NET · Forward model · Self-serve and mid-market
What has to happen in Self-serve and mid-market
Model as of
This page changes Self-serve and mid-market inside the complete NET model. Every other vertical, corporate programme and valuation assumption stays in the selected portfolio case.
Shares this vertical and portfolio case. Slider and horizon edits stay in your browser.
Self-serve and mid-market
Basis quarter$188M
Final quarter$371M
Implied CAGR+15%
Final revenue mix16%
Everything under $100,000 a year: the free tier converting to Pro and Business, the millions of small properties, and the mid-market accounts that have not crossed into the large cohort yet. It is 27% of revenue and shrinking as a share, but it grew 26% year over year in absolute terms, and it is the only funnel the large-customer cohort is fed from.
Last four quarters
2025 Q3
$152M
Estimated
2025 Q4
$166M
Estimated
2026 Q1
$179M
Estimated
2026 Q2
$188M
Estimated
Pro and Business plansMid-market contracts below the $100k lineUsage-based Workers and R2 spend on self-serve accounts
Sequential growth
+5.0%/qtr
decaying toward +2.5%
5% a quarter, below the 6.1% this residual ran at over the last year, as graduating accounts leave for the large cohort.
Self-serve and mid-market
Latest: $371M (2031Q2E)
| Period | Value |
|---|---|
| 2025Q2 | $149M |
| 2025Q3 | $152M |
| 2025Q4 | $166M |
| 2026Q1 | $179M |
| 2026Q2 | $188M |
| 2026Q3E | $197M |
| 2026Q4E | $207M |
| 2027Q1E | $216M |
| 2027Q2E | $225M |
| 2027Q3E | $234M |
| 2027Q4E | $243M |
| 2028Q1E | $252M |
| 2028Q2E | $261M |
| 2028Q3E | $269M |
| 2028Q4E | $278M |
| 2029Q1E | $287M |
| 2029Q2E | $296M |
| 2029Q3E | $305M |
| 2029Q4E | $314M |
| 2030Q1E | $323M |
| 2030Q2E | $333M |
| 2030Q3E | $342M |
| 2030Q4E | $351M |
| 2031Q1E | $361M |
| 2031Q2E | $371M |
Assumptions & reasoning
- This is the one line modelled on a growth rate, and it is a last resort rather than a choice. Cloudflare discloses the large-customer count and share but publishes nothing about the shape of the remainder — no count, no price, no retention. A driver built from invented inputs would look more rigorous and be less true, so the residual gets a rate and an honest label.
- It carries the highest margin in the model and the lowest capex, because self-serve costs almost nothing to sell and runs on the network the enterprise cohort already paid to build. That is also why it is the wrong place to look for growth: this cohort exists to feed the large-customer line, and every account that succeeds here leaves it.
- Being a residual, this line absorbs every error in the split above it. If the disclosed large-customer percentage is rounded up in a quarter, this line is rounded down by the same dollars. Do not read a wobble in its sequential growth as a change in the self-serve business.