16. #theinbounder
We expected that we could improve traffic by 10%
and conversion by 15% MoM.
0,00%
0,20%
0,40%
0,60%
0,80%
1,00%
1,20%
0
200.000
400.000
600.000
800.000
1.000.000
1.200.000
Jan 16 - 31 Feb-17 Mar-17 Apr-17 May-17 Jun-17 Jul-17 Aug-17 Sep-17 Oct-17 Nov-17 Dec-17
UGHH Projections
Site Visits Orders Average Order Receipts - Existing Sources Receipts - New Sources Total Receipts Conversion Rate
17. #theinbounder
We added TF-IDF optimized copy to the category
pages.
#theinbounder
We did the audience research to figure out who our
addressable and aspirational segments are
18. #theinbounder
We added TF-IDF optimized copy to the category
pages.We did the journey mapping to figure out their need
states and emotional experiences.
20. #theinbounder
We audited the content to determine the weaknesses
and built a strategy to go forward
We audited the analytics to improve our ability to
measure against our goals.
28. #theinbounder
We learned this at launch, so I angrily placed
Shopify on a subdomain.
www.undergroundhiphop.com Reverse proxy Shopify.undergroundhiphop.com
41. #theinbounder
Here’s a big picture look at what it’s like to fail
hard on an e-commerce site.
Best Record Store Day
of all time
When that super
good deal kicked in
Magazine investment
54. #theinbounder
Instead, I should have put the magazine on a
subdomain and tested content in isolation.
magazine.undergroundhiphop.com Server www.undergroundhiphop.com
86. IPULLRANK.COM @ IPULLRANK
Static Site Generators
Static Site Generators allows you to use a CMS and other tech to generate flat files for your site.
https://www.staticgen.com/
90. IPULLRANK.COM @ IPULLRANK
React won the Battle
Across the web, people are abandoning Angular and focusing more on React with Vue gaining a lot of ground in
the past couple years.
95. You need to compute the diff between the view source and the inspect
element versions of the page to understand what Google may or may
not see.
96. Use Screaming Frog – Crawl in text mode and in JavaScript
mode then compare the hash columns. Those that are different
have rendering disparities.
97. If you want to get fancy, crawl how Google crawls.
https://codeburst.io/a-guide-to-automating-scraping-the-web-with-
javascript-chrome-puppeteer-node-js-b18efb9e9921
98. You can even watch a Googler teach you to scrape Google with
it.
99. Nothing that requires a user action to be displayed
will ever be considered for ranking.
100. @justinrbriggs has a good step by step on how to do this.
https://www.briggsby.com/auditing-javascript-for-seo/
102. “It's becoming such a complex system, we
often can't really know how a change will
affect our own site until we roll it out.”
@WillCritchlow
103. IPULLRANK.COM @ IPULLRANK
I’M SURE ODN IS
GREAT…
…but you don’t need their platform to perform
effective split-testing.
104. Hypothesis Generation
“If we build X internal links from pages with 10 or more external links
to Y pages that we want to improve visibility, we will see an increase in
rankings and traffic.”
105. Bucketing Pages
Generate three buckets of pages. Two controls and one variant. Make
sure to choose pages that regularly get crawled and regularly get traffic.
109. IPULLRANK.COM @ IPULLRANK
Prophet by Facebook
Facebook’s Prophet can be used to Forecast impact of changes https://facebook.github.io/prophet/
110. IPULLRANK.COM @ IPULLRANK
CausalImpact Package
Google’s CausalImpact package
is one can be used to determine
impact of experiments.
https://google.github.io/CausalImpact/CausalImpact.html