# Your page scores 38 out of 100 for sameness against the 3 competitors you named — low enough that buyers can probably tell you apart.

You are telling a different story from Weco and the rest of the set, overlapping only on what you say you do and who it's for.

- **Your page:** https://autoscience.ai/autoresearch
- **Compared against:** Weco (weco.ai), Pavo (pavoai.com), Coral Autoresearch (coral-labs.ai)
- **Sameness Index:** 38 — Mostly your own
- **Ownable share:** 28% (of 59% distinctive)
- **Report:** https://grader.wynter.com/c/autoscience-ai-7j_YdyA

**How you compare · vs 366 SaaS sites** — 18 points less same than the average SaaS site. The average SaaS site scores 56; you scored 38. You are less same than 340 of the 366 SaaS sites. The best-differentiated site in the set is at 6.
_A single check can move several points between runs; the benchmark averages up to eight sites per category._

> With 3 competitors, commonness can only take 4 distinct values. The figures are honest at that resolution and no finer.

## 1. Sameness Index

| Category | Weight | Score | What it means |
| --- | --- | --- | --- |
| Messaging | 30% | 70 | The most expensive kind of sameness. A buyer cannot tell what job you do that the others do not. |
| Claims | 30% | 17 | Every shared claim is a line already read on another tab. Cut the ones nobody owns and spend the space on something they cannot. |
| Features | 15% | 53 | Expected in a mature category, and the least alarming of the six. Feature parity is normal; leading with it is the mistake. |
| Proof | 10% | 21 | Same kinds of proof as everyone means the proof stops working as proof. It is scored on the kind of evidence, not on which customers are named. |
| Structure | 10% | 0 | The generic SaaS template — hero, logos, three-feature grid, testimonial, CTA. Familiar is not the same as memorable. |
| Visual | 5% | 44 | Weighted lowest on purpose: buyers rarely decide on this. Worth knowing, rarely worth fixing first. |

**Mostly your own** — Most of what you say, your competitors don't.

## 2. Closest competitors

| Competitor | Echoes your claims |
| --- | --- |
| Weco | 54% |
| Pavo | 37% |
| Coral Autoresearch | 33% |

_Measured from your side: the share of what YOU say that each competitor also says, weighted by where you put it. This is not a Sameness Index — a competitor can echo most of your page and still score low on its own index by saying a great deal you do not._

## 3. Ownable share

**The share of your own page's claim space that is yours to keep.** This is not the inverse of the Sameness Index — it is measured on your claims alone, so the two do not add up to 100 and are not meant to.

28% of your page is unique, relevant and defensible — a strong share, and the thing to protect in any rewrite.

- **Distinctive message share:** 59% — prominent space spent on claims no competitor makes.
- **Ownable share:** 28% — the part of that which is also relevant to how buyers choose and hard to copy.

## 4. Where you sound like everyone else

Most of your set makes each of these and you are spending prominent space on them, ranked by how much. **These are not necessarily lines to delete** — some of what a page has to say is unavoidably shared, and claims about what kind of product you are were left out of this list for exactly that reason. The question each one raises is whether it has earned its space.

- **autonomous AI agent runs ML research/experiments** (Hero, 100% of the set) — also on Weco, Pavo, Coral Autoresearch
  - _"It identifies promising research, executes experiments within your stack, and ships verified pull requests that beat your baseline"_
- **beats or exceeds human researcher output** (Section, 100% of the set) — also on Weco, Pavo, Coral Autoresearch
  - _"Evaluate all the ideas that you normally wouldn't have capacity to"_
- **benchmark and leaderboard results as proof** (Section, 100% of the set) — also on Weco, Pavo, Coral Autoresearch
  - _"96th percentile in public ML competitions"_
  - Said your way: nobody else makes this exact claim, but everyone occupies the territory.
- **beats or exceeds human researcher output** (Section, 100% of the set) — also on Weco, Pavo, Coral Autoresearch
  - _"Mira autonomously outperforms nearly every human at ML research"_
  - Said your way: nobody else makes this exact claim, but everyone occupies the territory.
- **autonomous AI agent runs ML research/experiments** (Section, 100% of the set) — also on Weco, Pavo, Coral Autoresearch
  - _"24/7 experiments in flight"_
  - Said your way: nobody else makes this exact claim, but everyone occupies the territory.
- **optimizes whatever metric you define** (Section, 67% of the set) — also on Weco, Pavo
  - _"You name the metric. Mira optimizes for it."_
- **faster, smaller, cheaper models and compute** (Body, 67% of the set) — also on Pavo, Coral Autoresearch
  - _"Speed: Faster inference"_
  - Said your way: nobody else makes this exact claim, but everyone occupies the territory.
- **faster, smaller, cheaper models and compute** (Body, 67% of the set) — also on Pavo, Coral Autoresearch
  - _"Footprint: Smaller models"_
  - Said your way: nobody else makes this exact claim, but everyone occupies the territory.

