The method

What your site says to an investor

I read your site the way an investor or a business development team reads it, and I write down what it does and does not tell them. Five criteria, every finding quoted from your own pages.

Pipeline stage and TRL

Checks whether your site tells a reader where the program is: pre-IND, Phase I, TRL 3, or whatever stage applies. Vague timelines fail this.

IP position and patent status

Checks whether your patent position is visible: filed, pending, issued, or licensed, without requiring a reader to dig through a regulatory filing.

Is the commercial case visible

Detects where dense academic jargon buries your core commercial value proposition.

BD and licensing signals

Checks whether a pharma BD team can find what you are offering: licensing terms, co-development interest, or partnership criteria.

Differentiation from similar companies

Checks whether a reader can place you against comparable companies: indication named, modality stated specifically, and a line on how you differ.

What this is

A read of your site, not a score of it

Most SBIR and STTR stage biotech sites are written by the people who did the science, for people who already understand it. That is the right instinct and the wrong audience. The reader you need is a generalist investor, a licensing manager, or a business development analyst who has forty other companies to get through this week.

The BioSite Audit is the note I write after reading your site as that person. It says what your pages state clearly, what they leave a reader to guess, and what a reader would have to already know in order to follow you. Every finding is tied to a line of your own copy.

It is for companies with real science and a site that undersells it. It is not a design review, a speed test, or an SEO report, and it will not tell you anything about your traffic.

The five criteria

What I look for, and what good looks like

These five are what a diligence reader is actually trying to establish. They are the same five every time, so two audits are comparable to each other.

CRITERION 1

Pipeline stage and TRL

Can a reader tell where the programme actually is?

What good looks like
Your stage is stated in words a reader can act on: IND enabling, Phase I, CE marked, TRL 4, first in human planned for a named period. The indication is named. If there is a timeline, it is on the page rather than in a deck.
The common failure
Platform language with no programme behind it. The site describes a technology in detail and never says what is being developed, for whom, or how far along it is. The middle case is just as common: the word preclinical appears once, with no indication and no dates, which tells a reader almost nothing.
CRITERION 2

IP position and patent status

Can a reader confirm that the technology is protected, without calling you?

What good looks like
At least one hard, checkable identifier is visible: an issued patent number, a disclosed pending application, a named exclusive university licence, a trial registry identifier, or a peer reviewed paper cited with its journal or DOI.
The common failure
The phrases patent protected and proprietary platform, with no number attached to either. This is the single most fixable finding on an SBIR stage site, because the patent usually exists and simply never made it onto the website. A reader who cannot verify it has to assume the weakest version of it.
CRITERION 3

BD and licensing signals

Can a partner tell what you want from them, and how to start?

What good looks like
You say plainly what kind of relationship you are open to: licensing, co development, a discovery collaboration, distribution. And there is a way in that is specific to that conversation, such as a named person who handles it.
The common failure
The word partnerships appears in the navigation and nowhere else, and the only route in is a general contact form. A business development analyst cannot tell whether you are looking to out licence, in licence, or neither, so you never enter their shortlist. Nothing about this failure is visible from the inside, because you know the answer and the page does not say it.
CRITERION 4

Is the commercial case visible

After reading your homepage, can a non specialist say what you do and who it helps?

What good looks like
The mechanism is stated plainly, in two sentences or fewer, near the top of the homepage. The application or the patient population is named where a reader will see it without scrolling for it.
The common failure
The homepage reads like the abstract of a paper. Every sentence is accurate and the reader still cannot say what you sell or who benefits. This is the most common finding on strong science, and it is worth being clear about what it measures: this criterion is about how accessible the case is, not whether the science is sound. A rigorous, correct, well written site can fail it badly.
CRITERION 5

Differentiation from similar companies

Does the site give a reader enough to place you against comparable companies?

What good looks like
The indication is named, the modality is stated specifically rather than by category, and there is a sentence on how your approach differs from what is done today. Specific means CAR T, CRISPR base editing, or multiplex IVD PCR, not therapeutic or novel platform.
The common failure
Category words doing the work of specifics. Targeted therapy, next generation sequencing, and AI driven each cover hundreds of companies, so a reader who wants to understand where you sit has nothing to work with and moves on to a company that made it easy.

How the assessment works

Grounded in your own words

The report is built from what is on your pages and nothing else, so every line of it can be checked against the source.

The pages I read

Your homepage, plus a small number of pages linked from it that carry the substance: technology, pipeline, products, publications, clinical. If the material lives somewhere unusual, tell me in the request and I will read that too.

Evidence is quoted, not paraphrased

Every finding cites a line of text from your site, reproduced word for word. You can check any statement in the report against the page it came from. Where a criterion has nothing on your site to support it, the report says so directly rather than reaching for something loosely related.

Nothing is invented

No patent numbers, trial identifiers, development phases, or publications appear in the report unless they appear on your site. If your IP is not on your pages, the report tells you it is missing from the pages. It does not go looking elsewhere and it does not fill the gap with a plausible guess.

What it does not do

It will not evaluate your science, value your company, or tell you whether your patent is strong. It reads what a stranger can learn from your website in a few minutes, which is a narrow question and often a useful one.

Request one

Ask me to read your site

I run the assessment myself and send the report back by email, usually within two working days. Nothing is assessed on this page and nothing happens to your site when you submit this. It is a message to me.

This is where the report goes.
The public site. A page behind a login cannot be read.

Or email support@codonforge.com directly.

Want a full rebuild?

Tell me about your platform

CodonForge builds sites for Seed and Series A biotech companies. One call: you describe the science, I tell you what the site should do and what it will cost.

Start a project See the work