Pages · daily Sitemap diff — No bulk page changes lately
We diff its public Sitemap daily. Bulk launches or removals will appear here and alert watchers.
Refraction Traffic, SEO & Growth Analysis
Refraction is a code generation tool for developers that uses AI to generate code, unit tests, documentation, and more
- Traffic declining · 2 months running
- Code Generator
- Documentation Generator
- Unit Test Generation
SIGPIK VERDICT · Sep 2026
Refraction visits fell 59% in Sep, mostly lost from Direct
Reconstruct every step Refraction took →- Monthly visits
- 2.2K
- ↓ 59.4% · −3.3K MoM
- Market rank
- #16 / 28
- 0.3% of market traffic
- Domain Rating
- 41
- Flat vs last month
- AI referrals
- 0
- 0% of traffic, early
01 · How the growth happened
Refraction Traffic Trend & Channel Attribution
Monthly visit estimates · 2026-05 — 2026-09 · DR 41, flat
Sep 2026
2.2K↓ 59%
Where the new visits came from
Monthly visits split by channel share: Refraction lost 3.3K visits from Aug to Sep, with direct contributing 39%.
Monthly breakdown
| Channel | Jul | Aug | Sep |
|---|---|---|---|
| Total | 5.9K | 5.5K | 2.2K |
| Direct | 0 | 2.1K | 867 |
| Organic search | 0 | 1.8K | 774 |
| Referrals | 0 | 561 | 213 |
| Organic social | 0 | 300 | 124 |
| 0 | 186 | 70 | |
| Display ads | 0 | 208 | 61 |
| Paid search | 0 | 88 | 28 |
| Paid social | 0 | 42 | 14 |
| Channel | Change barAug → Sep | Change | Share of change | Current share |
|---|---|---|---|---|
| Direct | −1.3K | 39% | 38.7% ↑0.1pp | |
| Organic search | −988 | 30% | 34.6% ↑2.6pp | |
| Referrals | −348 | 11% | 9.5% ↓0.7pp | |
| 5 other channels | −527 | 16% | 13.3% ↓1.6pp |
02 · What you can borrow
Refraction SEO Keywords, Sitemap Pages & AI Referrals
Refraction’s acquisition moves across pages, search terms and AI referrals, each with its data source.
Search · source keywords — Non-brand terms bring 20% of search traffic
Non-brand terms are traffic anyone can compete for; higher CPC means higher commercial value.
| Keyword | Traffic | CPC | Rank |
|---|---|---|---|
| refraction code generatorBrand | 80 | — | 1 |
| unittest pythonNon-brand | 10 | $0.70 | 2 |
| refactoring techniquesNon-brand | 10 | — | 3 |
| refactoringNon-brand | 0 | $3.07 | 4 |
| unittestNon-brand | 0 | $1.69 | 5 |
Borrow: Search traffic is mostly branded, riding existing awareness; the non-brand needs it does not cover are open to you.
AI referrals · ChatGPT, Copilot… — AI referrals are early — not a growth source yet
0 visits from AI assistants (previous month 0), 0% of all traffic.
Borrow: No need to follow yet; watch whether it scales next month.
03 · Full growth history
Refraction Promotion History, Backlinks & Launch Timeline
Above is what we observe automatically. The reconstruction starts from these moments, searches launches, community threads, directory listings and backlinks from the same period, fills in what Refraction did at each step, and gives you a sequence to follow.
- Promotion timeline with original sources
- Traffic before and after each channel
- Exportable launch action list
Monthly visits 5.5K → 2.2K
Observed · traffic dataMonthly visits 3.8K → 5.9K
Observed · traffic dataMonthly visits 2.8K → 3.8K
Observed · traffic dataWhere it launched and who discussed it
Product Hunt, Reddit, X, directoriesWhich channels produced new backlinks
Ordered by first seenA launch sequence you can follow
Exportable action list
04 · Same-market benchmark
Refraction Competitors & Alternatives Traffic Benchmark
Sep 2026 · Code Generation and Understanding
| Product | Monthly visits | DR | MoM |
|---|---|---|---|
| RefractionThis product | 2.2K | 41 | -59.4% |
| CodeGPTSame market | 209.9K | 62 | +11.5% |
| TabnineSame market | 110.6K | 78 | -15.7% |
| Your Dropbox is now a skill server | sx blogSame market | — | 94 | — |
Each product uses its latest observed month; different months are not a same-period ranking.
Know the next time Refraction ships pages in bulk
We diff its Sitemap, backlinks and traffic daily and alert you only on real changes.
Method: visits and channel shares are third-party monthly estimates; channel change = monthly visits × channel share, month over month; site actions and traffic are shown by timing only, not as cause and effect.