1. Dealing With Algorithm Changes
Micah Fisher-Kirshner | Senior SEO Manager
www.become.comSeptember 13th
2011
2. Quick Background
• Become.com: price comparison website
• Competitors:
US: NexTag, Shopping, Shopzilla
UK: PriceRunner, Twenga
DE: Idealo, Billiger
JP: Kakaku
• Why does Google
‘hate’ us?
We are a vertical search engine
• Why should Google
‘love’ us?
We had the highest ROAS for CSEs
in 2010
3. The Event: Panda Attacks
• Monitoring system goes haywire
Traffic drops 20+%
, I knew I
should’ve called in
sick…
, I knew I
should’ve called in
sick…
4. Event Checklist
Q1. Is the data in?
Q2. Are other business teams affected?
Q3. Are other sites talking about an algo update?
Q4. What launched recently?
Q5. What areas are affected?
Q6. Is something broken?
Q7. Who’s saying and looking at what things?
Q8. What kind of sites are dropping?
Q9. What are the algo change theories?
Q10. Does our issue match any theories?
Q11. What can we do to quickly recover?
5. Q1 Is The Data Fully In?
• Maintain a good relationship with ops
• Massive events
require flexibility
within organization
• Utilize Google
Analytics hourly
reports with
advanced segments
Let Me
Check…
Let Me
Check…File @
Bug!
File @
Bug!
6. Q2 Who Else Is Affected?
• Find the limits of the event
SEO affects everything & everything affects SEO
• Communication is essential
This isn’t the time to send emails
Keep your departments near
7. Q3 Algorithm Update Rumors?
• Read, read, read!
Voracious
• Focus on forums and search news sites as they will get
information out quickly
Webmaster World
Search Engine Land
8. Q4 What Was Recently Launched?
• Keep an event log
Sometimes product launches a month back is the cause
• Bug the engineers
Not every detail is always written down
Watch for rollbacks that undo critical changes
9. Q3 Algorithm Update Rumors?
• Read some more!
• Remember: just because you find one issue, does NOT
mean this is the only problem
Go back to the same forums and search news sites
10. Q5 What Areas Are Affected?
• Segment in any way you can
Keyword groupings: keyword length, keyword traffic level, motive
Page groupings: home, category, product
Domain/site groupings
Affected all but one sub-domain, in internet explorer, only during certain hours
Bad display ad
11. Q3 Algorithm Update Rumors?
• Any mentions yet around the web?
One or two people doesn’t make it so
GET BACK TO WORK!GET BACK TO WORK!
12. Q6 Did Anything Break?
• Let’s assume everything has been white hat, right?
Sometimes broken or forgotten processes can lead to a broken website that looks like
black hat SEO
• Know what is fundamental to your site’s SEO
Backend functions are the easiest to miss
Worker transitions always miss certain processes
Go back to ops team to run through the SEO checklist
13. Q3 Algorithm Update Rumors?
• Can I find algorithm update confirmation now?
Yes!
• Now what?
Data collection time!
14. Q7 Who’s Talking?
• Recognize the regulars
Skip the broken track records
Always read the important people, even if it’s just one sentence
• Scrutinize the strangers
Read the long commentators, usually something to gleam
Short comments are junk comments
• Jerks will be jerks
Push past their annoyances and listen
15. Q8 What Sites Are Dropping?
• Your ranking data shows severity of impact
• Competitive ranking data is essential
Seeing who survives help provide answers about algo updates
Searchmetrics, Sistrix, seoClarity, and SEOMoz are just the beginning
16. Q9 What Are The Theories?
BeA WhiteHat SEO
Think like a Black Hat
17. Likely Pogosticking Impact
Strong Factor
Many data sources (Google toolbar,
Chrome, ISPs, etc)
Logical way to determine better
quality
Amit Patel’s quote:
“If people type something and then go
and change their query, you could tell
they aren’t happy[.] If they go to the
next page of results, it’s a sign they’re
not happy. You can use those signs that
someone’s not happy with what we gave
them to go back and study those cases
and find places to improve search.”
Confirming Factor
AdWords ad traffic quality
Paid traffic through botnets
Hate content farms? Wait for the
new click farms!
18. Q10 What Theories Fit?
• Find out everything you
can
Work your business connections
Read blogs for ‘in-depth’ analyses
Site C
Site A
Site B
• Jot down likely
possibilities
Make sure you have enough data!
Get out in front fast before others
19. THINK LIKE A SEARCH ENGINE
Q11 What Can We Do To Recover?
Build For The User
20. Learning To Love Statistics
• Understand how changes were made
Richard Baxter at SEOgadget emulates this:
Q[uestion]:
Would you trust the information presented in this article?
SEO translation:
“‘Trust’ could be measured by the links (citations) awarded to the article, the more authoritative
the link, the more trustworthy the article. […] Linking to low quality sites, penali[z]ed sites or
downright awful places on the [I]nternet is hardly likely to inspire trust, is it?
…
“If significant volumes of links aren’t available just yet, perhaps a shorter term solution could be
to analy[z]e the social buzz associated with the article. Was there a big discussion surrounding
the URL on Twitter? How many people shared the URL via Facebook?”
Coding clues and LinkedIn
• A/B test
Sub-domains, categories, page types
How to test when the algorithm is overall instead of page level?
21. Thanks And Good Luck!
Did you fix it yet?Did you fix it yet?
Learn more at: www.become.comwww.linkedin.com/in/micahf
k
Editor's Notes
Before next slide include sunny day with “Borrowed California Day” joke for NYC
Like An Earthquake Drill
Propensity to launch major changes on the same day as Google
Get Back To Work!
For example, if tedster from WMW and Danny Sullivan say there’s been an update, that’s confirmation enough
Short Comments Are Junk Comments Joke: 140 Characters or Less
In The Plex (April 2011) by Steven Levy
Make Joke on SVP of Knowledge with proper Job Position titling with LinkedIn