How Do I Know If My Link Building Is Working?

Judge link building on three things in order: are you gaining referring domains you'd have been glad to earn, are the target pages gaining impressions and average position in Search Console, and are those pages gaining organic sessions and conversions. Tier one moves in weeks, tier two in months, tier three last. If tier one is moving and tier two is flat after six months, the links are real but the wrong ones.

The awkward truth underneath the question is that you have no control group. You cannot run your site without the links and compare. Everything below is about getting as close to a causal answer as you can with the evidence available — and about noticing failure early enough to change course.

This page is about knowing the truth for yourself. If you need to present that truth to someone paying for it, how to report link building results to a client covers the packaging.

The three-tier metric hierarchy

Most link building measurement fails because it mixes tiers and then wonders why the numbers disagree. Separate them and each one answers a different question.

Tier	Metric	Question it answers	Moves in
1 · Output	Net new referring domains	Did we actually acquire links?	2–6 weeks
1 · Output	Placement quality and relevance	Are they links worth having?	Immediately
2 · Leading	Impressions, average position (target terms)	Is Google re-evaluating these pages?	1–3 months
2 · Leading	Position for the specific target keywords	Are we climbing?	2–6 months
3 · Outcome	Organic sessions to target pages	Is the climb producing traffic?	3–9 months
3 · Outcome	Conversions, signups, revenue	Was any of it worth doing?	4–12 months

The hierarchy runs one way. Tier 1 without Tier 2 means you're acquiring links that don't count. Tier 2 without Tier 3 means you're ranking for terms nobody valuable searches. Tier 3 is the only tier that justifies the budget, and it's the slowest and noisiest — which is exactly why you need the other two.

Metrics that matter, and the ones that don't

Track these:

Net new referring domains. New minus lost, on the pages you're targeting. The single most honest measure of link acquisition. Not total backlinks — one sitewide footer placement can add tens of thousands of "backlinks" and change nothing. The distinction is covered in referring domains vs backlinks.
Impressions and average position in Search Console, filtered to the target pages and terms. The most underused signal in link building. A page climbing from position 40 to 18 produces almost no traffic change but shows up clearly here, months before the rankings table looks different.
Links still live. A link you won in March and lost in June was never worth what you counted it for. Do backlinks expire covers the failure modes; re-verification belongs in your measurement, not just your audit.
Referral traffic from placements. Usually small. Occasionally the whole point — a link that sends twenty qualified visitors a month is doing a job whether or not Google noticed it.
Organic conversions on target pages. The number that ends the argument.

Stop tracking these as KPIs:

Total backlink count. Trivially inflated, universally recognised as a vanity metric.
Domain Authority or Domain Rating as a headline. Third-party estimates Google doesn't use, moving on their own recalculation schedule for reasons unrelated to your work. Useful as a rough per-placement filter, useless as a goal. See domain authority vs domain rating.
Sitewide organic traffic. Moves for a dozen reasons that have nothing to do with your links, in both directions. It will convince you a campaign worked when a seasonal spike happened, and that it failed when an unrelated page decayed.
Emails sent, prospects contacted. Effort, not progress. Worth logging so a slow month is explicable; never worth celebrating.

Set the baseline before you start

The most common reason people can't tell whether link building worked is that they never wrote down where they started. Before the first outreach email, record:

Referring domains — sitewide and for each target page.
Current position for every target keyword, dated.
Search Console impressions and average position for those terms over the previous 90 days.
Organic sessions and conversions for each target page, previous 90 days.
What else is changing — a redesign, a migration, a content refresh, a Google update. This is the list that saves you in month five when something moves and everyone claims credit.

Point 5 is the one people skip and the one that decides whether your measurement means anything. If you rebuilt the site the same quarter you started link building, you have permanently lost the ability to attribute either.

