How to automate P&L forecasting with gini

Getting 100% clear on the financial future of your company is the number one priority of every business owner. But in today’s volatile market, P&L forecasts go quickly out-of-date.

Instead of once a month, you now have to update your P&L forecast every time something changes — which can be as often as daily. Who has time for that?

The consequence of not updating your P&L forecasts often enough, though, is dire — 82% of companies that fail do so because of a lack of understanding of their cash flow and working capital.

The fast-changing environment we now find ourselves in calls for a new approach to P&L forecasting.

Fast.

Powerful.

Accurate.

Introducing: Machine learning predictive models (wait for it) baked into a simple spreadsheet format. 

That’s gini.  

Why automate your P&L forecasts with machine learning?

• It’s fast. Like, lightning fast. 
  Machine learning models analyse all your variables in every possible combination in seconds.

• It’s more powerful.
  Machine learning models can handle a large quantity and variety of data, so you can even include external market data in your forecasts.

• It’s more accurate. 
  ‍Machine learning models are better at identifying patterns in data than you are. And there’s less room for human error.

If you’re panicking at the mention of complex machine learning tech, don’t worry. 

We’ve designed gini for people with zero coding experience. Our machine learning models are built into a spreadsheet format, so you can run data predictions in a few clicks. No formulas or code needed.  

‍Read: How to level up your business forecasting with machine learning

How to run a P&L forecast with gini

This P&L dataset from a global hotel chain includes income and expenses from January 2018 to October 2020. Let’s see how easy it is to run a 12-month forecast on net profit.

Go to add-ons and choose “gini”.

On the gini sidebar, select:

• Target variable: Net profit
• Forecast period: 12 months
• Additional variables: Select all, except Total revenue and Total expenses. 
  (This will provide a more granular view of how individual variables are impacting profit.)

Click “Forecast”.

Once the machine learning models have analysed the P&L data, a new tab will appear with the forecasted values for all P&L variables populated from November 2020 to October 2021.

The forecasted values for net profit are highlighted in orange.

Below the table, the values are plotted in a graph with a 95% confidence interval between the dotted lines. 

It predicts a peak in profit over the December holiday season, and a dip around July and August 2021 — a seasonality trend picked up in the historical P&L data.

How to see what impacts your net profit most

On the sidebar, you can see a breakdown of the most impactful variables for your P&L forecast. 

gini’s machine learning model calculates this by analysing each variable’s importance based on the predictive power it contributes, while also accounting for complex interaction patterns.

In this case, income from Deluxe rooms and Suite rooms impact net profit most.

‍

How to run a scenario to see if/when you’ll need a loan

Let’s say the worst case happens, and the area goes into full lockdown for three full months from January 2021 to March 2021.

We don’t expect anyone to be paying for deluxe rooms, attending events, or eating in the restaurant, but we do expect minimal income from takeaway food and perhaps a few suite rooms. We could lower our marketing costs and lay off a few employees for this period.

Let’s go ahead and change our forecasted income values directly in the cells to see how net profit is impacted. Edited cells are highlighted in blue and appear on the sidebar automatically. 

(You can see the original value by hovering your cursor over the cell, and change it back by clicking “X” on the sidebar.)

Click “Run Scenario”.

After the machine learning models have run the analysis, an updated chart appears, showing the scenario-based forecast line for net profit in green, with the original baseline forecast line in grey to compare against.

Looks like, as long as things pick up straight away and you have sufficient reserves of working capital, you may not need a loan.

Running quick scenarios on your data like this is very helpful when it comes to preparing for the worst and best cases. Remember, growing at an unexpectedly high rate can be just as costly and also requires sufficient reserves of working capital.

With gini, you can run scenarios faster than doing it manually in a spreadsheet, and you get a more accurate outcome. The machine learning model doesn’t look at any one data point in isolation, it takes the whole picture into account, giving you a better understanding of how variables interact with one another over time.

If you have any questions about gini, please email us at predict.help@gini.co or schedule a one-on-one demo below.