• Hi All

    Please note that at the Chandoo.org Forums there is Zero Tolerance to Spam

    Post Spam and you Will Be Deleted as a User

    Hui...

  • When starting a new post, to receive a quicker and more targeted answer, Please include a sample file in the initial post.

Duplicating sales trend for next year using Excel

PP3321

Active Member
I have a data of daily sales in September 2015.

My requirement is to copy the trend for next year, i.e. September 2016
*Please find the attached.

Problems is that weekends change in 2016.
So I can not just copy & past from 2015 data.

What would analysis experts do in this case?

Should I instead:

-get weekly trend?
-get trend by day, rather than weekday name?
-get moving average?

Thanking you in advance...
 

Attachments

  • Forecast Problem.xlsx
    15.1 KB · Views: 4
Last edited:
The requirement is to automate the calculation on Excel.

So maybe I can do Index&Match using week number.

For example...first Monday of month, 2nd Monday of the month...etc
 
I tried vlookup. It works fine, but towards the end of month, if the week 5 does not exist in next year, it would not have any value...
 

Attachments

  • Forecast Problem 2 with vlookup.xlsx
    16.4 KB · Views: 5
Well, since you are doing forecasting based on week and day of week. You really need to have data for matching set.

Ex. For each month you need to find first Sunday (or Monday) of the month. If the Sunday does not fall on 1st of the month then, you extend range by x number of days backward. And do same for end of month (for Saturday or Sunday), but forward.

FYI - you have formula error in J2. You excluded J3 in range, throwing off % Daily Contribution calc.

Another approach. Is to calculate avg contribution of each day of week for a given month/segment. Then use that for forecasting.

Note: For an accurate forecast you'd need to build a model that suites your business. See link for what's typically needed.
http://chandoo.org/forum/threads/projecting-trends.28776/#post-172271
 
For simple time series forecast for data exhibiting seasonal trend. Do a search on "Seasonal naïve method" & "Drift method". I think the methods matches what you are looking to accomplish here.
 
Your basic data with and without weekends
upload_2016-9-8_22-11-18.png

Looking at the above there appears to me no use trying to use the monthly data for any prediction.

To look at the Daily data you will need to chart all the monday's together and see if there is a trend, then the same for Tues, Wed etc
 
Back
Top