Methodology
How StampSnap Estimates Stamp Values
A transparent look at how our AI computes value estimates — the data we use, how we calculate, and what you should know before relying on the results.
Important: All value estimates provided by StampSnap are AI-generated and for informational purposes only. They do not constitute professional appraisals. Stamp values can vary significantly based on precise condition, market fluctuations, and buyer demand. Always consult a certified philatelic appraiser before making insurance, purchase, or sale decisions.
1. Data Sources
Our value estimates draw from multiple data sources to produce a balanced reference price:
Stamp Catalogs
Scott (US), Stanley Gibbons (UK), Michel (Germany), and Yvert (France) catalog prices provide the baseline catalog value. These are the most widely recognized references in philately.
Auction Results
Recent sale prices from major stamp auctions help adjust catalog values toward real-world market prices. We aggregate data from public auction archives and dealer price lists.
Dealer Price Lists
Current retail prices from established stamp dealers provide a market-floor perspective. Dealer prices typically sit between catalog values and auction hammer prices.
Online Marketplaces
Listings and completed sales on platforms like eBay, Delcampe, and HipStamp give insight into current market liquidity and what collectors are actually paying.
Community Data
Anonymized, aggregated data from StampSnap user scans helps us identify trends and refine estimates over time. Individual user data is never shared or sold.
2. How Estimates Are Computed
The computation is a multi-step process combining the identified stamp's attributes with our reference data:
Step 1: Baseline Value
Once the AI identifies the stamp (country, catalog number, year, variety), we retrieve the baseline catalog value from our reference database.
Step 2: Condition Adjustment
The AI analyzes your photo for visible condition indicators — centering, cancellation intensity, perforation integrity, color freshness, and visible faults. Each factor adjusts the baseline up or down.
Step 3: Market Adjustment
Catalog values are adjusted based on recent market data to reflect current demand and typical sale prices vs. catalog listings.
Step 4: Value Range
Rather than a single price, we produce a low-to-high range (typically ±10%). This reflects the natural variability in stamp pricing — the same stamp in similar grade can sell for different prices depending on buyer, venue, and timing.
Step 5: Confidence Score
The overall confidence percentage reflects how well the submitted image matches known catalog data and how reliably the condition could be assessed. Low-confidence results may have wider value ranges.
3. What Affects Value
Our model considers these factors when computing the final estimate. Each is assessed from the submitted photo where visible:
Grade / Condition
The single biggest factor. Mint, lightly hinged, or used — each has a different price tier.
Cancellation
Heavy or light? Fancy cancel or socked-on-the-nose? Light cancels preserve value.
Perforations
Missing teeth, short perforations, or damaged edges reduce value significantly.
Watermark
Some watermarks are rarer than others on the same stamp design. Affects variety pricing.
Gum
Original gum (OG) adds premium. Disturbed gum, regummed, or no gum all grade differently.
Centering
Well-centered designs command dramatically higher prices than poorly centered examples.
4. Limitations
We believe transparency is essential. Here is exactly what StampSnap cannot do:
Not a Professional Appraisal
StampSnap does not employ certified appraisers. Our estimates are algorithmic, not expert opinions. For formal valuations (insurance, estate, legal), always consult a recognized philatelic expert.
Condition Assessment Is Visual Only
Our AI can only assess what is visible in the photo. Hidden faults (thin spots, repaired tears, re-gumming) cannot be detected from an image. In-person inspection by a professional is the only reliable way to assess true condition.
Catalog Coverage Is Not Exhaustive
While we strive to cover major world catalogs, many local issues, provisional stamps, and obscure varieties are not yet in our reference data. No estimate is provided for unidentified stamps.
Market Volatility
Stamp prices fluctuate. An estimate reflects available data at the time of calculation, not a guaranteed future sale price. Rare stamp auctions can produce unexpected results in either direction.
Photo Quality Limits Accuracy
Poor lighting, blur, shadows, or angle distortion reduce our ability to assess condition accurately. Higher quality photos produce more reliable estimates.
5. Update Frequency
Our reference data is updated on the following schedule:
- check_circle
Catalog prices — refreshed annually (in line with new catalog editions from Scott, Stanley Gibbons, Michel, and Yvert)
- check_circle
Market data — aggregated quarterly from auction archives and dealer price lists
- check_circle
AI model — retrained periodically with new stamp varieties and improved condition assessment capabilities
- check_circle
Community trends — analyzed monthly from anonymized usage data to identify shifts in collector interest
Last full data refresh: January 2026.
Have questions about our methodology?
We welcome feedback and questions from the philatelic community.
mail Contact Us