3. 3
Measurement has come a long way, but there
are still some people skeptical of the value of
audience segmentation.
4. 4
There are a lot of complaints in the internet
marketing echo chamber about personas not
being a valuable fact in lieu of “real data.”
5. 5
Users delete their cookies on a regular
basis.
AdBlock prevents tracking entirely for
analytics tools that are not whitelisted
There’s also a small percentage of weirdos
that don’t enable JavaScript.
16. 16
There are a number of ways to effectively measure
your personas.
17. @iPullRank
If your site has user profiles, users can self-
segment based on the data you request.
You can tag landing pages based on the
audience targeting options of your ad.
You can append data from sources like
FullContact and Towerdata Intelligence.
19. @iPullRank
On the back end,
Google Analytics allows
you to measure affinity
segments.
On the frontend,
Google allows you to
target different
audiences with affinity
semgents.
20. 20
Based on a sequence of behaviors you
might segment your users
Leverage affinity segments to determine
your personas
Sugar plum gummi bears. Apple pie yummy
cake with candy
22. @iPullRank
In Oli’s talk from last year, he defined
a user’s “Context Bucket” as their
desires, expectations and data that
represent what a user wants when
they land on your site.
23. @iPullRank
I counter that personas and context buckets are one and the
same thing. Personas are a long form version of what he’s
describing with context buckets.
This tells us why the available audience is
interested.
This tells us the specific needs that must be met.
24. @iPullRank
Oli’s framework aligns with persona and user journey modeling in helping to understand context, communicate with your
audience clearly, making sure your message is congruent with their needs and that you speak in a way that they believe is
credible.
25. @iPullRank
Market segmentation is a marketing strategy that involves dividing a
broad target market into subsets of consumers who have common
needs and priorities, and then designing and implementing strategies to
target them. Market segmentation strategies may be used to identify the
target customers, and provide supporting data for positioning to achieve
a marketing plan objective.
26. @iPullRank
A segment is a representation of your users or audience based on the combination of two or more factors or data points.
27. @iPullRank
This is an example of the same types of users measured based on specific identifiable data points.
28. @iPullRank
Cohorts are segments of people that are tied together by similar experience. Colloquially we refer to these as generations.
A common cohort that you hear about is marketing as in “Facebook is losing the Millenials to other services like Instagram”
29. @iPullRank
In terms of analytics, cohorts are defined by users that visited in the same timespan. For example, everyone here is in the
NYC Conversion Road Trip cohort.
30. @iPullRank
Personas are archetypal representations of actual people in your target audience. Depending on whose model you follow
they typically include a user story, user needs, engagement insights. They are often given alliterated names and have a
quote that personifies them quickly.
32. @iPullRank
There are variety of persona types, but the most important distinction to make in the marketing context is the distinction
between those that are actively looking to buy your product (buyer personas) and those that are just interested in
consuming your content (audience personas). There is often overlap between the two, but they should be treated
differently.
33. @iPullRank
Keywords while the best proxy for intent were just a type of abstraction for a need a user is looking to fulfill. As marketers
we’ve focused on the need rather than the user. Personas are also an abstraction, but they put more focus on users.
Metaphorically speaking a keyword is a stick figure while a persona is an action figure meaning a persona is a more robust
representation of a user and their intent.
36. 36
There are a lot of complaints in the internet
marketing echo chamber about personas not
being a valuable fact in lieu of “real data.”
37. @iPullRank
When performing an affinity mapping session you want to do 3 distinct rounds of sharing and discussion. Each round has a
different color post-it. Give the team 20 minutes to jot down their thoughts for each round then go around the room
discussing each post-it until there’s a consensus of what goes on the board. Discuss groupings helps segments to form.
38. @iPullRank
It is highly effective to do the affinity mapping session with stakeholders over Google+ Hangouts with a tool called
http://mural.ly. Mural.ly is described as “Google Docs for Visual People.” It allows for collaboration and even has a template
specifically for affinity mapping.
