kevin Indig - Internal Link Building on Steroids (Tech SEO Boost )
- 2. Kevin Indig | @Kevin_Indig | #TechSEOBoost
Kevin Indig
Internal link building on steroids
… or how to flatten power curves
- 4. Kevin Indig | @Kevin_Indig | #TechSEOBoost
Power curves 101
- Vilfredo Pareto: 80/20
- People + riches
- Factors + impact
- Startups + VC returns
- 5. Kevin Indig | @Kevin_Indig | #TechSEOBoost
0
50,000
100,000
150,000
200,000
250,000
300,000
1
7
13
19
25
31
37
43
49
55
61
67
73
79
85
91
97
103
109
115
121
127
133
139
145
151
157
163
169
175
181
187
193
199
205
211
217
223
229
235
241
247
253
259
265
271
277
283
289
295
301
307
313
319
Clicks
Number of URLs
KEYWORD CLICK CURVE
- 6. Kevin Indig | @Kevin_Indig | #TechSEOBoost
0
100
200
300
400
500
600
1
12
23
34
45
56
67
78
89
100
111
122
133
144
155
166
177
188
199
210
221
232
243
254
265
276
287
298
309
320
331
342
353
364
375
386
397
408
419
430
441
452
463
474
485
496
507
518
529
540
551
562
573
584
595
606
617
628
639
650
661
672
683
694
705
716
Linkingdomains
Number of URLs
DOMAIN POP DISTRIBUTION
- 7. Kevin Indig | @Kevin_Indig | #TechSEOBoost
0
50
100
150
200
250
1
67
133
199
265
331
397
463
529
595
661
727
793
859
925
991
1057
1123
1189
1255
1321
1387
1453
1519
1585
1651
1717
1783
1849
1915
1981
2047
2113
2179
2245
2311
2377
2443
2509
2575
2641
2707
2773
2839
2905
2971
3037
3103
3169
3235
3301
3367
3433
3499
3565
3631
3697
3763
3829
3895
3961
4027
4093
4159
4225
4291
Monthlycrawls
Number of URLs
CRAWL RATE DISTRIBUTION
- 8. Kevin Indig | @Kevin_Indig | #TechSEOBoost
Why is that important?
- 9. Kevin Indig | @Kevin_Indig | #TechSEOBoost
SEO is not getting easier
- 10. Kevin Indig | @Kevin_Indig | #TechSEOBoost
Internal linking = one of the
strongest levers
- 11. Kevin Indig | @Kevin_Indig | #TechSEOBoost
“Use the Powa of
Indernal Lings”
- 12. Kevin Indig | @Kevin_Indig | #TechSEOBoost
www.kevin-indig.com
Tech SEO Lead @ Atlassian
Mentor @ German Accelerator
- 13. Kevin Indig | @Kevin_Indig | #TechSEOBoost
Q: How can we optimize
internal linking?
- 14. Kevin Indig | @Kevin_Indig | #TechSEOBoost
Crawl + calculate internal PR
- 15. Kevin Indig | @Kevin_Indig | #TechSEOBoost
Use tools to get recommendations
- 16. Kevin Indig | @Kevin_Indig | #TechSEOBoost
Problem: Most internal link models are inaccurate!
- 17. Kevin Indig | @Kevin_Indig | #TechSEOBoost
PageRank exists between and within sites
- 18. Kevin Indig | @Kevin_Indig | #TechSEOBoost
Internal PageRank is only half of the equation
Page A Page B
Page C
Page D
Site A
- 19. Kevin Indig | @Kevin_Indig | #TechSEOBoost
External PageRank is the other side of the equation
Page A Page B
Page C
Page D
Site A Site B
Page A Page B
Page C
Page D
- 20. Kevin Indig | @Kevin_Indig | #TechSEOBoost
What we need is a model that
combines internal and external PR
- 21. Kevin Indig | @Kevin_Indig | #TechSEOBoost
Solution: the “True Internal PR” model (TIPR)
CheiRank Backlinks
Log files
TIPR
PageRank
- 22. Kevin Indig | @Kevin_Indig | #TechSEOBoost
What can you do with TIPR?
