The Importance of Internal Linking to Boost Site Performance and UX | Seer Interactive

Internal linking gets treated as a cleanup task: fix the orphans, add a breadcrumb, move on. Seer
Interactive's argument is that it deserves to be treated as a strategy, with a stated goal, a
hypothesis, and a metric chosen to match. Their case study is useful less for the numbers than for the
demonstration that the same tactic aimed at two different goals produces two different kinds of
result.

Two programmes, two goals

Results as reported by Seer Interactive. The measurement row is the one to read before quoting the others.

The contrast is the point. In the first engagement the objective was commercial routing: too many
visitors were landing on pages that could not convert them. Adding 80 or more internal links pushed
traffic toward product pages, and the reported outcome was a 192% increase in conversion rate, 76%
more transactions, and 116% more revenue, with 58% fewer visitors landing on non product pages.

That last figure is the most informative one in the whole case study. It tells you the
mechanism. Conversion rate did not improve because the product pages got better at
persuading; it improved because the mix of people arriving changed. Internal links redistributed
attention, and the conversion metric followed.

The second engagement aimed at rankings instead. Links were pointed at specific URLs targeting
competitive terms, and the average gain was 2.5 positions, with a 7.5% conversion rate increase and
13% more transactions on the affected pages. Smaller numbers, but a different objective and a
different metric.

Read the numbers with the method attached

Seer describe applying the scientific method: observation, hypothesis, tooling, implementation,
measurement. The measurement step used two week windows before and after the change, with no control
group.

This is worth stating plainly, because a 192% conversion rate increase is the kind of figure that
gets pulled into decks without its caveats. A before and after comparison over two weeks cannot
separate the internal linking change from anything else that happened in the same window: seasonality,
a promotion, a core update, other SEO work, a paid media shift. The result may well be real. The
design simply cannot tell you how much of it the links caused.

Compare this with the SearchPilot split tests covered elsewhere on this site. Their
nearby location links test
reported a 7% uplift, and their
related article links test
reported 16% on donor pages. Those figures are an order of magnitude smaller than 192%, and they are
also far more defensible, because half the pages acted as a live control absorbing exactly the same
seasonality and algorithm changes.

Both kinds of evidence are useful. They just support different claims.

A hierarchy for reading SEO case studies. Most published case studies, including the ones here, sit in the top row.

The practical rule: use before and after case studies to decide what to try, and use
controlled tests to decide what to promise. Presenting a 192% figure to a stakeholder as an
expected outcome is how SEO programmes acquire targets they cannot hit.

The part worth copying: opportunity discovery

Where the case study is genuinely instructive is the tooling. Seer built the opportunity list with
Screaming Frog for crawl data and Power BI to prioritise at scale, and implementation required only
CMS changes, so no engineering time was needed.

That last constraint shapes everything. A linking programme that needs a development sprint
competes with the product roadmap and loses. One that an editor can ship through the CMS gets done.
When you scope internal linking work, scope it to what the CMS can do.

The prioritisation logic is reproducible without Power BI. You need two data sets joined on URL:

# Step 1: crawl for existing internal inlink counts
Screaming Frog > Internal tab > filter: HTML
Bulk Export > Links > All Inlinks

# Step 2: join against Search Console query data
#   page | clicks | impressions | avg_position | inlink_count

# Step 3: the opportunity filter that matters
impressions > 500
AND avg_position BETWEEN 8 AND 25
AND inlink_count < 10

# These are pages Google already surfaces, ranks just off
# the money positions, and that the site barely links to.

That filter is the whole method in three conditions. Pages with real impressions are pages Google
already understands. Pages at positions 8 to 25 are close enough that a nudge changes traffic
materially, since moving from 12 to 8 is worth far more than moving from 60 to 45. Pages with few
inlinks are the ones where adding links actually changes their internal standing. Pages that satisfy
all three are where internal linking pays.

Our internal link analyzer
produces the inlink side of that join, and the
one followed internal link check
surfaces the extreme cases where a valuable page has essentially no internal support.

Finding where the links should go

Knowing which pages need links is half the job. The other half is finding pages that can
credibly link to them, which means pages already discussing the topic.

# Find pages that mention a target phrase but do not yet link to it
# (classic internal link opportunity discovery)

# In Screaming Frog: Configuration > Custom > Search
#   Filter 1: Contains     "running shoes"
#   Filter 2: Does Not Contain  "/category/running-shoes/"

# Any URL matching both is a page that discusses the topic
# but does not link to the page you want to promote.

# Quick command line version against a URL list:
while read u; do
  body=$(curl -s "$u")
  echo "$body" | grep -q "running shoes" \
    && ! echo "$body" | grep -q "/category/running-shoes/" \
    && echo "OPPORTUNITY: $u"
done < urls.txt

The output is a list of pages that mention your target topic and do not yet link to the page you
want to promote. Those are natural link placements, which matters for two reasons: the anchor text
writes itself from the surrounding sentence, and the link is genuinely useful to someone reading that
paragraph. Links inserted into content that has nothing to do with the target are the ones that read
as manipulation and tend not to survive editorial review anyway.

A workable process

1. Pick one goal. Routing traffic to converting pages and lifting rankings on
   competitive terms are different programmes with different target pages. Running both at once makes
   the results uninterpretable.
2. Build the opportunity list from the crawl and Search Console join above, then cut
   it to the twenty or thirty highest value targets. A list of 500 opportunities never gets worked.
3. Find contextual placements using the mention-without-link search, and prefer
   linking pages that already receive traffic.
4. Write anchor text from the sentence, descriptive and varied, never the same exact
   phrase 80 times.
5. Ship in batches and stagger them. Shipping everything at once means one
   uninterpretable before and after. Staggering across weeks lets you see which batches moved.
6. Hold a comparison group if you possibly can. Even without a testing platform, you
   can leave a matched set of similar pages untouched for the first month. That single step lifts your
   evidence from the top row of the table to the second.

What internal linking cannot do

The routing result is a useful corrective to the idea that internal links are purely a ranking
lever. Much of the commercial value in the first case study came from users following links to better
pages, which is a UX outcome that happens to be measured in revenue.

But the same framing sets the limits. Internal links distribute authority and attention that
already exist. They do not create demand, they do not fix a product page that lacks the information a
buyer needs, and they do not rescue thin content. If the destination page cannot convert the traffic
it already gets, sending it more traffic converts nothing.

For the wider framework, our
complete internal linking guide
covers the mechanics, and
topic cluster architecture
covers how to decide which pages should support which before you start adding links at all.

FAQ

Source: https://www.seerinteractive.com/blog/the-value-of-internal-links/