39. @iPullRank
Asking open ended questions using an online survey tool will allow you to get a sense of the trending issues on the mind of
you target demographic. The downside is that data is very difficult to analyze at scale and often comes down to just using
word clouds of responses.
40. @iPullRank
Ethnographic research is an iterative process of observing users or audiences in their natural environments and distilling
conclusions from those observations. This has been brought into the online realm and nicknamed “netnographic research.”
41. 41
Focus groups involve sit downs with actual or
potential consumers of the either your product,
service, or content. These are typically run by a
moderator, but be warned that moderators or
highly opinionated respondents may coerce the
opinion of the group. (Mad Men Sidebar: Faye
Miller is Don Draper’s best girlfriend).
42. @iPullRank
User interviews are just like focus groups, but just focusing on one user at a time. You can do this to see how users are
interacting with your product or content. You can leverage Google+ Hangouts for this or use a tool like
http://www.luckyorange.com or http://www.usertesting.com to watch a user’s session after the fact to see where they are
getting hung up.
45. @iPullRank
At my previous agency I built a methodology of layering data from the market segmentation tools and validating those data
points with the Social advertising inventories to collect the demographic and psychographic data points and then build the
user stories, key characteristics and engagement insights.
46. @iPullRank
Layering data entailed seeking out commonalities in multiple datasets. If there were commonalities across the datasets
then we assumed that a data point from one was valid to the other and we could use both to build our personas.
47. @iPullRank
While Experian Simmons is quite an expensive tool that is out of reach for many there are free Mosaic Interactive Guides for
the US, Germany, Australia, France, Brazil, Scotland, and Spain. I’ll get you started with the US guide http://guides.business-
strategies.co.uk/mosaicusa2011/html/visualisation.htm
48. @iPullRank
Once you know what questions you need answered you can increase the sample size and get more definitive insights
across a wider group of your audience. SurveyMonkey Audience is also great for this. They have great resources on survey
design: http://help.surveymonkey.com/articles/en_US/kb/Design-Tips-How-to-create-and-administer-effective-surveys
49. @iPullRank
Your analytics is the most relevant data source to you and you may have a lot of information specifically about user needs if
you have internal search activated. You also have a lot of information on the rest of their behaviors especially if you are
already doing any cohort analysis.
50. @iPullRank
User profiles are a great way to truly understand precisely what your users are here for. You can fire custom variables into
Google Analytics when these users login to then create advanced segments which allow you to see what actions different
customer types are taking.
_gaq.push([‘_setCustomVar',1,'user
Segment',userSegmentName ,1]);
51. @iPullRank
Speaking of user profiles, they are great for scraping to get insights. I scraped Moz’s active user base last year and did some
in-depth analysis. http://moz.com/blog/what-i-learned-from-scraping-seomozs-active-user-base If you have your own user
profiles you can do the same with a few database queries.
52. @iPullRank
Ethnographic research is an iterative process of observing users or audiences in their natural environments and distilling
conclusions from those observations. This has been brought into the online realm and nicknamed “netnographic research.”
53. @iPullRank
Google’s Consumer Barometer provides data on buying habits around specific product types and what devices they use to
buy them. http://consumerbarometer.com
54. @iPullRank
All of these internal and external data sources begin to give you a very clear hypothesis of who these users are. Now the
only thing left to do is making sense of the data points in context of your business and its goals.
56. @iPullRank
One of Moz's key business goals is to increase the number of SMBs that signup for free services that become monthly
subscribers. Therefore the goal of this persona exercise will be to discover a key segment of Moz's audience that is very
likely to share and link to content, but hasn't purchased a Moz Analytics pro membership yet
57. @iPullRank
Identifying the demographics allows to figure out where does the persona fall. Based on this data it’s equally valid to have a
segment that is 25-34 and female desktop user as it is to have one that 55-64 and male mobile user. The percentages give
the likelihood that that person is part of your audience.
