Deep page links for a past-paper library: a worked sketch
Imagine a small site that hosts old exam papers with mark schemes. This is an invented walk-through, built to show reasoning, not to report results.
ExamArticles editorial team ·
The site, the numbers and the outcome below are invented for teaching. No real customer is described, and nothing here predicts what your site would see.
Before and after, on paper
| Before | After the plan | |
|---|---|---|
| Pages | 400 papers, one weak homepage | Same papers, a clearer homepage and five hub pages |
| Links go to | Almost all to the homepage | Mix of homepage, hubs and two best papers |
| Anchors | Mostly exact subject words | Brand name, plain phrases, a few subject words |
| Pace | Twenty orders in one afternoon | Two or three orders each week |
| Measurement | Glance at rankings | Search Console clicks by page, monthly |
The reasoning, step by step
Four hundred papers sound like a lot to promote. They are not, because most of them answer a question few people ask.
So the sketch picks hub pages: one per subject and level, each listing papers with a short note on how to use the mark scheme. Those hubs are the pages a teacher would actually pass to a student, which makes them natural link targets.
Homepage links stay in the plan, since a visitor who arrives there should see who runs the library and which exam boards it covers. Plan for roughly one in three links to land on the homepage; the rest go to hubs and a couple of strong papers.
What we would measure, and when
Pick two numbers before ordering and ignore the rest. Impressions for the hub pages and clicks to the homepage for your brand name are enough.
Search Console shows both for free. Compare four-week windows, not days, because exam queries rise and fall with the school calendar and a January bump may simply be mock season.
If you want to check the donors afterwards, Ahrefs, Semrush or Majestic can show whether the new links were picked up. Their indexes differ from ours and from each other, so a link missing in one is not proof it is missing everywhere.
Read the scorecard again before the second batch. Sketch or not, the habit is worth keeping.
The sketch as a four-week routine
- 1
Week 1: tidy
Rewrite the homepage first screen, add dates to each paper, and remove duplicates. Nothing is ordered yet.
- 2
Week 2: first batch
Add two or three hub-page links to a basket from donors that pass the scorecard. Pay and note the date.
- 3
Week 3: second batch
Add one homepage link and one link to the strongest paper. Vary the anchor text from week 2.
- 4
Week 4: look, do not buy
Read Search Console for the target pages. If nothing changed yet, wait. Google can take weeks to recrawl.
What the sketch leaves open
Would this have worked for a real library?
We cannot say. The sketch is a method, and real results depend on the pages, the donors, competition and Google's systems, none of which an example can predict.
Are exam papers copyright-safe?
Not automatically. Boards set their own terms for past papers, so read them before hosting anything. Ask the board if unsure; links to a page you should not host would only make the problem more visible.
Should hub pages link to each other?
Yes. Internal links help visitors and crawlers, and they cost nothing. See the mistakes list for the common slips.