When what’s good for
users isn’t good for Google
slideshare.net/willcritchlow
@willcritchlow
Will Critchlow | SearchPilot
This was a WINNING conversion rate test
-25%
organic
traffic!
In the olden days:
“Can you just SEO this page?”
UX is a ranking
factor
Our call to adventure:
UX is a ranking
factor
UX is a ranking
factor
How users see
your site
UX is a ranking
factor
UX is a ranking
factor
UX is a ranking
factor
Affects organic
search performance
Controversial Accepted
CTR for algorithm
evaluation
This has been
confirmed
repeatedly.
“I’ve talked explicitly about using
clickthrough rate for experiments”
Paul Haahr - @haahr
https:/
/youtu.be/eGSGZMI4Z_I?t=979
Controversial Accepted
Human UX
ratings
Qualitative ratings
of pages that rank:
used to evaluate
algorithms
“... make this page a frustratingly poor
user experience”
-- reason for poor page quality rating
Search quality rater guidelines [PDF]
Controversial Accepted
ML models
trained on ratings
E.g. Panda -
trained to like the
things humans
like
Back in 2011, I was suggesting we run our own
Panda-like quality surveys (WBF here, instructions here)
Sidenote: totally worth running
these surveys for executive buy-in
Since Panda went real-time, quality issues don’t necessarily cause
obvious drops correlated with algorithm history dates
Client site 1 Client site 2
Would you trust information from this website? 72% 64%
Client site 1 Client site 2 Key competitor
Would you trust information from this website? 72% 64% 81%
Client site 1 Client site 2 Key competitor
Would you trust information from this website? 72% 64% 81%
Is this website written by experts? 50% 46% 65%
Client site 1 Client site 2 Key competitor
Would you trust information from this website? 72% 64% 81%
Is this website written by experts? 50% 46% 65%
Would you give this site your credit card details? 29% 21% 43%
Client site 1 Client site 2 Key competitor
Would you trust information from this website? 72% 64% 81%
Is this website written by experts? 50% 46% 65%
Would you give this site your credit card details? 29% 21% 43%
Are there any noticeable errors on this page? 6% 4% 1%
Client site 1 Client site 2 Key competitor
Would you trust information from this website? 72% 64% 81%
Is this website written by experts? 50% 46% 65%
Would you give this site your credit card details? 29% 21% 43%
Are there any noticeable errors on this page? 6% 4% 1%
Does this page provide original content or info? 76% 72% 85%
Client site 1 Client site 2 Key competitor
Would you trust information from this website? 72% 64% 81%
Is this website written by experts? 50% 46% 65%
Would you give this site your credit card details? 29% 21% 43%
Are there any noticeable errors on this page? 6% 4% 1%
Does this page provide original content or info? 76% 72% 85%
Would you recognize this site as an authority? 44% 33% 58%
Client site 1 Client site 2 Key competitor
Would you trust information from this website? 72% 64% 81%
Is this website written by experts? 50% 46% 65%
Would you give this site your credit card details? 29% 21% 43%
Are there any noticeable errors on this page? 6% 4% 1%
Does this page provide original content or info? 76% 72% 85%
Would you recognize this site as an authority? 44% 33% 58%
Does this website contain insightful analysis? 72% 62% 81%
Client site 1 Client site 2 Key competitor
Would you trust information from this website? 72% 64% 81%
Is this website written by experts? 50% 46% 65%
Would you give this site your credit card details? 29% 21% 43%
Are there any noticeable errors on this page? 6% 4% 1%
Does this page provide original content or info? 76% 72% 85%
Would you recognize this site as an authority? 44% 33% 58%
Does this website contain insightful analysis? 72% 62% 81%
Would you consider bookmarking pages on this site? 44% 38% 56%
Client site 1 Client site 2 Key competitor
Would you trust information from this website? 72% 64% 81%
Is this website written by experts? 50% 46% 65%
Would you give this site your credit card details? 29% 21% 43%
Are there any noticeable errors on this page? 6% 4% 1%
Does this page provide original content or info? 76% 72% 85%
Would you recognize this site as an authority? 44% 33% 58%
Does this website contain insightful analysis? 72% 62% 81%
Would you consider bookmarking pages on this site? 44% 38% 56%
Are there excessive adverts on this website? 2% 2% 8%
Client site 1 Client site 2 Key competitor
Would you trust information from this website? 72% 64% 81%
Is this website written by experts? 50% 46% 65%
Would you give this site your credit card details? 29% 21% 43%
Are there any noticeable errors on this page? 6% 4% 1%
Does this page provide original content or info? 76% 72% 85%
Would you recognize this site as an authority? 44% 33% 58%
Does this website contain insightful analysis? 72% 62% 81%
Would you consider bookmarking pages on this site? 44% 38% 56%
Are there excessive adverts on this website? 2% 2% 8%
Could pages from this site appear in print? 54% 54% 59%
Controversial Accepted
Site speed
measured
directly
Real User Metrics
including site
speed from
Chrome
https:/
/moz.com/blog/google-chrome-usage-data-measure-site-speed
https:/
/moz.com/blog/google-chrome-usage-data-measure-site-speed
First announced by SearchPilot CTO, Tom
Anthony
Controversial Accepted
CTR as a direct
ranking factor
This has been
explicitly denied
by Google reps.
