SOLVED

Forecast function doesn't return good data

Copper Contributor

Hello,

 

I've tried to use forecast from excell to make my work easier, but the result returned is too far from reality.  Have you experienced this?

8 Replies

@Paul8585 

Please provide some specific examples.

best response confirmed by Paul8585 (Copper Contributor)
Solution

@Paul8585 

 

More specifically, attach an Excel file with one or more examples.

 

Click the "browse" link near the bottom of the reply window.

 

If you do not see a "browse" link, upload the Excel file to a file-sharing website, and post the download URL in a reply here.  I like box.net/files; others like dropbox.com.  You might like onedrive.live.com because it uses the same login as this forum.  Do not use Google Sheets.

 

If the forum does not allow you to enter a URL (yet), spell it out as text.  For example, this discussion is at techcommunity dot microsoft dot com /t5/excel/forecast-function-doesn-t-return-good-data/td-p/3592388.

 

-----
There are many reason why FORECAST not return "good data".

 

First and foremost, because your actual data does not fit a linear treadline closely.

 

Second, because you are using a different "flavor" of the (new) family of FORECAST functions, not FORECAST per se (without qualifier).

 

Finally, because you are misusing the FORECAST function.  We cannot know one way or the other unless and until you show us your formula and data -- again, preferably in an Excel file that we can download.

@JoeUser2004 

Thank you Joe! I've attached the file. I need to forecast next 5 months.

Thank you!

@Paul8585 

Your data have a yearly pattern, so you should use FORECAST.ETS.

In C46:

 

=FORECAST.ETS(B46:B50,C3:C45,B3:B45)

 

See the attached version.

Thank you Hans!
Still I think it's not returning good values. If I look on the historical October is best month of the year(except 2021), at same level with august. After using FORECAST.ETS october 2022 it's under september 2022, that's why i think it's not returning good values, it's should be at least august 2022.

@Paul8585 

I don't know how FORECAST.ETS works, but my guess would be that the forecast is weighed in favor of the most recent data. so that follows the trend of 2021.

1 best response

Accepted Solutions
best response confirmed by Paul8585 (Copper Contributor)
Solution

@Paul8585 

 

More specifically, attach an Excel file with one or more examples.

 

Click the "browse" link near the bottom of the reply window.

 

If you do not see a "browse" link, upload the Excel file to a file-sharing website, and post the download URL in a reply here.  I like box.net/files; others like dropbox.com.  You might like onedrive.live.com because it uses the same login as this forum.  Do not use Google Sheets.

 

If the forum does not allow you to enter a URL (yet), spell it out as text.  For example, this discussion is at techcommunity dot microsoft dot com /t5/excel/forecast-function-doesn-t-return-good-data/td-p/3592388.

 

-----
There are many reason why FORECAST not return "good data".

 

First and foremost, because your actual data does not fit a linear treadline closely.

 

Second, because you are using a different "flavor" of the (new) family of FORECAST functions, not FORECAST per se (without qualifier).

 

Finally, because you are misusing the FORECAST function.  We cannot know one way or the other unless and until you show us your formula and data -- again, preferably in an Excel file that we can download.

View solution in original post