This work is licensed under CC BY 4.0 - Read how use or adaptation requires attribution

New Account Estimation at Pearson

Contributed by JohnAllen Crites of Pearson.


Project Title Estimation
Internship Title FinOps Analysts Intern
Education Status Full-time – Finance Major
Department Joined Infrastructure & Operations
FinOps Capabilities Forecasting & Budget Management
Period of Internship 12-weeks
Lifecycle Inform


Background The company has 5 separate business units that request AWS accounts from a centralized Cloud Management team.

Product owners are required to submit an estimate & secure budget before an AWS account can be provisioned. 

Product owners can consult with engineering leads, FinOps Analyst, or self produce and estimate. 

Estimates can be for new Cloud first workloads or migrations from data centers


Problem Statement The centralized FinOps team believes inconsistency in estimates is a result of lack of training across business units and wants to establish centralized governance and tooling. FinOps recommends a standard template and at a minimum all production estimates are peer reviewed by the FinOps team.


Scope Pull and organize actuals for the previous 12 months and compare to estimates submitted during account provisioning. Over 100 AWS accounts.
Out of Scope Existing Cloud workloads re-architectured, requires re-forecast but are out of scope for this project using first time estimation required to acquire an AWS account.




Key Tasks
  1. Pull all actual costs for every AWS account created in the last 12 months and organize by team & business unit.
  2. Gather all submitted budget estimates per AWS account from each Product Owner
  3. Develop Excel formulas to compare accuracy of budget estimates to actuals.
  4. All budget estimates are lessestimates less than actual by 10%. Gather insights from Product Owner where project is in life-cycle.
  5. All estimates greater than actual by 10%.  Gather insights from Product Owner where project is in life-cycle.
  6. Create an Excel report that highlights variances by business units and team

Types of insights : 

  • Understand who created the estimate
  • What was their process for creating the estimate
  • Tools used
  • Assumptions in estimates
  • Any gaps in the estimates
  • Which personas reviewed estimate
  • Understanding how estimate relates to budget
Tools Used AWS Cost Explorer

AWS Aurora Mysql

AWS QuickSight


Skills Used Excel formulas



Personas Interactions Technical Project Managers

Product Owners


FinOps Analysts

Key Discoveries One business unit “X” had a significantly higher variance of estimate to actuals

  • “X” unit did not have a template
  • “X” unit did not account for discounted commitment agreements with csp vendor
  • “X” over or under estimated commitment coverage
  • FinOps Analysts did not peer review estimates
  • All teams that estimated under or over actuals did not have any estimating training or understand the budget management process.

Teams that estimated the closest to actuals

  • Had significant more experience running workloads in Cloud
  • Baselined off similar workloads
  • Consultant or partnered with FinOps team to complete estimates
  • Exponential distribution was used over linear



Mentor Part of a cohort of interns

Manager met twice a week

Peer mentor assigned

Certification No Certification achieved during internship

12 leadership learning hours

Third party CFM training tool

Tools All AWS services were new tools learned by the intern.