← Google Interview Insights

Google·Product Manager·Onsite - Product Sense / Strategy·Senior

SeniorPrefer not to say
Jun 2026

Summary

Interviewed at Google for what seemed like a product or engineering estimation role. One question, classic Fermi-style, and I left feeling like I either nailed it or completely missed the point.

Questions Asked (1)

Q1

Estimate the total cost of storing all of Google Earth's photos.

System DesignTechnical Trade-offsProduct Analytics & Metrics
Author's notes

I started by trying to break down the surface area of the Earth, then guessed at image resolution and coverage overlap.

Create a free account to read the full note

AI HintsAI Generated

Suggested Approach

Break down the problem into estimating the total volume of imagery data (in petabytes) and then applying cloud storage costs per GB/month. Use a top-down approach: estimate the number of high-resolution photos, average file size, and total storage needed, then multiply by Google Cloud storage rates. State assumptions clearly and consider compression and tiered storage.

Pro tip: Differentiate between raw imagery and processed tiles; Google Earth likely stores both, but the majority of cost comes from the raw data. Also, mention that Google uses its own infrastructure, so costs are internal and lower than public cloud rates.

1. Clarify scope and assumptions

Define what 'Google Earth's photos' includes: satellite imagery, aerial photos, Street View? Assume we're estimating storage cost for all imagery data. State assumptions about resolution, coverage, and update frequency.

2. Estimate total data volume

Calculate the number of images and average file size. For example, Earth's land area ~150 million km²; at 1m resolution, each km² needs ~1 million pixels; with 3 bytes per pixel (RGB), that's ~3 MB per km² uncompressed. Multiply by area and consider multiple layers/zooms.

3. Convert to storage units and apply cost

Convert total data to GB or TB. Use a cost per GB/month for cloud storage (e.g., Google Cloud Storage ~$0.02/GB/month for standard). Multiply to get monthly cost, then annualize.

4. Adjust for real-world factors

Account for compression (e.g., JPEG reduces size by 10x), redundancy (replication factor 3), and tiered storage (cold data cheaper). Also consider that Google uses custom infrastructure, so cost may be lower.

5. Sanity check and present range

Compare with known data points (e.g., Google Earth has petabytes of data). Present a range (e.g., $10M-$100M/year) and highlight key drivers.

Key Points to Mention

  • Total data volume estimation: area × resolution × layers
  • Compression and deduplication impact on storage size
  • Storage cost per GB/month and tiered storage options
  • Redundancy and replication for durability
  • Google's internal infrastructure vs. public cloud costs
  • Update frequency and versioning of imagery

AI-generated suggestions, not part of the candidate's original notes. May be inaccurate — verify before relying on them.