ads are actually profitable:
Break down revenue and orders by UTM campaign for the last 30 days, then pull spend for the same campaigns from my ad platform (Meta, Google Ads). Show me which campaigns made money after ad spend and which lost it.
Explain a revenue change:
Break down revenue by channel for last week versus the week before, then use site analytics to compare sessions and conversion rate for the same channels. Tell me whether the change came from traffic or from conversion.
Find traffic that doesn’t pay:
Show conversion rate and revenue by channel for the last 30 days, then compare with traffic by channel from my analytics. Which channels get attention but don’t convert?
Explain churn in the customers’ own words:
Show churned subscriptions by month for the last six months, then pull support tickets and CRM notes for the customers who cancelled most recently and summarize the reasons they gave.
Prioritize support by money at stake:
List customers with failed payments or active subscriptions, match them to open tickets in my support tool, and rank by revenue at risk.
Judge which email campaigns produced revenue:
Break down revenue by UTM campaign for the last 60 days, then compare with sends, opens and clicks from my email tool. Which campaigns drove revenue rather than just engagement?
Writing Prompts That Work
A good prompt doesn’t need to be complicated. A few simple habits can help your AI give you more useful answers and get to the information you need faster.
Ask for the specific answer, not the raw data. “Break down revenue by channel for last month” beats “list orders from last month”. You get a meaningful analysis instead of just a list of numbers.
Say the period plainly. “Last month”, “the last 90 days”, “Q2” all work. Be specific for a clearer result.
One question at a time. Stacking four questions into one prompt gets you a shallow answer to each.
Name the thing. Use the actual checkout name, promo code, product or email address rather than “my main checkout”.
Let it look things up first. If your AI seems unsure what it can measure, ask it to check.