LearningLeader.com is Ryan Hawk's leadership podcast: 700 plus episodes, a decade of guests, and a back catalogue most listeners never found. We spent 24 hours on the plumbing. Ten weeks later his search traffic had doubled, it held after his book launch faded, and the searches carrying it are not his name.
Client
Ryan Hawk, The Learning Leader Show
Work shipped
July 9 to 13, 2026
Measured through
September 20, 2026
Source
Google Search Console
The result
Clicks per day
51→101
+97%
Impressions per day
9,560→20,872
+118%
Non-brand clicks per day
8→37
+359%
Non-brand impressions per day
3,522→12,429
+253%
Daily averages. Before is June 29 to July 8, the ten days of history we have prior to the work. After is August 3 to September 20, seven weeks that deliberately start once the July 21 book launch spike had already decayed. Non-brand means every search that is not his name, the show's name, or the book's title.
Search impressions per day, June 29 to September 20. The green band is the week the work went live. The clay band is the book launch. The line does not spike and settle back, which is what a launch looks like. It steps up and keeps climbing eight weeks after the launch finished, which is what consolidation and crawl work looks like.
The obvious objection
It was not the book
Ryan launched a book on July 21, three weeks into the measurement window. Any honest read of these numbers has to deal with that, so we split every search into brand and non-brand and waited for the spike to decay before drawing the comparison.
Brand searches are now below where they started. 11 clicks a day before the work, 10 a day in the four weeks to September 20. Whatever the book did, it is finished, and the traffic did not leave with it.
Non-brand searches went from 8 clicks a day to 37. Brand fell over the same period, so the growth is all on the side of people who were not looking for Ryan.
The launch peak averaged 15,478 impressions a day. The ten days to September 20 averaged 23,686, or 21,160 once the noise page described further down is removed. Either way it is above the peak, eight weeks after it.
Traffic that was really a launch is back near baseline by now. His brand traffic did exactly that. The rest did not, and the impression side is still climbing.
The clean proof
One page, one redirect, 29 times the clicks
Episode 537 is Aron Ralston, the hiker from 127 Hours who cut off his own arm. It existed at two addresses, one of 543 duplicate pairs on the site. We pointed the duplicate at the real page and changed nothing else about it.
Both URLs are counted together, so this is not one page borrowing from the other.
The searches are aron ralston, 127 hours true story, 127 hours real guy. Nothing to do with a leadership book.
Its average position improved while its reach grew nine times over, so this is not a page showing up more often in worse places.
Being straight about it: that one page is 23% of the clicks and 25% of the impressions in the after window. Take it out entirely, along with the noise page named further down, and the rest of the site is still up 55% on clicks and 61% on impressions.
The mechanism
Fewer addresses, more traffic
This is the number that explains the rest. Two matched ten-day windows, one before the work and one at the end of August:
Ten-day window
Before
After
Distinct URLs earning impressions
915
785
Impressions those URLs earned
101,165
209,994
Share landing on the real episode page
18%
62%
Google is showing fewer of Ryan's pages and sending twice the traffic. That is consolidation doing exactly what it should: 469 duplicate addresses stopped competing with the pages they were copies of, and 311 real episode pages started surfacing in their place. A visit that used to land on a dead-end copy now lands on the episode itself, with the player, the show notes and the links on it.
It has held. In the ten days to September 20 the same count reads 819 addresses, still well below the 915 before the work, and they earned 240,952 impressions.
The work
What 24 hours bought
All of it additive and reversible, none of it visible to his readers, and nothing risky went near the launch.
543 duplicate episode addresses consolidated to one clean URL each.
545 older episodes given a real headline. They had none, so Google was inventing one.
Structured data for Ryan, the show and every episode, self-updating as he publishes.
74 high-value pages given rewritten search titles and descriptions.
266 "keep listening" links across 51 pages, so a reader who finds one episode finds the next.
A searchable all-episodes hub at /episodes/, self-updating.
A safe speed pass. Image compression, caching, a font-load fix. Mobile PageSpeed came off a flat 53.
Guests who now bring their own audience
Searches that earned Ryan nothing in the ten days before the work, and what they have earned since it shipped, July 14 to September 20:
Search
Clicks before
Clicks since
jack raines
0
46
chris borland
0
31
maurice clarett
0
25
matt eversmann
0
18
jesse itzler net worth
0
17
will guidara
0
16
Where we are honest
What this does not prove
The baseline is ten days. Ryan's Search Console property was created around June 29 and Google does not backfill history before that. Ten days is enough to see a shift clearly. It is not a long runway, and we will not pretend otherwise.
Marquee guest names still do not convert. A search for katie gatti tassin puts his episode at position 1.9. Google showed it 930 times in the four weeks to September 20 and it earned one click. Those people want the guest, not an interview about the guest, and no title we write fixes that. We would rather say it now than have it found later.
Part of the September reach is noise. Since September 1 one page, /podcast/, has been surfacing for generic searches about learning: machine learning, learning through play. That is 63,152 impressions in the four weeks to September 20 and 75 clicks. We left it in the totals at the top because it is what Search Console shows. Strip it out and impressions per day read 19,532 instead of 20,872, still up 104%, and the clicks lose about three a day out of 101.
Clicks grew slower than impressions. That is what happens when a 700-episode back catalogue starts surfacing around position 9. The reach came first. Turning more of it into clicks is the next body of work, not a finished one.
We report the work delivered and the measured change, never predicted rankings. Search results move for reasons no agency controls.
How we measure
The same four rules on every site
Everything above comes out of one place, and none of it is our own dashboard.
Your Search Console, not our tool. You own the property, you see the same numbers we do, and you can check any figure in this document yourself.
A before window we locked before touching anything. Whatever the site was doing in the weeks before the work is the line we are measured against, spikes and all.
Brand split out from the rest. Searches for your own name were coming anyway. The number that counts is people who found you without knowing you existed.
A one-page report every month, including what it cannot prove. What we changed, what those pages did next, and the parts that moved for reasons we did not cause. Then a call to go through it together.
Since then
What we built next
A sample chapter page for the book, live August 11, as the lead magnet for his email list.
Rebuilt signup forms. The popup went from 1,119px to 630px on an iPhone, so it fits without scrolling, and a bug that showed the form twice on every post is gone.
An AI assistant on the episodes page, live September 1. Ask it a question and it answers from the show's transcripts, with every episode it cites linked back to the real page.
He put us on a monthly plan in September. Every episode in the archive gets the same treatment by mid-November: the transcript on the page, the questions people actually ask about that guest, and the answers in the guest's own words.