Is Anyone Using AI (or Anything Smarter Than Gut Instinct) to Predict Sales Trends?

[Donnie-M]

I’ve been thinking a lot about how to get ahead of the curve—whether it’s knowing what inventory to restock, what specials to run, or even which team member to schedule more hours for a specific day.

I’ve started experimenting with combining sales data, day-of-week trends, and even weather patterns to make decisions. AI (my helpful Overlord) is part of the mix… and sometimes it’s spooky accurate. Other times, not so much.

So I’m curious—

Are you doing anything fancy to forecast what’s coming?

Anyone using spreadsheets, tools, Square reports, or third-party apps to help plan ahead?

Or do you still roll with your instincts and experience? (No shame, that’s real power too.)

Let’s crowdsource some insight—human, digital, or otherwise.

And if your forecasting model “turned you into a newt”… I trust you got better. 🦎

#SmarterSelling #ForecastOrFreestyle #AIOverlordAdvises #HatOrBust #TurnedMeIntoANewt

[TheRealChipA]

Hey there @Donnie-M.  So I have a budget forecasting and cash flow tool that I use every day, and have for years.  At the beginning of the year, I import my P&L and Balance Sheet from the previous year, as well as a number of Square CSVs.  Then I fill in various assumptions about how this year is expected to be different from the past year and it spits out a proforma budget that I can adjust as needed.

On a daily basis, I import revenue account journal entries and compare those to expected revenue in 7-, 14-, 21- and 28- day increments.  My cash flow forecast table has access to my bills and expense journals so that I can quickly spot any possible cash flow issues and correct them before they become problems.  I even download the Square Trends CSV and use it to for various “what ifs.”

After 10+ years in business, I don’t really need to do labor budgeting.  I have three ice cream season types — high season, a transition season before and after high season, and winter.  For each of those, my schedule is set and hasn’t really changed much at all in recent years.  But I do revenue and cash flow analysis religiously every day to help me spot potential trouble before it bites me.

So, I guess I and my spreadsheet are the AI, in this case.

At one time a few years ago, I gave an app called Profit Rover a go.  If you are at all interested, I’d recommend checking them out with their 14-day free trial.  It uses AI to analyze a lot of things on your Square history — sales, for sure, but also labor and scheduling.  It even makes product and service pricing recommendations.  I was impressed, but since I have my spreadsheet and know that it works it wasn’t worth the monthly fee to me.  If I were starting a new business, I’d probably use it for that.

[TheRealChipA]

@Donnie-M I also have another menu costing and inventory tracking spreadsheet that uses the Item Details CSV to show me theoretical inventory and alert me to supplies I need to check/order.  That one took a while to create, but it is highly accurate and is my other “daily use” tool.

[DinaLRosenberg]

Wow- this is all impressive. I have it rolling in my head- no surprise but I'm a gut instinct type of girl LOL

[JTPets]

With AI I built a script for google sheets that downloads my catalog and another that downloads 91 days of sales unit history (1/4 year or 13 weeks) automatically every few hours from Square without my intervention

I then use that information to create an order based on 91 days of history an order for X (I use 45 for most of my items, or 21 for freezer items due to space) number of days of stock wanted.

This allows me to predict what I need to get through a 1 week delivery cycle with enough inventory to account for a higher customer base. While it isnt perfect I find for my long shelf life retail items it is almost perfect. 

I then combine that data with a manual catalog export to line up with a default vendor and BAM I have a drop down order generator. I really wish Square would clone it! @nika *nudge nudge

All of this is automatically assembled (except for the vendor portion) and updates every few hours or manually with a button click

[BrianaJo]

My husband (co-owner & roaster) and I both use our gut instinct. We both have years of experience in different ways where we complement each other very well. We'll review data from the Square reports and then depending on what it is we're doing either him or I will know how to address it. 

[Twpchair]

I use Square myself and Excel. My store booker uses Google Docs and Square (now) by force 😆. My accountant is old-school and on the 10th of the month, he needs our monthly reports. Our state taxes are due to be sent on the 15th of the month. He does Exel only 70 years old and won't fix what isn't broken type thing. Since around January, I have been sending both Excel sheets and Square. Each month he will tell me about this Square thing that is pretty accurate and detailed. I think 🤔 he's coming around 😆. I used to use multiple paid apps and since Square has become more all-in-one and done. I have dropped two already and saved over 40.00 a month. Myself only I don't use Google Docs. Just our store bookkeeper does. I use AI with Square for just about everything from inventory to sales prior to the days, vs week best days, card usage vs cash flow, and which days. You name it.

[elioth]

Back when I was helping my parents business we were implementing a lot of ML models (my background is in the quantitative sciences) to do demand forecasting (different locations had different demand patterns and the weather seemed to influence a lot which products got sold), we also implemented a stocking guide using the same forecasting model.

The two most difficult problems were pricing (we just looked at "nearby competitors" prices and try to be in the middle of the pack) and staff scheduling (this is one of those problems that looks easy on paper, but actually eats a lot of the margins when you run a food business). We came up with a couple of algorithms to do this based on the expected demand and adjust seasonally. For example, for pricing taking goods that tended to sell well together, making one relatively cheaper and the other one more expensive, so that in the end the margin was higher. For staff scheduling it was the same, trying to schedule people according to the demand is easier said than done, and in the end only an algorithm can do this efficiently.

We did this fetching data from the Square POS for app integrations (which is a headache in itself, as the data needs a lot of cleaning), but we came up with a workflow that made sense for us (my parents are not tech savvy, so reporting everything in a way that was useful for them without consumed a lot of time, but in the end it was worth it, as they don't have to be crunching numbers in excel all day). 

As a matter of fact, I'm now trying to package some of these algorithms into an app integration, it is still a work in progress, but feel free to DM me if you'd like to try it.