No problems
@Guillermo Cuadrado. Yep linear approach based on previous. Its quite difficult (but not unachievable) to forecast based on whats coming ; if you can define a cookie cutter approach to things like migration then a formula per se would be achievable ; in that lets say with your migration you had a 'migration size' or a way of migrating on a (per app) basis ; where you could try and boxify an app migration into a tiered approach (think small migration, medium migration etc) then you could say ok in the next 1/4 we are going to do 4 app migrations of 2 medium and 2 small at an expected cost of $x. Of course, no two applications are the same and there are always different challenges when it comes to app migrations where you find servers that no one knows about that apparently that application uses :)
For projects if you had a direct tie in to your project management office, theoretically based on the project budget and its estimates during inception phase be able to forecast those numbers and see where you land against actuals.
Again, not easy things to do, but, achievable given the time and the right people :)
Original Message:
Sent: 01-12-2021 02:53
From: Guillermo Cuadrado
Subject: Cloud expense forecast tools in excel ???
Thanks, @Greg Winfield. We've done something similar for Azure subscriptions, but loading the data onto Cost Transparency (no Cloudability yet).
Our central Cloud organization collects data from DevOps teams and puts together a forecast based on it. Still at an early infancy stage, with lots of teething problems.
As you mention, I believe that what @Justin Sheldon was after is an algorithm. You propose a linear approach, which is as good as any. The only scenario it won't work is if the organization would face a large migration project-like we are. Then, knowing the past would be pretty much meaningless. For established Cloud applications, a linear approximation would work well. Actually, our current Cloud spend has been pretty stable, thus averages helped us too.
------------------------------
Regards, Guillermo
------------------------------
Original Message:
Sent: 01-11-2021 20:47
From: Greg Winfield
Subject: Cloud expense forecast tools in excel ???
Justin,
The forecasting feature in Apptio Cloudability provides some of this capability in the tool so have a look in there.
In Excel if you did a list of cost by service on lets say a per month basis, over a 12 month period and then ran an average against that you could project your costs based in previous types of spending in Excel. This data could be easily exported out of Cloudability and then this done.
eg
| Jan | Feb | Mar | est Apr |
| Service | | | | |
| AWS EC2 | 10000 | 10500 | 8000 | 9500 |
| AWS S3 | 2000 | 3000 | 9000 | 4667 |
| AWS RDS | 20000 | 20500 | 30500 | 23667 |
From this if you could work out the percentage of change then you could apply that down to a granular level across your Cloud fleet to give you forecasts based on previous actuals?
Hope this helps.
Original Message:
Sent: 12-28-2020 10:06
From: Justin Sheldon
Subject: Cloud expense forecast tools in excel ???
thanks for the response. I'm looking to start simple and create a template where I can input the basics (i.e. product, size, usage, cost). We're primarily an AWS shop and are also dealing with poor quality tagging of resources as well. When I work with our EA team and ask about how to forecast future months, I get some "deer in headlights" looks :).
Original Message:
Sent: 12-21-2020 03:35
From: Guillermo Cuadrado
Subject: Cloud expense forecast tools in excel ???
We are still struggling with this, but our Cloud management team does a round with all registered DevOps teams and tries to get an estimation of the consumption for the coming months. For now data is incomplete, and they present a forecast. We have this in Cost Transparency.
I guess what you're after is some kind of heuristic that will tell you how the consumption will likely evolve. We are in the verge of a large Cloud migration program, so I guess I'll know more in the coming months, @Justin Sheldon.