kevin Indig - Internal Link Building on Steroids (Tech SEO Boost )

Editor's Notes

1. I should have called this presentation: “lessons of power curves in SEO”.
2. Power curves: a universal principle; also called “power laws” or “pareto 80/20 principle” Vilfredo Pareto Say hat a few minority has a big impact on a majority. We see that principle everywhere in SEO
3. We see that principle everywhere in SEO: a few keywords bring in the most traffic…
4. A few pages receive the most backlinks
5. - A few pages get crawled the most - It’s important to understand what moves the needle in SEO nowadays because …
6. … SEO is not getting easier! - We need to find more efficient tactics to not waste time and energy.
7. References of all bible verses -> idea of internal linking not new -> bar chart = chapters, length = number of verses Or, as my homie Arnie would say…
8. - Question for the crowd:
9. The problem with these oldschool approaches is that they’re inaccurate. Let me explain...
10. But all of our internal PageRank and internal link optimization models forget external PageRank. Most internal link models are just missing one crucial component: backlinks
11. We often calculate internal PR for internal link optimization But for the full picture, we need to take backlinks into account…
12. When we add links from other sites into the model, the whole PageRank equation changes! So, how do we solve this issue?...
13. I created a model called “TIPR” to solve this issue. It combines PageRank, CheiRank, Backlinks, and Log Files PageRank and CheiRank to create the internal link graph Backlinks to make the internal link graph accurate Log files to monitor changes; much better than rankings or organic traffic High correlation between crawl frequency and rankings
14. Primary Goal: Calculate ”real” internal PageRank with crawl data, backlinks, log files to optimize internal linking accurately Secondary Goal: Monitor crawl rate to identify technical problems Tertiary goal: Understand what impacts crawl rate and therefore leads to better rankings Let’s go through this for a minute
15. Identify pages with high PageRank (internal + external) and give to “MoneyMaker” pages with lower PR This is where we get back to power curves
16. - First, I tried the model on a smaller site (~3,000 pages)
17. Simple model: pulled crawl frequency, traffic, links, and other metrics Flattened the curve by linking from strong pages to weaker ones
18. - +160% organic traffic over time within 15 months
19. - Site with roughly 40K pages
20. Crawled site with ~40,000 URLs Thanks to Audisto for providing the data here. German engineering!
21. Used AHREFs for this, but you can also use other tools or even semrush Feature: most linked pages Then you can either use URL rating (prop. Metric) or domain pop (which I found to correlate heavily with most prop. metrics)
22. Thanks to Nozzle, for providing keyword data to better understand impact of TIPR Nozzle.io
23. - Some observations: power curve for crawl frequency
24. Anybody have a guess? A single file that’s being crawled double as much as any other file? > Robots.txt
25. Majority of URLs receive 0-10 links
26. Notice the perfect Power curve!
27. - Outgoing internal links are much more spread out but still unevenly distributed
28. - Comparison: should be much more evenly spread.
29. - Comparison: should be much more evenly spread.
30. Crawl frequency is monthly Dummy data
31. -
32. - Now we have a true view of which pages are strong and which are weak
33. This is the average PR and CR per directory We clearly see that the categories directory has way more PageRank than others “Addons” has much higher CheiRakn than other directories
34. This guy had just 46 outgoing links When we changed that…
35. - We saw our crawl rate go up
36. - We saw traffic go up
37. Robots.txt is the most crawled url across the board (probably depends on change rate) Google spends longest time on xml sitemaps Finding a set of 404s on one of our site, Google reacted with a reduction in crawl rate and ranking across the board
38. How thinks there is one? How thinks there are two?
39. Google transitions from search to discovery engine. For its 20th anniversary, Google announced It will focus on user journeys and recommendations. Big shift! What does that mean for SEOs and webmasters? Two things…
40. First, SEO becomes a winner takes it all game.