Some kind of
SERP interaction
affecting results
we see
My theory.
We sell custom cigar humidors. Our custom cigar
humidors are handmade. If youʼre thinking of
buying a custom cigar humidor, please contact
our custom cigar humidor specialists at
custom.cigar.humidors@example.com
Keyword stuffing: Doesn’t mean keyword density is the correct metric
Most organisations refuse to accept they overlap
SEO CRO
Search Engine
Optimisation
Conversion Rate
Optimisation
But in truth there is a lot of overlap
SEO CRO
While some SEO changes donʼt impact CRO
SEO
Meta info - invisible
to UX CRO
And some CRO changes donʼt impact SEO
SEO Non-indexed pages
CRO
An awful lot of changes affect both
Crossing the threshold:
Weʼre not in
Kansas anymore
We should be building SEO
hypotheses from UX fundamentals
BUT
-25%
organic
traffic!
Tweet me: @willcritchlow #BrightonSEO
We publish winning and losing tests at searchpilot.com
We publish our case study email newsletter because itʼs the resource we wish existed
Conversion rate
optimisation focuses
on converting more
of your existing traffic
(bottom of the
funnel)
Conversion Rate
Optimisation
Search engine
optimisation focuses
on getting additional
traffic to your
website (top of
funnel)
Conversion Rate
Optimisation
Search Engine
Optimisation
Elixir:
Full funnel testing
combines these
approaches to do
both at the same
time
Full Funnel Testing
How do SEO and
product teams work
together well?
Supreme ordeal:
Green
Totally safe:
Not in scope for
CRO testing
Amber
Needs care:
Potential for
impact, but not
overlapping
Red
Direct overlap:
Overlapping
changes or
hypotheses
We then use a flow
chart for amber tests
Road back:
1.
Are there
conversion rate
tests running or
planned on this
site section?
2.
Do the
hypotheses affect
the same part of
the page?
3.
Do the tests target
the same user
behaviour levers?
We use a RACI matrix to keep the right people in the loop
Test Type SEO Teams
Product
Teams
Broader
Marketing
Engineering Ad Teams Leadership
Green Tests Accountable Informed Informed Informed Informed Informed
We use a RACI matrix to keep the right people in the loop
Test Type SEO Teams
Product
Teams
Broader
Marketing
Engineering Ad Teams Leadership
Green Tests Accountable Informed Informed Informed Informed Informed
Amber Tests Responsible Accountable Informed Consulted Informed Informed
We use a RACI matrix to keep the right people in the loop
Test Type SEO Teams
Product
Teams
Broader
Marketing
Engineering Ad Teams Leadership
Green Tests Accountable Informed Informed Informed Informed Informed
Amber Tests Responsible Accountable Informed Consulted Informed Informed
Red Tests Responsible Accountable Informed Consulted Informed Informed
In summary
UX is a ranking
factor?
Not everyone agrees
Controversial Accepted
We just need to know that UX is aligned with Google’s goals
But we don’t need the futurist
perspectives or the tinfoil hat conspiracy
theories for my purposes here
We just need to know that UX is aligned with Google’s goals
1. Panda (ML-trained on UX signals)
2. Quant and qual measurement of algo
changes using UX metrics
To know that we should be building
SEO hypotheses from UX
fundamentals
BUT: then we affect users and search
So: we need to play nicely together
Test Type SEO Teams
Product
Teams
Broader
Marketing
Engineering Ad Teams Leadership
Green Tests Accountable Informed Informed Informed Informed Informed
Amber Tests Responsible Accountable Informed Consulted Informed Informed
Red Tests Responsible Accountable Informed Consulted Informed Informed
And ideally, test
both the SEO and
CRO impact of our
ideas
Full Funnel Testing
Questions: @willcritchlow
Lessons from
100s of SEO
tests
The restaurant
4.10pm
Image credits
● Hornet
● Beach
● Paul Haahr
● Google
● Smile
● Pug
● Chrome
● Tinfoil hat
● Broken plate
● Confused monkey
● Unicyclist
● Horse
● Phone
● Devil cat
● Coloured smoke
● Kansas
● Brighton
● On stage in San Diego

BrightonSEO - SearchPilot - Will Critchlow - When what’s good for users isn’t good for Google

  • 1.
