Azure Cost Advisor - Cost savings is not aligned with current costs

Question

I had a lot of recommendations from Azure to save costs on my resources deployed under my subscription. The potential saving recommended on Azure advisor is far more than my current charges. For example it has suggested to Resize/stop the under utilized virtual machines to save upto 2000$ but I am actually paying 250 $ dollars annually.

It has also suggested to Auto scale my cosmos DB to save more than 10,000$ annually while I am paying only 5000$ annually.

Please advise why the Azure advisor is showing the inaccurate savings amount and how is this being calculated on azure side. currently I am on Pay as you go plan and paying based on my usage.

My another query is

It has also suggested to opt for reserved instance on the subscription level and wanted to check how the price is suggested over there because it is also seems inaccurate.

Answer 1

Hello Mohammed Shaik, Azure Advisor’s cost-saving recommendations are based on several factors and assumptions, which can sometimes lead to discrepancies between the suggested savings and your actual costs. Here are some reasons why you might see these differences.

1. Retail rates vs Actual costs - Azure Advisor often calculates potential savings based on retail rates, which might not account for any discounts, promotions or special pricing you currently have. This can make the recommended savings appear higher than what you're actually paying
2. Utilization metrics - Recommendations are based on utilization metrics over a specific period (e.g., the last 7, 14, or 30 days). If your usage patterns vary significantly, the recommendations might not accurately reflect your typical usage
3. Assumptions of resource usage - The recommendations assume that resources can be resized or shut down without impacting performance or availability. In reality, some resources might be underutilized by design, or resizing them may not be feasible
4. Lookback period - The lookback period for analyzing resource usage can affect the recommendations. If the period used for analysis doesn't match your typical usage patterns, the recommendations might not be accurate.

Regarding the Reserved Instances recommendations, the pricing is based on the assumption that you will commit to a one-year or three-year term, which typically offers significant discounts compared to pay-as-you-go pricing. However, if your actual usage is lower or if you have other discounts, the suggested savings might not align with your actual costs.

To get more accurate recommendations, you can:

• Adjust the lookback period in Azure Advisor to better match your typical usage patterns
• Review the specific metrics and assumptions used in the recommendations
• Consider any special pricing or discounts you have when evaluating the recommendations

Please do not forget to "Accept the answer” and “up-vote” wherever the information provided helps you, this can be beneficial to other community members.