# RunRL landing page growth score

Lazyweb publishes a public search API over its indexed pages; the guide is at https://www.lazyweb.com/agent-access.

Source: https://www.lazyweb.com/yc-p25/runrl
Source observed: 2026-09-30T02:44:50.501Z
Last checked: 2026-10-20T07:27:43.316+00:00
Indexed: 2026-09-30T06:56:18.174+00:00

---

# RunRL landing page growth score

We assessed the public landing page for [RunRL](https://runrl.com/) as part of Lazyweb's YC P25 growth-score reports. RunRL provides a platform for improving models with reinforcement learning. Its stated use cases include drug design and red-teaming; it also describes trained models for interpretability, formal verification, drug design, and automated red-teaming.

## Overall result

RunRL's growth score is **47 out of 100**, updated **September 20, 2026**. The weakest category is **Trust**, scored **32**.

| Category | Score | Assessment |
| --- | ---: | --- |
| Trust | 32 | Poor |
| CTA | 42 | Poor |
| Information hierarchy | 52 | Fair |
| Mobile web | 52 | Fair |
| Value props | 58 | Fair |

[Fix Trust](https://www.lazyweb.com/plans/checkout?plan=individual&source=growth_score_feature_build_fixes)

## Ranking

| Rank | Company | Growth score |
| ---: | --- | ---: |
| 1 | [OpenPipe](https://www.lazyweb.com/growth-score/site/openpipe.ai) | 76 |
| 2 | [Together AI](https://www.lazyweb.com/growth-score/site/together.ai) | 73 |
| 3 | [Adaptive ML](https://www.lazyweb.com/growth-score/site/adaptive-ml.com) | 64 |
| 4 | RunRL | 47 |

## Score over time

The report records these observations:

1. **September 20, 2026:** 47, live report; overall −19, CTA −30, Trust −26.
2. [June 16, 2026](https://www.lazyweb.com/growth-score/history/2026-06-16): 66; overall +2, CTA +4, Mobile web +4.
3. [May 16, 2026](https://www.lazyweb.com/growth-score/history/2026-05-16): 64; Value props +4, Information hierarchy −2.
4. [April 14, 2026](https://www.lazyweb.com/growth-score/history/2026-04-14): 64; Information hierarchy +6, Value props −4.
5. [March 24, 2026](https://www.lazyweb.com/growth-score/history/2026-03-24): 64; first tracked report, with strongest Mobile web at 72 and weakest Value props at 58.

## Trust — 32, poor

Trust measures proof that the product is real, competent, and safe to try. The report compares the visible trust signals on several pages.

![RunRL trust-section capture](https://zlfyzdmohcskkucuunmk.supabase.co/storage/v1/object/public/sites/growth-score/site/runrl.com.jpg?v=2026-09-20T20%3A57%3A24.978%2B00%3A00)

The visible RunRL signal is a **“Backed by YCombinator”** footer line. The comparison reads:

| Company | Trust signal |
| --- | --- |
| RunRL | “Backed by YCombinator” footer line |
| Adaptive ML | “Adaptive ML has been acquired by Datadog” announcement banner |
| OpenPipe | “Trusted by top companies” logo row featuring Zapier, Y Combinator, Flow, and Method |
| Together AI | “TRUSTED BY” logo strip with Cohere, DeepMind, LG AI Research, VFS.Global, and ElevenLabs |

The trust-category ranking is:

1. [OpenPipe — 76](https://www.lazyweb.com/growth-score/site/openpipe.ai)
2. [Together AI — 72](https://www.lazyweb.com/growth-score/site/together.ai)
3. [Adaptive ML — 66](https://www.lazyweb.com/growth-score/site/adaptive-ml.com)
4. RunRL — 32

[Build proof section](https://www.lazyweb.com/backlog?site=runrl.com#opt-trust-signals) to choose proof signals and receive copy and placement for a stronger trust section.

## CTA — 42, poor

The CTA assessment considers copy, size, verb, and cost-feel. RunRL's visible CTA is **“book a call.”** The report identifies these problems:

- There is no button, only an inline text link.
- A call feels costly and committal.
- The CTA has zero visual footprint above the fold.
- Two equal links split the action.

