FAQs

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Suspendisse varius enim in eros elementum tristique.

Getting started
Using giniPredict
Forecasting
Scenarios
Features
Error messages
Security
Pricing
Getting started

How to get started (Video)

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Suspendisse varius enim in eros elementum tristique. Duis cursus, mi quis viverra ornare, eros dolor interdum nulla, ut commodo diam libero vitae erat. Aenean faucibus nibh et justo cursus id rutrum lorem imperdiet. Nunc ut sem vitae risus tristique posuere.

How do I import my data from Xero or other accounting software?

For Xero
giniPredict lets you import data from Xero seamlessly.

Open a new Google Sheet.

Click on Add-ons -> giniPredict -> Import from Xero.

Follow the instructions to import a report from Xero. We recommend importing at least 24 months of data.

Click “Import” and wait for a new sheet to appear with your Xero data.

For other accounting software

Export the data you want to use from your accounting software to a Google Sheet or CSV file. 

Open a Google Sheet and import the data (File -> Import).

In a new tab under Add-ons, open giniPredict’s template and copy-paste the imported data into the correct format on the template.
Note: You may need to delete any non-functional header rows.

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Suspendisse varius enim in eros elementum tristique. Duis cursus, mi quis viverra ornare, eros dolor interdum nulla, ut commodo diam libero vitae erat. Aenean faucibus nibh et justo cursus id rutrum lorem imperdiet. Nunc ut sem vitae risus tristique posuere.

How can I input my data from Netsuite?

Export the data you want to use from Netsuite to a CSV file (top left menu). 

Open a Google Sheet and import the data from the CSV file (File -> Import).

In a new tab under Add-ons, open giniPredict’s template and copy-paste the imported data into the correct format on the template.
Note: You may need to delete any non-functional header rows.

How can I input my data from Google Analytics?

In Google Analytics, find the data or report you want to use, and set it to “monthly” data.

Export it to a CSV file (top right menu). 

Open a Google Sheet and import the data from the CSV file (File -> Import).

In a new tab under Add-ons, open giniPredict’s template and copy-paste the imported data into the correct format on the template.
Note: You may need to transpose the data first.

How should my data be formatted?

Currently, our tool is designed for monthly time series data, formatted with the months in columns along the top and the variables (expenses, revenue, etc) in rows. 

Each variable can be labelled with the data type (e.g. Expenses) and data subtype (e.g. Maintenance costs). You can use different types of data formats (numerical, percentages, currency, etc).

It’s okay to have empty rows, but only if they are completely empty (not labelled). It’s best to delete header rows.

You can use giniPredict’s template (under Add-ons) to get set up quickly.

Can I have formulas in my data?
(e.g. Total marketing spend as a sum of all marketing channels)

Yes you can have formulas in your Google Sheet. giniPredict’s machine learning model will look at the value of the cell and will function as expected.

What if I have missing data?

If there are missing values in your dataset, giniPredict’s machine learning model will do its best to fill in the gaps. It needs an absolute minimum of 3 values per variable (row) to do that. Still, missing data is likely to reduce the accuracy of the forecast. 

As a general principle, the more data you input, the better, as the model has more to learn from and can produce more accurate results.

Can I run forecasts on weekly or daily data as well as monthly data?

Currently, giniPredict only supports monthly data. But we hear you — weekly and daily forecasts are in the development pipeline.

Can I use giniPredict with Microsoft Excel?

Currently, giniPredict is an add-on for Google Sheets only.
We plan to make it available for Microsoft Excel in future.

For now, you can open your Excel spreadsheet in Google Sheets to use giniPredict.

Using giniPredict

Do I need an internet connection to run forecasts with giniPredict?

As a Google Sheet add-on, giniPredict needs an internet connection to run. It won’t work in offline mode.

What kind of data can I use giniPredict with?

You can run forecasts on any quantitative monthly time series data with giniPredict, including financial data (revenue, expenses, etc), marketing and sales (web traffic, ad performance, user growth), product demand and resource planning, and much more.

You can even plug in any external data you have on market trends, employment stats, weather, commodity prices and so on, to understand how it impacts your company performance.

What is the minimum amount of data I need in order to get a reasonable forecast?

giniPredict works best with a minimum of 36 data points (months). 
The more historical data you put in, the more the model has to learn from, the more accurate the forecast.

How does giniPredict work?

giniPredict’s customised machine learning model predicts future trends based on
historical time series data using state-of-the-art modelling techniques. It examines the similarities and correlations between each variable in every possible combination to model linear and non-linear complex relationships while taking different levels of seasonality into account for a flexible forecast.
The model is also robust to outliers, shifting trends and missing data (to a certain extent). 

It generates a baseline prediction and provides a 95% confidence interval for guidance. It also assigns an importance score to each variable based on the predictive power it contributes, so you can see which variables have the most impact on the target variable.

When running forecasts, bear these principles in mind:

The further into the future you forecast, the less accurate it will be.

The more data you give it, the more the model learns, the more accurate the forecast.

The higher the quality of your data, the more accurate the forecast.

Can I make any adjustments to the machine learning model?

No, it’s not possible to make adjustments to giniPredict’s customised machine learning model. Our model works well with any historical time series data, and allows you to run multiple types of forecasts, from cash flow to sales, to product demand to marketing ROI, for any industry and field, without needing to code.

What should I do about anomalies in my historical data?

giniPredict’s machine learning model automatically identifies and ignores any one-off anomalies and outliers in the data, excluding them from the forecast to get a more realistic result.