Calculate
“accurate” internal
PageRank
Identify technical
problems
Monitor
optimization
progress
- 23. Kevin Indig | @Kevin_Indig | #TechSEOBoost
The TIPR process
Analysis Recommendations Monitoring
- 24. Kevin Indig | @Kevin_Indig | #TechSEOBoost
TIPR – step by step
1. Crawl site
2. Calculate internal PR and CR
3. Add backlinks to get “true internal PR”
4. Add crawl rate from log files to understand impact of (internal + external)
links over time
5. Sort and rank metrics
6. Optimize for Money Maker Pages
- 25. Kevin Indig | @Kevin_Indig | #TechSEOBoost
“Robin Hood” principle: take from
the rich, give to the poor
- 26. Kevin Indig | @Kevin_Indig | #TechSEOBoost
Let’s talk some results
- 27. Kevin Indig | @Kevin_Indig | #TechSEOBoost
Dry testing the model at small scale
- 30. Kevin Indig | @Kevin_Indig | #TechSEOBoost
Crawl to find PR and CR
- 32. Kevin Indig | @Kevin_Indig | #TechSEOBoost
Exported server log files
and tracked keywords
- 33. Kevin Indig | @Kevin_Indig | #TechSEOBoost
1710
1033
96
3 1
0
200
400
600
800
1000
1200
1400
1600
1800
1 5 20 50 100
NumberofURLs
Crawl frequency
CRAWL FREQUENCY DISTRIBUTION
- 34. Kevin Indig | @Kevin_Indig | #TechSEOBoost
1710
1033
96
3 1
0
200
400
600
800
1000
1200
1400
1600
1800
1 5 20 50 100
NumberofURLs
Crawl frequency
CRAWL FREQUENCY DISTRIBUTION
Guess who this lone fella is?
- 35. Kevin Indig | @Kevin_Indig | #TechSEOBoost
19
2
16
401
519
584
921
172
289
10
4129
10
69
11
99
6
222120
2
60
115
150
41
17
25
30
69331211254
39
311253341
30
211212124334122
30
111111111111141111112111111111111111111111111111111111112131111111112221112115234111
0
100
200
300
400
500
600
700
800
900
1000
0
5
7
9
11
13
15
17
19
21
23
25
27
29
31
33
35
37
39
41
43
45
48
50
52
54
56
58
61
63
65
70
72
75
79
81
83
86
88
91
94
101
108
110
114
120
126
131
134
143
154
161
180
190
200
216
237
274
290
319
323
326
331
335
344
352
369
375
377
379
603
706
1009
2989
2991
2995
Numberofinlinks
Number of URLs
INLINKS PER URL
- 36. Kevin Indig | @Kevin_Indig | #TechSEOBoost
19
2
16
401
519
584
921
172
289
10
4129
10
69
11
99
6
222120
2
60
115
150
41
17
25
30
69331211254
39
311253341
30
211212124334122
30
111111111111141111112111111111111111111111111111111111112131111111112221112115234111
0
100
200
300
400
500
600
700
800
900
1000
0
5
7
9
11
13
15
17
19
21
23
25
27
29
31
33
35
37
39
41
43
45
48
50
52
54
56
58
61
63
65
70
72
75
79
81
83
86
88
91
94
101
108
110
114
120
126
131
134
143
154
161
180
190
200
216
237
274
290
319
323
326
331
335
344
352
369
375
377
379
603
706
1009
2989
2991
2995
Numberofinlinks
Number of URLs
INLINKS PER URL
- 37. Kevin Indig | @Kevin_Indig | #TechSEOBoost
0
50
100
150
200
250
300
350
400
450
0
6
8
10
12
14
16
18
20
22
24
26
28
30
32
34
36
38
40
42
44
46
48
50
52
54
56
58
60
62
64
66
68
70
72
75
77
79
81
83
85
87
89
93
95
97
99
102
105
109
113
119
123
130
142
146
151
155
168
170
183
189
206
235
242
362
Numberofoutgoinglinks
Number of URLs
OUTGOING INTERNAL LINKS PER URL
- 38. Kevin Indig | @Kevin_Indig | #TechSEOBoost
0
100
200
300
400
500
600
700
800
900
1000 0
5
7
9
11
13
15
17
19
21
23
25
27
29
31
33
35
37
39
41
43
45
48
50
52
54
56
58
61
63
65
70
72
75
79
81
83
86
88
91
94
101
108
110
114
120
126
131
134
143
154
161
180
190
200
216
237
274
290
319
323
326
331
335
344
352
NumberofInlinks/Outlinks
Number of URLs
INCOMING VS. OUTGOING INTERNAL LINKS
Inlinks Outlinks
- 39. Kevin Indig | @Kevin_Indig | #TechSEOBoost
0
100
200
300
400
500
600
700
800
900
1000 0
5
7
9
11
13
15
17
19
21
23
25
27
29
31
33
35
37
39
41
43
45
48
50
52
54
56
58
61
63
65
70
72
75
79
81
83
86
88
91
94
101
108
110
114
120
126
131
134
143
154
161
180
190
200
216
237
274
290
319
323
326
331
335
344
352
NumberofInlinks/Outlinks
Number of URLs
INCOMING VS. OUTGOING INTERNAL LINKS
Inlinks Outlinks
Optimum
- 40. Kevin Indig | @Kevin_Indig | #TechSEOBoost
How do we flatten this curve?