Getting closer to causality

You can't run a true experiment, but you can stop guessing:

Use target-page cohorts. Pick the pages you're building links to and track them as a group against a comparable group you're deliberately not building links to. Same site, same publishing cadence, same Google updates. If the linked cohort pulls away from the unlinked one, that's the strongest evidence a single site can generate. It isn't a randomised trial — you chose the pages, and that choice carries bias — but it removes most of the sitewide noise.

Stagger the work. Building to five pages at once and seeing everything move tells you less than building to two pages in Q1 and three in Q2. Staggering gives each cohort its own before-and-after.

Date every link. You cannot correlate what you cannot place in time. Record the date each link went live, not the date you won the agreement — the clock starts when the page is live and indexed.

Check the linking page is actually indexed. An unindexed linking page passes nothing, and a campaign full of them looks identical to a campaign that isn't working.

When to judge, and what "not working" looks like

Links have to be crawled, indexed and processed before they influence anything, and on competitive terms the visible effect commonly takes three to six months — how long backlinks take to work covers the mechanics. Judging at 30 days is measuring noise.

At the six-month mark, four failure signatures are worth telling apart, because they need opposite responses:

Nothing acquired. Tier 1 flat. Not a link problem — an outreach or offer problem. Fix the pitch, the asset, or the prospect list before touching anything else.
Links acquired, no movement anywhere. Tier 1 up, Tier 2 flat. Usually the links are low quality, unindexed, nofollowed, or from sites with no topical relationship to yours. Audit what you actually got against what makes a bad backlink.
Impressions up, positions flat. Google is re-evaluating but you haven't cleared the competitive bar yet. This is the most misread signature: it means keep going, and it's the one most people quit on.
Rankings up, conversions flat. The links worked and the keyword choice didn't. A content and targeting problem wearing a link building costume.

Only the second signature means the link building itself is broken.

Where the honest answer usually lands

For most sites the finding is that links are being acquired too slowly to outpace competitors who are also acquiring them. That's not a measurement problem, it's a supply problem — and it's the one thing the dashboard can't fix.

Backlinkster addresses supply directly: site owners in related niches trade one-for-one in-content links, and every placement is verified live by code, so the "still live?" column stays accurate on its own instead of decaying between audits. Five swaps a month on the free plan; see the plans.

Frequently asked questions

How do I know if my link building is working? Track three tiers in order: net new referring domains, then Search Console impressions and average position for your target terms, then organic sessions and conversions on the target pages. Tier one should move within weeks, tier two within months, tier three last. Movement stopping at one tier tells you where the problem is.

What link building metrics actually matter? Net new referring domains, placement quality and relevance, whether links are still live and indexed, impressions and average position for target terms, and conversions on target pages. Total backlink count and headline Domain Authority are vanity metrics — inflatable, third-party, and unrelated to what Google uses.

How long before I can tell if link building is working? Expect three to six months for ranking movement on competitive terms, longer in hard niches. Link acquisition itself should be visible within weeks. If nothing has moved at all by month six — not even impressions — investigate rather than wait.

Why are my rankings flat after building lots of links? Most often the links don't count: unindexed linking pages, nofollowed placements, sites with no topical relationship, or links bought in bulk from sites that exist to sell them. Check quality and indexation before adding volume. Rising impressions with flat positions is a different case and usually means keep going.

How do I prove links caused a ranking improvement? On a single site you can't prove it outright. You get close by tracking a linked cohort of pages against a comparable unlinked cohort, staggering campaigns so each has its own before-and-after, dating every placement, and recording every other change to the site so you can rule alternatives out.

Should I track Domain Authority to measure progress? No. It's a third-party estimate Google doesn't use, it's recalculated on a vendor's schedule, and once it becomes a goal you'll optimise for the score rather than the outcome. Use it as a rough filter when judging an individual prospect, never as a KPI.

The bottom line

Measuring link building well is mostly a discipline problem, not a tooling one. Write down the baseline before you start, separate output from leading indicators from outcomes, date everything, and keep a list of what else changed. Then give it two quarters before you draw conclusions — and when you do, read which tier stalled, because that tells you whether to fix the outreach, the link quality, or the keyword you picked in the first place.