Question

Best Practices on how to make a content forecast

  • 11 January 2023
  • 2 replies
  • 33 views

Userlevel 5
Badge +3

Hi there,

My current task is to create a Content Forecast for the two coming years (2023-2024) so we can have a rough estimate of how many user questions, discussion, and comments we can expect (inside and outside of ideation) in order to have an estimate of how much content will need to be moderated

How would you approach this?

I came up with an idea but maybe somebody has got a better technique :)

First off: I already have a member forecast & a session forecast for the next to years. This will be an important factor for the forecast calculations.



Basically I have a list of all numbers for the past year with which I calculate a ratio
(simplified version here)

Month Questions Sessions Members Ratio:
Questions 
p. Session
Ratio:
Questions p.Member
January 100 10000 5000 0.01 0.02
February          
etc.          

 

Then I calculate a monthly ratio, e.g. (topics p. session) and (topics p. member) of each of the above content types.
→  e.g. for January 0.01 questions p. session and 0.02 questions p. member


Then I multiply those ratios by the forecasted # of members or sessions and then use an average of that number.

So for example, the formula for the Forecasted # of Questions for July 2023 (click on “learn more”)
 


Formula: =AVERAGE(A*B,C*D)

Legend:

  • A: Forecast of Sessions in July 2023
  • B: Average Ratio of Questions / Session as calculated above
  • C: Forecast of Members in July 2023
  • B: Average Ratio of Questions / Members as calculated above


    Optional: I am considering to add a “maturity factor” or “growth factor” as the variable “E” in the formula below 
    ​​​​​​Formula: =AVERAGE(A*B,C*D)*E

    So I can use that to make a more accurate model based on how much I think that the content will grow proportionally to the community growth. The factor will probably be lower than 1 (e.g. 0.8) as the community reaches maturity and content saturation we cannot expect the ratio to stay constant in the coming years)


I do that for each content type that I want to forecast, by just copy-pasting the formula.
 

Any other ideas? 


2 replies

Userlevel 7
Badge +2

Interesting. Thanks for sharing. 

I haven’t done this, so take my advice with a grain of salt…

But I’d expect that your content ratio is likely slow down over time. At first, it will increase with increases in community adoption (more users = more content), but at some point you will develop a knowledge base that covers most of the “low hanging fruit.”

Simpler and easier questions will be asked and answered, and the recommendation engine and other features will route users to those questions. You’ll likely also see a change in the type of questions to more niche and/or more complex and/or more in the weeds questions. That’s what we’ve seen in our community 3+ years old. After we hit 2,000+ questions (and moved to Insided), we saw a dramatic shift in the types of question asked and now question volume is slowing a bit (despite adoption and engagement increasing).

So I guess my advice is to use your ratio for years 1 and 2, but then to forecast a smaller ratio (even if users are still increasing). 

 

Userlevel 5
Badge +3

Hi @ddpancratz thank you so much for sharing your knowledge here - helps a lot!

 

But I’d expect that your content ratio is likely slow down over time. 

Yup, I think so too, so in the formula above (in the spoiler) I added the variable “E” which will pull the number down a bit.

Have a great weekend!

Daniele

Reply