58. @iPullRank
I’ve used a combination of Moz’s Q&A, Twitter’s advanced search and data that I scraped from Moz’s user profiles to
uncover user needs. I’ve done it this way because these data sources are available for this site. Another site would require
other methods.
59. @iPullRank
Facebook Graph Search, Twtrland, and Twitter Advanced Search all allow me to look precisely at the people that are
interested in Moz based and review their demographics to then examine their psychographics based on the other features
of your profile.
60. @iPullRank
This research process yielded a persona that called Mozzy Smurf that is always on the run trying to live the 4-Hour week
lifestyle. He’s an avid traveler, lover of Moz’s content, but he expects more out of the product. Mozzy Smurf would become a
subscriber if their were different pricing options, an iPhone app, multi-seat accounts and more premium content.
The User Needs
or Context
Bucket
87. @iPullRank
A journey can be as granular as you want it to be, but keeping it basic is best because you’ll have to map and manage
content for each phase. Specifically what content goes with which phase?
89. @iPullRank
User desires
travel, but
doesn’t quite
know where
to go.
User is
looking to
book a
complete
itinerary
User is
planning out
everything
they need to
do for the
trip.
User is
traveling
from their
home to their
custom
itinerary trip.
User has
arrived at the
destination.
User is
enjoying their
trip.
User is
heading
home.
User has had
an issue
during their
trip.
User wants to
share their
trip with
friends and
family.
EXPLORE BOOK PREPARE TRAVEL ARRIVE STAY DEPART PROBLEM REPORT
How do I find
a place to go?
How do I find
all the
features of the
trip I want and
book it?
What do I
need to get
ready for this
trip?
How to do I
get to the
airport? What
if my flight is
canceled?
What do I do
once I land?
How to find
the things in
my itinerary?
How do I get
back to the
airport? Can I
take this stuff
home?
How do I get
help when I’m
on a trip?
Where’s the
best place to
share my
experience
online?
Trip.Me had no content for
any of these phases in the
user journey
95. @iPullRank
PAGE VALUE READABILITY SCORE TWEETS
CONVERSIONS SENTENCE COUNT LIKES
CONVERSION
RATE
# IMAGES CONTENT
FORMAT
TRAFFIC BOUNCE RATE PINS
ORGANIC
SEARCH TRAFFIC
WORD COUNT TIME ON
SITE
SOCIAL MEDIA
TRAFFIC
READING TIME # VIDEOS
96. @iPullRank
SECTION CATEGORY
WHAT IS IT? REDUNDANCY
TIMELINESS VOICE AND TONE
ADHERENCE
QUALITY ACTIONABILITY
LINKWORTHINESS SHAREWORTHINESS
CONVERSION NOTES SUGGESTIONS
TARGET PERSONA NEED STATE
120. @iPullRank
Bring your quantitative and qualitative data into one sheet and use pivot charts on the audit data to determine insights,
mainly what you should be testing.
122. 122
$0.00
$100.00
$200.00
$300.00
$400.00
Fairly Easy Difficult Easy Standard Fairly Difficult Very Confusing
Building pivot charts comparing the page value metric vs. the page’s readability score quickly tells you where you should consider
performing measurement tests.
123. 123
PAGE VALUE VS.
FORMAT
REDUNDANCY VS.
SECTION
PAGE VALUE VS
PAGE SPEED
# IMAGES VS
READABILITY
SCORE
# VIDEOS VS
WORD COUNT
FORMAT VS LINKS
# VIDEOS VS TIME
ON PAGE
PERSONA VS.
FORMAT
RESPONSE CODE
VS LINKS
TIME ON PAGE VS
SOCIAL SHARES
PAGE VALUE VS
SOCIAL SHARES
LINKS VS.
DIRECTORY
LINKS VS QUALITY
WORD COUNT VS
TIME ON PAGE
PERSONA VS.
CONVERSION RATE
SECTION VS
ACTIONABILITY
TIME ON PAGE VS
READING TIME
PERSONA VS. PAGE
VALUE