The CTA-category ranking is:

1. Together AI — 81
2. OpenPipe — 73
3. Adaptive ML — 62
4. RunRL — 42

[Rewrite this CTA](https://www.lazyweb.com/backlog?site=runrl.com#opt-cta-writing) to generate primary and secondary CTA options grounded in RunRL's offer and visitor objections.

## Information hierarchy — 52, fair

The report identifies three design issues:

1. **Inline text-link CTAs:** The primary actions **“email”** and **“book a call”** are body-text links without button weight, making them difficult to see at a glance.
2. **Footer credibility row:** **“Backed by YCombinator”** and **Promontory** appear as low-contrast links. The strongest trust signal reads as an afterthought and adds two exits.
3. **San Francisco link:** An outbound link has styling equal to the conversion links, diverting attention without conversion value.

![Information hierarchy annotations on the desktop capture](https://zlfyzdmohcskkucuunmk.supabase.co/storage/v1/object/public/sites/growth-score/site/runrl.com.jpg?v=2026-09-20T20%3A57%3A24.978%2B00%3A00)

[View layout plan](https://www.lazyweb.com/backlog?site=runrl.com#opt-information-hierarchy) for a sequenced plan to keep, move, merge, or remove these elements.

## Mobile web — 52, fair

The mobile assessment treats the phone experience as a first-class experience. It identifies:

1. **Text-link CTAs:** “email” and “book a call” are inline text links rather than tappable buttons, so they are easy to miss and difficult to hit with a thumb.
2. **Three-line wrapped headline:** The desktop headline **“RunRL: Reinforcement learning as a service”** wraps into three heavy lines and consumes approximately a quarter of the first viewport.

![Mobile web annotations on the mobile capture](https://zlfyzdmohcskkucuunmk.supabase.co/storage/v1/object/public/sites/growth-score/site/runrl.com-mobile.jpg?v=2026-09-20T20%3A57%3A24.978%2B00%3A00)

[View mobile fixes](https://www.lazyweb.com/backlog?site=runrl.com#opt-mobile-web) for a prioritized implementation checklist.

## Value props — 58, fair

RunRL's landing-page headline is:

> **RunRL: Reinforcement learning as a service**

Its supporting statement is:

> We provide a platform for you to improve your models with reinforcement learning.

### What works

- The product is immediately clear.
- The page names real use cases: drug design and red-teaming.
- The wording avoids buzzword inflation and overclaiming.

### What could improve

- There are no numbers or measurable outcomes.
- The subhead restates the headline.
- There is no visual showing the product working.

The value-props ranking is:

1. OpenPipe — 76
2. Adaptive ML — 62
3. RunRL — 58
4. Together AI — 56

## Competitors

The report infers these competitors from product, audience, and value proposition:

| Rank | Company | Destination | Score/status |
| ---: | --- | --- | ---: |
| 1 | [OpenPipe](https://www.lazyweb.com/growth-score/site/openpipe.ai) | openpipe.ai | 76 |
| 2 | [Together AI](https://www.lazyweb.com/growth-score/site/together.ai) | together.ai | 73 |
| 3 | [Adaptive ML](https://www.lazyweb.com/growth-score/site/adaptive-ml.com) | adaptive-ml.com | 64 |
| 4 | RunRL | runrl.com | 47 |
| — | [Predibase](https://www.lazyweb.com/growth-score/site/predibase.com) | predibase.com | failed |

[Back to YC P25](https://www.lazyweb.com/yc-p25)

---

[Read this page’s summary](https://www.lazyweb.com/api/public/v1/page-tldr?url=https%3A%2F%2Fwww.lazyweb.com%2Fyc-p25%2Frunrl&trace=t_bc7e1c4e5b07)

## Ask a follow-up about this page

Need a more specific answer? Ask the exact question you have. It returns a fresh, cited answer from Lazyweb’s knowledge base. No account or API key required.

Open this URL with your own question, URL-encoded. Only q is required. No account or setup.

`GET https://www.lazyweb.com/agent/faq?q={question}&trace=t_bc7e1c4e5b07`

[Tool guide and examples](https://www.lazyweb.com/agent-access) · [OpenAPI](https://www.lazyweb.com/api/public/v1/openapi.json) · [Page TLDR form](https://www.lazyweb.com/agent-tools/page-tldr?url=https%3A%2F%2Fwww.lazyweb.com%2Fyc-p25%2Frunrl). Opening a form does not execute a tool.