How do fixed variables (e.g. fixed costs) impact the forecast? 

giniPredict’s machine learning model is designed to examine the relative changes for each variable and how they interact with each other. Fixed variables (e.g. rent) are not informative to this evaluation and are not likely to impact the forecasted variable (e.g. sales).

It may still be worth including fixed variables in your dataset if you want to run scenarios to see the impact on cash flow if your rent changes in future, for example. 

What types of data can I forecast with giniPredict?

You can use giniPredict to run forecasts on any quantitative monthly time series data, from cash flow and profit and loss statements to web traffic, foot traffic, product demand, user analytics, advertising performance and much more. 

Forecasting

How do I run a forecast?

Once you’ve imported your monthly time series data into a Google Sheet and formatted it according to our template, go to Add-ons and choose giniPredict. 

On the sidebar, select the variable and time period for your forecast, plus the additional variables to factor in. You must include at least one additional variable — the more the better. You can exclude “total” variables (e.g. Total revenue) to get a more granular view of how individual variables impact each other.

Click “Run forecast”.

A new tab will appear, with future values populated in a table (the forecasted variable in orange) and a chart underneath. The chart shows the historical trend in grey and the forecasted trend in orange, with upper and lower bounds indicating a 95% confidence interval.
Hover your cursor over the chart to see the specific values of each month.

A new sidebar shows you which variables have the most impact on the forecasted variable and gives you the option of running a scenario.

Do I need to select my dataset range every time I want to forecast?

You don’t need to select a data range when you want to forecast. giniPredict automatically factors in all the data in the worksheet you’re viewing when you open the add-on.

How far ahead can I forecast?

You can forecast as far ahead as you like, but bear in mind that the further ahead you forecast, the less accurate the outcome. For giniPredict, we recommend forecasting 6 to 12 months ahead maximum, to maintain a confidence interval of 95%.

Why do I need to select additional variables when running a forecast?

giniPredict’s machine learning model analyses the relationships between the forecasted variable (e.g. sales) and other related variables (e.g. advertising expenses) to provide you with a realistic forecast. At least one additional variable is required to run forecasts and scenarios.

What do the upper and lower dotted lines on the forecast chart mean?

The dotted lines above and below your baseline forecast indicate a 95% confidence interval.

How does giniPredict know which is the most impactful variable?

giniPredict’s machine learning model analyses each variable’s importance based on the predictive power it contributes, while also accounting for complex feature interaction. 

How do I compare the forecast to actual data? 

giniPredict’s machine learning model automatically backtests the data as you import more data going forward. As you keep importing new data every month, the model learns from it and adapts to increase the accuracy of the forecast. 

You can check previous forecasts against actual data to compare it manually in a separate tab, if you like.

Scenarios

How do I run a scenario?

Once you’ve run a forecast, you’ll be given the option to run a scenario from the same forecast output tab. 

To run a scenario (e.g. increase marketing spend for 3 months to see how it would impact sales):

Change the predicted values of the related variable in question by typing the new values directly into the cells (e.g. add $2000 to marketing spend for May, June, July 2021). 

The altered cells will be automatically highlighted in blue in the sheet and displayed in the sidebar.

Click “Run scenario”.

giniPredict will update your target variable forecast values and generate new rows showing the percentage change from your base forecast (“with base”) and previous scenario (“with last”).

The chart shows Scenario 1 (in blue) with the original baseline forecast (in orange) to compare against. You can toggle between the two using the dropdown cell above the chart. 

You can continue to tweak values and run successive scenarios to find the optimal outcome.

How do I compare different scenarios?

When you run a scenario, giniPredict will update your target variable forecast values and generate new rows showing the percentage change from your base forecast (“with base”) and previous scenario (“with last”).

The chart shows Scenario 1 (in blue) with the original baseline forecast (in orange) to compare against. You can toggle between the two using the dropdown cell above the chart. 

If you run a second scenario, the table and chart will be updated with the new set of values in a different colour. To compare the different outcomes visually, you can toggle between scenarios using the dropdown cell above the chart.  

Features

How do I recommend a new feature I’d like giniPredict to have?

Please share any feedback, recommendations and ideas you have with us at predict.help@gini.co.

How is giniPredict different from Scenario Manager in Excel?

Scenario Manager in Excel allows you test scenarios with your data at a single point in time.

giniPredict shows you a scenario-based forecast for 6 to 12 months into the future, factoring in historical trends and correlations in your data. It also takes multiple variables into account to give you a more realistic picture.

How is giniPredict different from Goal Seek in Excel?

With Goal Seek in Excel, you work backwards from a target using a straightforward formula. 

With giniPredict, you can tweak variables and run scenarios to get a better understanding of how each variable relates to and impacts your target over time.

Error messages

Network Error

Something’s wrong with your internet. Please refresh your browser and try again.

Google Error

Our server had a problem when talking to Google's servers.
This may be because you’ve recently installed and enabled the add-on, and Google hasn’t recognised that you’ve enabled it yet.
Please refresh your browser tab and try again in a minute.

Server Error 500: <...error message...>

Something went wrong on our server. Please contact predict.help@gini.co and provide us with the error message.

Server Error 503:

Our server was temporarily unavailable for the action you requested. Please try again in a minute.

Security

How is my data kept private?

Your data is always encrypted and never stored by the giniPredict model or gini the company. While we keep track of the user requests made to the model for analytics purposes, the data is completely anonymous, ensuring it is never possible to match a user to a dataset. 

For information on your personal data, please see our Privacy Policy.

Pricing

How much does it cost?

As a beta product, giniPredict is completely free.
We will inform you of our launch date and pricing plans soon.