Transcript
Intro
A patent and trademark firm went from about four hundred visitors a month to nearly four thousand, in under three months. That's plus eight hundred and thirty percent. A result like that deserves to be measured so clearly that nobody can argue with it, and that's today's job.
That one was a rescue. A previous rebuild had left the old addresses dead, so we redirected every one to its right new page, brought back the English pages and made the site fast. Clear work, clear result. Off-page work needs a method of its own, because links act slowly and quietly, and you want every win they bring on the record.
So today: how we measure off-page work and show the wins, to clients and to you. Real Search Console data from a dental clinic, a restaurant and an electronics spare-parts store, read the way we read it every month.
Quick reminder of where we stand. We're not the links-don't-matter crowd. Off-page work is part of the campaigns we run, and in many industries it moves results. That's exactly why it needs careful measurement: when links work you want to see it, and when a link gets ignored, nobody tells you.
Since December 2022, Google says its spam system neutralises unnatural links, and any credit they passed is simply lost. Nobody sends you a message when that happens. The rankings just don't move, and your own data is the only place you'll ever notice.
Search Console first
Rule one: Search Console first, third-party tools second. Search Console counts real impressions and real clicks from Google. Tool traffic numbers are estimates: keywords you rank for, times their search volume, times a guessed click rate, added up. Great for sizing up a competitor. It isn't your traffic.
Search Console has blind spots of its own. Google hides queries that only a handful of people search, to protect their privacy. For our dental clinic, those hidden queries brought fifty nine percent of September's clicks. So when the query table doesn't add up to the total, nothing is broken. Judge on page and property totals, and use the query table for direction.
Two more traps. Google marks the newest days as preliminary, so a report pulled this morning can end in a dip that isn't real. And months have different lengths, and the current month is never finished. Compare clicks per day, and compare full months with full months. Half a month against a whole one will flatter you or scare you, and either way it's wrong.
One more filter if you serve several countries or languages. Our dental clinic treats patients from abroad, so links pointing at one language's pages should show up first in the countries that speak it. Split by country and page together, or a gain in one market drowns in the noise of another.
Read it in order
Now the order. Filter Search Console to the pages your links point at, not the whole site. Read impressions first, because they move earliest: Google is showing the page for more searches. Then clicks. Position comes last, because a site that starts showing up for lots of broad new searches can see its average position sink while its traffic climbs.
Here's that order on a real win. Our restaurant client, from the second half of June to September: impressions per day grew about three point seven times, and clicks per day rose seventy percent. Reach came first, and the customers came right behind it.
Second split: brand against non-brand. Off-page work should mostly show up in searches that don't contain your name, because people who already know the restaurant find it anyway. From July to September, the restaurant's non-brand impressions grew about four times, and non-brand clicks about two point six times. That's where the new customers came from.
Third, give yourself a reference. Pick a few similar pages on the same site that got no new links, and read them the same way over the same weeks. When the linked pages pull ahead of the untouched ones, that gap is your evidence that the links are working.
Timing and everything else
Next, timing. Google says crawling alone can take anywhere from a few days to a few weeks. A new link has to be found, the page carrying it crawled and indexed, and only then can its value pass on. So we judge off-page work over months, on review dates everyone agreed before the campaign started.
Here's the dental clinic, month by month, in clicks per day. Before us, it sat around nineteen and a half. We replaced machine-translated pages in eleven languages with proper translations, moved every language to clean addresses with redirects and hreflang, and cleaned up a messy link profile. Off-page work started in July. By September it was averaging nearly thirty five a day.
Put the work on the timeline and the story reads itself. Translations, clean addresses and the link cleanup land in June, off-page work starts in July, and every month after that beats the one before. Twenty point nine, twenty three point two, twenty six point one, then thirty four point eight a day in September.
The start date is the other place a report can mislead. Picture ReelRadar's team starting a link campaign on a site that was already climbing for months. Credit the campaign with growth from last winter, and it takes credit for work it never touched. Our rule: every number is measured from when our work started, a real dated day before it against a real dated day after.
Here's how we report it to clients. Links work as part of a programme, together with technical fixes and content, so we report the programme's result: the whole lift, measured from the day we started. Across many real campaigns we've watched off-page work move results, and that's our field work.
Keep a change log. Every technical fix, every content release, every link that went live, each with a date, plus Google's update dates from its Search Status Dashboard. When a chart moves, the log shows which piece of work it lines up with, so every win in the report has a date and a name.
Check the links themselves
Before you credit any link, check the link itself. Is it still on the page, and is that page indexed? Is the link followed, not nofollow or sponsored? And does it point at the page whose numbers you're reading? A link that fails any of these can't explain your chart.
The indexing check is the one everybody skips. If the page carrying your link isn't in Google's index, Google may never have seen the link at all. A link on an orphan page nobody indexed can sit there for months, doing nothing, looking great in a spreadsheet.
And don't treat Search Console's Links report as the full answer. Google says it isn't a complete list, can include links that have since been removed, and doesn't show which ones are nofollow. Keep your own record of every placement and when you last saw it live. A link can pass every check and still count for nothing, so the traffic data gets the final word.
Report the win
Last, the report. Our rule: lead with a real day from before the work and a real day after it, both dated, then put the averages beside them. For the clinic: on the twenty seventh of May, before we started, it got eight clicks. On the seventeenth of September, fifty eight. That's about seven times. September averaged thirty four point eight a day, up seventy seven percent, and impressions per day were up one hundred and twenty three percent.
Here are three clients, measured the same way: a real day before our work against a real day after it. The clinic went from eight clicks on the twenty seventh of May to fifty eight on the seventeenth of September. The restaurant, from twenty seven on the fourteenth of June to seventy on the eleventh of September. And the spare parts store, from forty on the ninth of August to two hundred and four on the sixteenth of October.
And the store. On the ninth of August, two weeks before we started, it got forty clicks. On the sixteenth of October, two hundred and four. That's five times. The first half of October has been running at around one hundred and fifty three a day.
Here's how reach turns into customers. In August, the restaurant's impressions per day grew seventy two percent on July, and the number of searches it appeared for grew forty three percent. Clicks rose eleven percent, and September added another twelve percent on top. Reach arrives first, and the customers follow it.
A report clients trust has a short list of parts. The baseline, and where it came from. A real day before the work and a real day after it, both dated, with the averages beside them. Impressions and clicks per day, brand and non-brand kept apart. And the change log, so everyone can see what shipped and when.
Verdict
So let's rate the claim behind most link case studies. Rankings moved the week after the links went in, so the links did it.
Placebo. Google says crawling alone takes days to weeks, so a jump one week later usually has another cause. Picture ReelRadar's new links going live the same week as a site redesign, with a Google update a few days later. Timing is a clue. It's never proof.
So what does real link evidence look like? The linked pages pull ahead of similar pages that got no links, in a stretch with no other big release, and the same lift shows up again on the next campaign. That's the pattern we look for in our field work.
Wrap up
Here's the whole method on one screen. Search Console first, estimates second. Impressions, then clicks, then position, for the pages the links point at, brand and non-brand apart. Give it months, check the links are live and indexed, and report a dated day before and a dated day after, with the averages beside them.
The link campaign measurement sheet is free on the episode page at irongoo.com, with the checks, the reading order and a report template with the before and after days and the averages already laid out. And next time someone shows you a case study chart, ask them one thing: what else happened that month?
