Everyone’s Customers Are Wrong and
         Their Data Is Lying



              Evan Hamilton
       Head of Community, UserVoice
DATA!
Big Data! Social Data! Mobile Data!
          It’s the future!
Actually...
Data doesn’t tell the whole story.
Data doesn’t tell the whole story.




A/B test: will this carousel increase clicks?
                 ^ incredibly ugly
Data doesn’t tell the whole story.




“Hey, look! The version of the page with the carousel is
          making more people click through!”
Data doesn’t tell the whole story.




“Oh wait...it’s making more people click, not actually
                       convert.”

                Tricky data at work!
Data doesn’t tell the whole story.




  We created confusion, not conversions.
Data doesn’t tell the whole story.
Data doesn’t tell the whole story.




         Hey, that looks good!
Data doesn’t tell the whole story.




Oops. They “forgot” to report refunds (amongst other things.)
PEOPLE!
 Feelings! Empathy! Gut!
    It’s the only way!
Actually...
People don’t tell the whole story.
People don’t tell the whole story.




“Wow, maybe we should build this. Let’s ask for more
                feedback first...”
People don’t tell the whole story.




Oh. So they want this feature because they’re not
        responding to customers. Next.
There’s always more context.

• What is the pain point?
• Are these paying users? Freeloaders?
• Are these people using your product in the
 intended way?
  • (We had someone using our forum to name their baby.
 Cute, but not our target market.)
• Are they using your main features?
• Early “adopters”?
 • (Robert Scoble is not your target market. He is fickle.)
People don’t tell the whole story.




      “Makes sense, let’s build it!”
People don’t tell the whole story.




But customers found this confusing. (And it made us throw
              up in our mouths a little bit.)
People don’t tell the whole story.




“Yet another online identity” is what we should have
                     focused on.
People don’t tell the whole story.




                           v2
        No password. No signup. No registration.
Solved the problem without a complicated sign-in process.
Customers often give you bad solutions to
             real problems.

(Don’t misinterpret this as good solutions
          or not real problems.)
Everything is wrong.
    Now what?
Data still has to be interpreted.
Data still has to be interpreted.


Requests can lead you astray.
Data still has to be interpreted.


  Requests can lead you astray.


Combine data + customer stories.
Combine data + customer stories.




Back when we just had UserVoice Feedback, we’d lose
lots of customers after 1 year. So we asked them why
                  they were leaving.
Turns out they were growing up. New staff, new
department heads...& feedback forums didn’t have a
        definitive place in the organization.
So we built a customer service tool into our product. Not
what we necessarily planned to build...but what customers
                        wanted.
It increased retention immensely. We now have a place in
                     the organization.

We got here because data alerted us to something, & user
              stories filled out the details.
Don’t get lost in the data rat-hole.
Don’t get lost in the data rat-hole.

Don’t chase 1% when you can get 15%.
Don’t get lost in the data rat-hole.

Don’t chase 1% when you can get 15%.

  Don’t get lost in minor homepage
             optimization.
You’ll never know what you could get unless
          you try something crazy.
Tried this yet? If not, go do that
before trying another shade of blue.
Tips, tricks, and tools

• (At least) Google Analytics
  • Hell, even Bit.ly links give you a basis to work from
• Mixpanel
• The ol’ fake button
  • How many people click a (fake) “calendar” button in your app? If 40% of
  users do, you should build it!

• Test with segments
  • Try tests on nonpaying users if you don’t want to scare off paying users
• Talk to customers (constantly)
• UserVoice
Thank you!

       Evan Hamilton
Head of Community, UserVoice
       @evanhamilton

Everyone’s Customers Are Wrong and Their Data Is Lying

  • 1.
    Everyone’s Customers AreWrong and Their Data Is Lying Evan Hamilton Head of Community, UserVoice
  • 2.
    DATA! Big Data! SocialData! Mobile Data! It’s the future!
  • 3.
  • 4.
    Data doesn’t tellthe whole story. A/B test: will this carousel increase clicks? ^ incredibly ugly
  • 5.
    Data doesn’t tellthe whole story. “Hey, look! The version of the page with the carousel is making more people click through!”
  • 6.
    Data doesn’t tellthe whole story. “Oh wait...it’s making more people click, not actually convert.” Tricky data at work!
  • 7.
    Data doesn’t tellthe whole story. We created confusion, not conversions.
  • 8.
    Data doesn’t tellthe whole story.
  • 9.
    Data doesn’t tellthe whole story. Hey, that looks good!
  • 10.
    Data doesn’t tellthe whole story. Oops. They “forgot” to report refunds (amongst other things.)
  • 11.
    PEOPLE! Feelings! Empathy!Gut! It’s the only way!
  • 12.
  • 13.
    People don’t tellthe whole story. “Wow, maybe we should build this. Let’s ask for more feedback first...”
  • 14.
    People don’t tellthe whole story. Oh. So they want this feature because they’re not responding to customers. Next.
  • 15.
    There’s always morecontext. • What is the pain point? • Are these paying users? Freeloaders? • Are these people using your product in the intended way? • (We had someone using our forum to name their baby. Cute, but not our target market.) • Are they using your main features? • Early “adopters”? • (Robert Scoble is not your target market. He is fickle.)
  • 16.
    People don’t tellthe whole story. “Makes sense, let’s build it!”
  • 17.
    People don’t tellthe whole story. But customers found this confusing. (And it made us throw up in our mouths a little bit.)
  • 18.
    People don’t tellthe whole story. “Yet another online identity” is what we should have focused on.
  • 19.
    People don’t tellthe whole story. v2 No password. No signup. No registration. Solved the problem without a complicated sign-in process.
  • 20.
    Customers often giveyou bad solutions to real problems. (Don’t misinterpret this as good solutions or not real problems.)
  • 21.
  • 22.
    Data still hasto be interpreted.
  • 23.
    Data still hasto be interpreted. Requests can lead you astray.
  • 24.
    Data still hasto be interpreted. Requests can lead you astray. Combine data + customer stories.
  • 25.
    Combine data +customer stories. Back when we just had UserVoice Feedback, we’d lose lots of customers after 1 year. So we asked them why they were leaving.
  • 26.
    Turns out theywere growing up. New staff, new department heads...& feedback forums didn’t have a definitive place in the organization.
  • 27.
    So we builta customer service tool into our product. Not what we necessarily planned to build...but what customers wanted.
  • 28.
    It increased retentionimmensely. We now have a place in the organization. We got here because data alerted us to something, & user stories filled out the details.
  • 29.
    Don’t get lostin the data rat-hole.
  • 30.
    Don’t get lostin the data rat-hole. Don’t chase 1% when you can get 15%.
  • 31.
    Don’t get lostin the data rat-hole. Don’t chase 1% when you can get 15%. Don’t get lost in minor homepage optimization.
  • 32.
    You’ll never knowwhat you could get unless you try something crazy.
  • 33.
    Tried this yet?If not, go do that before trying another shade of blue.
  • 34.
    Tips, tricks, andtools • (At least) Google Analytics • Hell, even Bit.ly links give you a basis to work from • Mixpanel • The ol’ fake button • How many people click a (fake) “calendar” button in your app? If 40% of users do, you should build it! • Test with segments • Try tests on nonpaying users if you don’t want to scare off paying users • Talk to customers (constantly) • UserVoice
  • 35.
    Thank you! Evan Hamilton Head of Community, UserVoice @evanhamilton