    When what’s goodfor users isn’t good for Google slideshare.net/willcritchlow @willcritchlow Will Critchlow | SearchPilot
  • 2.
    This was aWINNING conversion rate test
  • 3.
  • 4.
    In the oldendays: “Can you just SEO this page?”
  • 5.
    UX is aranking factor Our call to adventure:
  • 6.
    UX is aranking factor
  • 7.
    UX is aranking factor How users see your site
  • 8.
    UX is aranking factor
  • 9.
    UX is aranking factor
  • 10.
    UX is aranking factor Affects organic search performance
  • 11.
  • 12.
    CTR for algorithm evaluation Thishas been confirmed repeatedly.
  • 14.
    “I’ve talked explicitlyabout using clickthrough rate for experiments” Paul Haahr - @haahr
  • 15.
  • 16.
  • 17.
    Human UX ratings Qualitative ratings ofpages that rank: used to evaluate algorithms
  • 20.
    “... make thispage a frustratingly poor user experience” -- reason for poor page quality rating Search quality rater guidelines [PDF]
  • 21.
  • 22.
    ML models trained onratings E.g. Panda - trained to like the things humans like
  • 23.
    Back in 2011,I was suggesting we run our own Panda-like quality surveys (WBF here, instructions here)
  • 24.
    Sidenote: totally worthrunning these surveys for executive buy-in Since Panda went real-time, quality issues don’t necessarily cause obvious drops correlated with algorithm history dates
  • 25.
    Client site 1Client site 2 Would you trust information from this website? 72% 64%
  • 26.
    Client site 1Client site 2 Key competitor Would you trust information from this website? 72% 64% 81%
  • 27.
    Client site 1Client site 2 Key competitor Would you trust information from this website? 72% 64% 81% Is this website written by experts? 50% 46% 65%
  • 28.
    Client site 1Client site 2 Key competitor Would you trust information from this website? 72% 64% 81% Is this website written by experts? 50% 46% 65% Would you give this site your credit card details? 29% 21% 43%
  • 29.
    Client site 1Client site 2 Key competitor Would you trust information from this website? 72% 64% 81% Is this website written by experts? 50% 46% 65% Would you give this site your credit card details? 29% 21% 43% Are there any noticeable errors on this page? 6% 4% 1%
  • 30.
    Client site 1Client site 2 Key competitor Would you trust information from this website? 72% 64% 81% Is this website written by experts? 50% 46% 65% Would you give this site your credit card details? 29% 21% 43% Are there any noticeable errors on this page? 6% 4% 1% Does this page provide original content or info? 76% 72% 85%
  • 31.
    Client site 1Client site 2 Key competitor Would you trust information from this website? 72% 64% 81% Is this website written by experts? 50% 46% 65% Would you give this site your credit card details? 29% 21% 43% Are there any noticeable errors on this page? 6% 4% 1% Does this page provide original content or info? 76% 72% 85% Would you recognize this site as an authority? 44% 33% 58%
  • 32.
    Client site 1Client site 2 Key competitor Would you trust information from this website? 72% 64% 81% Is this website written by experts? 50% 46% 65% Would you give this site your credit card details? 29% 21% 43% Are there any noticeable errors on this page? 6% 4% 1% Does this page provide original content or info? 76% 72% 85% Would you recognize this site as an authority? 44% 33% 58% Does this website contain insightful analysis? 72% 62% 81%
  • 33.
    Client site 1Client site 2 Key competitor Would you trust information from this website? 72% 64% 81% Is this website written by experts? 50% 46% 65% Would you give this site your credit card details? 29% 21% 43% Are there any noticeable errors on this page? 6% 4% 1% Does this page provide original content or info? 76% 72% 85% Would you recognize this site as an authority? 44% 33% 58% Does this website contain insightful analysis? 72% 62% 81% Would you consider bookmarking pages on this site? 44% 38% 56%
  • 34.
    Client site 1Client site 2 Key competitor Would you trust information from this website? 72% 64% 81% Is this website written by experts? 50% 46% 65% Would you give this site your credit card details? 29% 21% 43% Are there any noticeable errors on this page? 6% 4% 1% Does this page provide original content or info? 76% 72% 85% Would you recognize this site as an authority? 44% 33% 58% Does this website contain insightful analysis? 72% 62% 81% Would you consider bookmarking pages on this site? 44% 38% 56% Are there excessive adverts on this website? 2% 2% 8%
  • 35.
    Client site 1Client site 2 Key competitor Would you trust information from this website? 72% 64% 81% Is this website written by experts? 50% 46% 65% Would you give this site your credit card details? 29% 21% 43% Are there any noticeable errors on this page? 6% 4% 1% Does this page provide original content or info? 76% 72% 85% Would you recognize this site as an authority? 44% 33% 58% Does this website contain insightful analysis? 72% 62% 81% Would you consider bookmarking pages on this site? 44% 38% 56% Are there excessive adverts on this website? 2% 2% 8% Could pages from this site appear in print? 54% 54% 59%
  • 36.