- 41. Kevin Indig | @Kevin_Indig | #TechSEOBoost
So, what do we do with this information?
URL Crawl frequency Domain pop PageRank CheiRank
/URL1 200 300 0.0810 0.3555
/URL2 150 200 0.0300 0.3422
/URL3 300 100 0.0690 0.3000
/URL4 50 50 0.0220 0.2908
- 42. Kevin Indig | @Kevin_Indig | #TechSEOBoost
Rank, take average, re-sort
URL Crawl frequency Domain pop PageRank CheiRank Average
/URL1 2 1 1 1 1.25
/URL2 3 2 3 2 2.5
/URL3 1 3 2 3 2.25
/URL4 4 4 4 4 4
- 43. Kevin Indig | @Kevin_Indig | #TechSEOBoost
Rank, take average, re-sort
URL Crawl frequency Domain pop PageRank CheiRank Average
/URL1 2 1 1 1 1.25
/URL3 1 3 2 3 2.25
/URL2 3 2 3 2 2.5
/URL4 4 4 4 4 4
- 44. Kevin Indig | @Kevin_Indig | #TechSEOBoost
Look for pattern in URLs and optimize accordingly
3.99000 0.03000 0.00972 0.02000 0.01000
0.03000 0.03000
0.07000
0.01000 0.00000
0.00000
0.05000
0.10000
0.15000
0.20000
0.25000
0.30000
0.35000
0.40000
0.45000
Categories Apps Add-ons Vendors Plugins
Average PageRank and CheiRank by directory
PageRank CheiRank
- 46. Kevin Indig | @Kevin_Indig | #TechSEOBoost
2,958 incoming links
1,094 outgoing links46 outgoing links
12 incoming links
- 47. Kevin Indig | @Kevin_Indig | #TechSEOBoost
What happened when we rolled out
the changes?
- 50. Kevin Indig | @Kevin_Indig | #TechSEOBoost
Recap: TIPR
Crawl PR + CR Backlinks Log files
Power
curves
- 51. Kevin Indig | @Kevin_Indig | #TechSEOBoost
More lessons
Robots.txt XML sitemaps 404 errors
- 52. Kevin Indig | @Kevin_Indig | #TechSEOBoost
Limitations of the model
• Way more ranking factors than PageRank
• Only suitable for a certain size of sites
• Just tested on a few sites (yet)
• Still trying to find the right weighting
- 53. Kevin Indig | @Kevin_Indig | #TechSEOBoost
Taking the concept one step further
• Automating the model
• Predicting success with staging environments
- 54. Kevin Indig | @Kevin_Indig | #TechSEOBoost
Thanks for your attention
Thanks to Catalyst, Audisto, and Nozzle.
@Kevin_Indig
www.kevin-indig.com
Editor's Notes
- I should have called this presentation: “lessons of power curves in SEO”.
- 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
- We see that principle everywhere in SEO: a few keywords bring in the most traffic…
- A few pages receive the most backlinks
- - A few pages get crawled the most
- It’s important to understand what moves the needle in SEO nowadays because …
- … SEO is not getting easier!
- We need to find more efficient tactics to not waste time and energy.
- 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…
- - Question for the crowd:
- The problem with these oldschool approaches is that they’re inaccurate.
Let me explain...
- 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
- We often calculate internal PR for internal link optimization
But for the full picture, we need to take backlinks into account…
- When we add links from other sites into the model, the whole PageRank equation changes!
So, how do we solve this issue?...
- 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
- 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
- Identify pages with high PageRank (internal + external) and give to “MoneyMaker” pages with lower PR
This is where we get back to power curves
- - First, I tried the model on a smaller site (~3,000 pages)
- Simple model: pulled crawl frequency, traffic, links, and other metrics
Flattened the curve by linking from strong pages to weaker ones
- - +160% organic traffic over time within 15 months
- - Site with roughly 40K pages
- Crawled site with ~40,000 URLs
Thanks to Audisto for providing the data here.
German engineering!
- 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)
- Thanks to Nozzle, for providing keyword data to better understand impact of TIPR
Nozzle.io
- - Some observations: power curve for crawl frequency
- Anybody have a guess? A single file that’s being crawled double as much as any other file?
> Robots.txt
- Majority of URLs receive 0-10 links
- Notice the perfect Power curve!
- - Outgoing internal links are much more spread out but still unevenly distributed
- - Comparison: should be much more evenly spread.
- - Comparison: should be much more evenly spread.
- Crawl frequency is monthly
Dummy data
- -
- - Now we have a true view of which pages are strong and which are weak
- 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
- This guy had just 46 outgoing links
When we changed that…
- - We saw our crawl rate go up
- - We saw traffic go up
- 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
- How thinks there is one?
How thinks there are two?
- 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…
- First, SEO becomes a winner takes it all game.