## 5. Buried gold

Unique, relevant, and sitting in body copy. The cheapest wins here — you already have the differentiation and the page hides it.

- **Improves inference speed** (Body, 0% of the set)
  - _"Speed: Faster inference"_
- **Reduces compute cost/resource use** (Body, 0% of the set)
  - _"Cost: Resource efficiency"_
- **Augments rather than replaces your researchers** (Body, 0% of the set)
  - _"Does not replace researchers; runs experiments your researchers don't have bandwidth for and reports results, team reviews and steers"_
- **Improves model quality/accuracy** (Body, 0% of the set)
  - _"Quality: Smarter models"_
- **Customer retains ownership of model and code** (Body, 0% of the set)
  - _"You own the model and the code; nothing is licensed back"_
- **Your data never leaves your network** (Body, 0% of the set)
  - _"Your model and data remain yours; with on-prem deployment, nothing leaves your network"_

## 6. Claim inventory

| Claim | Type | Placement | Common | Sayability | Relevant | Ownership |
| --- | --- | --- | --- | --- | --- | --- |
| Agent proposes, runs experiments, ships code changes | capability | Hero | 67% | some | yes | Commodity |
| Positions as autoresearch/automated research category | category | Hero | 67% | anyone | yes | Commodity |
| Only submits changes that beat the baseline | proof | Hero | 33% | some | yes | Contested |
| Deployable on-prem or in your cloud | capability | Hero | 0% | some | yes | Unique and owned |
| Explores more ideas than the team has capacity for | outcome | Section | 33% | anyone | yes | Contested |
| Benchmark ranking in public ML competitions | proof | Section | 0% | some | yes | Unique and owned |
| Outperforms human ML researchers autonomously | outcome | Section | 0% | some | yes | Unique and owned |
| Runs experiments continuously, 24/7 | capability | Section | 0% | anyone | yes | Unique for now |
| Optimizes any metric you specify | capability | Section | 33% | anyone | yes | Contested |
| Ingests published research and applies it to your model | capability | Section | 0% | some | yes | Unique and owned |
| Thousands of new ML papers published weekly | proof | Section | 0% | anyone | no | Unique for now |
| You never miss relevant new research | outcome | Section | 0% | anyone | no | Unique for now |
| Improves inference speed | outcome | Body | 0% | anyone | yes | Unique for now |
| Produces smaller models | outcome | Body | 0% | anyone | no | Unique for now |
| Reduces compute cost/resource use | outcome | Body | 0% | anyone | yes | Unique for now |
| Setup needs only repo access and a metric | quality | Body | 33% | anyone | yes | Contested |
| Adds new model capabilities/features | outcome | Body | 0% | anyone | no | Unique for now |
| Augments rather than replaces your researchers | quality | Body | 0% | anyone | yes | Unique for now |
| Improves model quality/accuracy | outcome | Body | 0% | anyone | yes | Unique for now |
| Improves sample efficiency of data | outcome | Body | 0% | anyone | no | Unique for now |
| Customer retains ownership of model and code | quality | Body | 0% | anyone | yes | Unique for now |
| Your data never leaves your network | quality | Body | 0% | some | yes | Unique and owned |

## What we compared

- **Mira (by Autoscience)** (your page) — https://autoscience.ai/autoresearch
- **Weco** — https://www.weco.ai/platform
- **Pavo** — https://www.pavoai.com/interventions
- **Coral Autoresearch** — https://coral-labs.ai/autoresearch/

## What this does and does not tell you

An AI read each page and wrote down the claims it makes, then grouped the claims that say the same thing. **No score in this report was written by a model** — every number is computed from those groupings in our own code, with the weights shown above.

## The only way to know if it matters

This report can tell you where your messaging overlaps. It cannot tell you whether a buyer would care, or which of your reasons to buy actually land. Put the page in front of real B2B buyers in your target market and ask them.

[Test it with real buyers](https://wynter.com)