  • 37.
    Site speed measured directly Real UserMetrics including site speed from Chrome
  • 38.
  • 39.
  • 40.
  • 41.
    CTR as adirect ranking factor This has been explicitly denied by Google reps.
  • 42.
    Some kind of SERPinteraction affecting results we see My theory.
  • 43.
    We sell customcigar humidors. Our custom cigar humidors are handmade. If youʼre thinking of buying a custom cigar humidor, please contact our custom cigar humidor specialists at custom.cigar.humidors@example.com Keyword stuffing: Doesn’t mean keyword density is the correct metric
  • 44.
    Most organisations refuseto accept they overlap SEO CRO Search Engine Optimisation Conversion Rate Optimisation
  • 45.
    But in truththere is a lot of overlap SEO CRO
  • 46.
    While some SEOchanges donʼt impact CRO SEO Meta info - invisible to UX CRO
  • 47.
    And some CROchanges donʼt impact SEO SEO Non-indexed pages CRO
  • 48.
    An awful lotof changes affect both
  • 49.
    Crossing the threshold: Weʼrenot in Kansas anymore
  • 50.
    We should bebuilding SEO hypotheses from UX fundamentals
  • 51.
  • 52.
    Tweet me: @willcritchlow#BrightonSEO We publish winning and losing tests at searchpilot.com We publish our case study email newsletter because itʼs the resource we wish existed
  • 53.
    Conversion rate optimisation focuses onconverting more of your existing traffic (bottom of the funnel) Conversion Rate Optimisation
  • 54.
    Search engine optimisation focuses ongetting additional traffic to your website (top of funnel) Conversion Rate Optimisation Search Engine Optimisation
  • 55.
    Elixir: Full funnel testing combinesthese approaches to do both at the same time Full Funnel Testing
  • 56.
    How do SEOand product teams work together well? Supreme ordeal:
  • 57.
    Green Totally safe: Not inscope for CRO testing
  • 58.
  • 59.
  • 60.
    We then usea flow chart for amber tests Road back:
  • 61.
    1. Are there conversion rate testsrunning or planned on this site section?
  • 62.
    2. Do the hypotheses affect thesame part of the page?
  • 63.
    3. Do the teststarget the same user behaviour levers?
  • 65.
    We use aRACI matrix to keep the right people in the loop Test Type SEO Teams Product Teams Broader Marketing Engineering Ad Teams Leadership Green Tests Accountable Informed Informed Informed Informed Informed
  • 66.
    We use aRACI matrix to keep the right people in the loop Test Type SEO Teams Product Teams Broader Marketing Engineering Ad Teams Leadership Green Tests Accountable Informed Informed Informed Informed Informed Amber Tests Responsible Accountable Informed Consulted Informed Informed
  • 67.
    We use aRACI matrix to keep the right people in the loop Test Type SEO Teams Product Teams Broader Marketing Engineering Ad Teams Leadership Green Tests Accountable Informed Informed Informed Informed Informed Amber Tests Responsible Accountable Informed Consulted Informed Informed Red Tests Responsible Accountable Informed Consulted Informed Informed
  • 68.
  • 69.
    UX is aranking factor?
  • 70.
  • 71.
    We just needto know that UX is aligned with Google’s goals But we don’t need the futurist perspectives or the tinfoil hat conspiracy theories for my purposes here
  • 72.
    We just needto know that UX is aligned with Google’s goals 1. Panda (ML-trained on UX signals) 2. Quant and qual measurement of algo changes using UX metrics
  • 73.
    To know thatwe should be building SEO hypotheses from UX fundamentals
  • 74.
    BUT: then weaffect users and search
  • 75.
    So: we needto play nicely together Test Type SEO Teams Product Teams Broader Marketing Engineering Ad Teams Leadership Green Tests Accountable Informed Informed Informed Informed Informed Amber Tests Responsible Accountable Informed Consulted Informed Informed Red Tests Responsible Accountable Informed Consulted Informed Informed
  • 76.
    And ideally, test boththe SEO and CRO impact of our ideas Full Funnel Testing
  • 77.
  • 78.
    Lessons from 100s ofSEO tests The restaurant 4.10pm
  • 79.
    Image credits ● Hornet ●Beach ● Paul Haahr ● Google ● Smile ● Pug ● Chrome ● Tinfoil hat ● Broken plate ● Confused monkey ● Unicyclist ● Horse ● Phone ● Devil cat ● Coloured smoke ● Kansas ● Brighton ● On stage in San Diego