Sample report — complete & ungated
Constraint Brief: clinical trial supply company
This is a complete Constraint Brief from a real engagement, published in full and unedited. The client is not named and every identifying detail has been removed — the same confidentiality your own clients get, and the reason you can put this in front of a prospect without explaining whose site it is. Every figure is labelled MEASURED or NOT MEASURED, so you can judge the standard of the reasoning before you order.
In a white-labelled order, this report would carry your agency name and logo instead of DeepGro’s.
Download this brief as a PDF
The same report in the format a partner actually receives — a designed, five-page document you can forward to a prospect without editing it. On a white-labelled order it carries your agency’s name and logo instead of ours.
Before you read it
How to read the figures
Every number here is labelled MEASURED (counted directly from the live site or its sitemaps — reproducible today), or NOT MEASURED (something only the client’s own analytics can show). Nothing is asserted without a label, and nothing is estimated to look more impressive than it is.
This brief deliberately contains no revenue projection. Sizing the opportunity requires organic sessions, enquiry conversion rate and enquiry value — three numbers held internally. We say so rather than inventing them.
01
Executive summary
The subject runs a substantial content operation — 176 blog posts and a 26-page clinical glossary, competently written, averaging 1,237 words per post. It is not under-invested in content.
The problem is that this content barely connects to the ten pages that sell anything. In a sample of 18 posts, the average post contained 5.5 in-content links to other blog posts and 0.6 to any service page. Seven of the eighteen contained none at all. Authority accumulates in the editorial layer and stays there.
02
The constraint
Constraint identified
The content engine has no route to the commercial pages
A routing problem, not a volume problem. The editorial layer is roughly nine times better connected to itself than to revenue.
| In-content links per post | Average | Basis |
|---|---|---|
| To other blog posts | 5.5 | MEASURED · 18-post sample |
| To any service page | 0.6 | MEASURED · 18-post sample |
| To the contact page | 0.2 | MEASURED · 18-post sample |
| Posts with zero service link | 39% | MEASURED · 7 of 18 |
Header, navigation and footer links were excluded from these counts. Global nav links to /services/ from every page; that is not editorial linking and counting it would have hidden the finding.
03
The shape of the site
250 URLs across five sitemaps. Twelve of them can take an order.
| Layer | URLs | What it does |
|---|---|---|
| Blog posts | 176 | Informational — regulatory round-ups, explainers, sourcing commentary |
| Glossary terms | 26 | Definitions. Almost pure top-of-funnel |
| Event pages | 18 | Trade shows, across two competing URL patterns; five already past |
| Service pages | 10 | The commercial layer |
| Core pages | 11 | Home, about, why-partner, contact and similar |
| Case studies | 2 | The only proof assets on a high-consideration purchase |
| Categories & legal | 6 | Archives, privacy, cookies |
17 informational pages for every commercial page. Not wrong in itself — a specialist supplier should own the vocabulary of its field — but the two layers are barely joined.
MEASURED · counted from sitemap_index.xml and its five child sitemaps.
04
Why it is happening
The content brief carries no commercial obligation. Posts answer a question and end. Nothing in the workflow requires a post to route the reader to the service that solves the problem it just described.
The services hub is the weakest page on the site. /services/ — parent of every commercial page — renders no H1 and carries an auto-generated 40-character meta description reading “Services Archive”. It is an untouched CMS archive template. Link equity arriving there lands on a page that cannot rank.
Proof is thin at the point of decision. Two case studies support a purchase involving regulated supply chains and procurement cycles measured in months.
05
Supporting findings
| Finding | Where | Severity |
|---|---|---|
| Services hub has no H1 and an auto-generated meta description | /services/ | Critical |
| Editorial layer not routed to commercial pages | Across /blogs/* | Critical |
| Two URL patterns for the same events, both indexable, each self-canonical | Event pages | High |
| Five event pages for past dates still indexable, no archive treatment | Event pages | High |
| Two case studies against ten service pages | /case-studies/ | High |
| No H1; title tag and URL describe different things | About page | Medium |
Organization schema present but declares no sameAs entities | Homepage | Medium |
| 1,982 words under a single H2 | Main service page | Medium |
| 61 script elements on one template — worth field-testing for responsiveness | Blog template | Medium |
| Correct Service and Offer schema, 54-character title, clean canonical, H1 present — the template the other nine should follow | Main service page | Strength |
| No AI crawler restrictions in robots.txt; the glossary is unusually well suited to AI citation | /robots.txt | Strength |
06
What we would do, in this order
1. Rewrite the services hub
Give it an H1, a written introduction, a real meta description and descriptive links to all ten service pages. Half a day. First, because every later linking decision points at it.
2. Editorial linking pass over the top 40 posts
Not all 176. The 40 carrying the most authority, two or three in-content links each into the relevant service page, anchors naming the service. This is the constraint and the fastest lever — the authority already exists.
3. Fix the content brief so it stops recurring
Every future post names the service it routes to before it is written. Without this, step 2 has to be repeated every year.
Five further items — the event duplication, stale event pages, missing H1, sameAs, and script weight — are real and worth fixing. None is what is currently holding enquiries back, and doing them first would consume a quarter without moving the number. The full diagnostic sequences all of them.
07
What this brief cannot tell you
Everything above is externally verifiable. The site, its sitemaps and its markup were analysed directly, so every figure can be re-checked independently — by you, or by anyone you present this to. A finding a client can confirm carries more weight than one they must take on trust.
NOT MEASURED: current traffic, which queries rank, conversion rate, enquiry value, and real-world page performance. We had no analytics, Search Console or CRM access. Linking figures come from 18 of 176 posts — the pattern held across the sample, but exact averages will shift on a full crawl.
What more data would change: the size of the prize, not the diagnosis. That 176 posts link to each other nine times more often than to the pages that sell anything is a counted fact about this site’s structure.
Want one of these for a prospect you are pitching?
Send the domain and the market they compete in. It comes back within 24 hours, branded as your agency’s work, for $95. One brief covers one website.