AnalyticsPredictive AnalyticsCloud ServicesBusinessTechnicalAzureEnterpriseSales

Estimating Cloud Spend — Rules of Thumb for AWS, Azure, and GCP

By February 9, 2018 No Comments

If you ever find yourself estimating cloud spend, I have some advice for you.

  1. Go through all the pricing calculators
  2. Outline all the resources you will need
  3. Call up professional help if you so choose
  4. Write everything down
  5. Sum up the monthly numbers
  6. and then…just straight up double your estimate

Direction changes regularly during a major development initiative, we have all experienced this. Give yourself some wiggle room and develop a monthly budget range.

Do not forget to consider periods of heightened development, UAT or QA

Double or even triple your estimated monthly costs when you are in heavy development cycles.

It is so easy to lose sight of the huge amounts of data you are often moving, the pressure & load testing, the unusual query volumes (not to mention the volume of poorly written queries), the constant abuse of negative testing, the list goes on… All of this adds up to increased spend when you are on the cloud.

Potentially double down again, for every extra environment you choose to add

Don’t forget — you are also going to want a separate development, QA and what the heck, at least two production environments. Then there is the georeplication check box and other disaster recovery mechanisms… So best to double spend estimates another time.

Moving to the cloud, you now have all of these fancy buttons to do what was a major headache before: deploying a backup, restoring a copy, spinning up a separate sandbox environment — all of this can be done with one hand while simultaneously drinking coffee or giving your pet some much deserved extra attention. Shiny new toys are played with, meant to be played with.

The cloud however, is pay-to-play… so keep all of this in mind.

Need a little guidance when it comes to the cloud? Aptitive’s cloud advisory service team may be able to help. Contact us for more information. 

A version of this post originally appeared on Medium.