Introduction to Website Optimizer Webinar

By | November 9, 2019


Greetings. This is Tom Leung, product
manager for Google Website Optimizer. This is the Website Optimizer
introduction webinar. I’m glad you could join me. The purpose of this webinar
is to really go over a couple of things. So let’s talk about what we’re
going to cover today. We’re going to talk about what’s
wrong with lagging page or website design, why
it’s broken, and how testing helps fix it. We’re going to talk about why
online marketing happens before and after the click, and
how not enough attention is given to optimization
after the click. The more attention you give to
it, the much better returns you’re going to get
for all of your activities before the click. We’re also going to talk about
how to think about your online marketing strategy for
the long haul. We’re going to talk about why
it’s important to think about constantly enhancing how you
drive, measure, and convert your traffic. Then we’ll talk about the tools
that are provided by Google to do those
three things. We’re also going to look at
some examples of website optimizer tests, how you can
take a page, you can make some improvements to it, and see
some dramatic results. We’re going to talk about how
Website Optimizer works. Specifically, what does
this tool do? How do I set it up? What comes out at
the other end? We’re going to look at a demo. We’re going to show you,
step-by-step, how to set up a test, from A to Z. For this
webinar we’re going to do a very simple test. It’s an A-B
split test. It’s the easiest test to run. Often will give you some of
the biggest results, and generally people can set
up a test that in five minutes or less. We will probably take a little
bit longer because I’m going to go slow, in case you haven’t
done this before. And lastly, we’re going to talk
about some up frequently asked questions, some common
misperceptions about testing, Website Optimizer, clear the
decks of all those things. So, let’s get on with it. Thanks again for coming by. So what’s wrong with
the current system? Isn’t it enough to focus your
efforts on acquiring targeted traffic and then analyzing
the activity of that traffic on your site? Isn’t that the cutting edge? It’s certainly fundamental to a
successful program, but it’s not enough. And specifically, acquiring and
analyzing traffic alone, is a little bit like spending
all of your energy finding these perfect seeds, these
clicks from all the search engine marketing tools that
are available out there. You get the targeting, the
geotargeting, you picked the perfect ad creative. You use all kinds of techniques
to exclude certain keyword searches, and you create
these perfect account structures. And then, you spend many, many
hours printing out these very detailed beautiful charts about
what’s going on, on your side, and what pages are doing
well and not doing well. But that’s it. For a lot of people,
that’s all they do. And in a way, that’s like
taking these beautiful, precious seeds and just throwing
them in the air and half of them land on asphalt,
half of them land on soil. And then you have all these
detailed reports that show you that, “wow,” the ones that are
landing on the asphalt, they’re not really
taking root. So we’re going to talk about how
testing changes the game a little bit. We also should acknowledge that
as marketers, we’re smart people, right? We’ve been doing this for a
while, you and I, but we’re not fortune tellers. Unless we’re selling to other
product managers who have the same psychographic and
demographic profile as we do, we really don’t know what
customers want. And even if we were selling to
people exactly like us, we know that everybody
is not the same. You really only know what works
when you actually get the content out there and you
actually see if people will convert or not. The other thing to keep in mind
is that there are lots of highly paid expert opinions
out there. And they’re often influenced by
a lot of great experience and training, but they
can be wrong. And they can be horribly wrong,
sometimes, For the same reason that you and I can be
wrong about what we think is best for presenting information
to customers, even consultants can be wrong, too. Now, the best consultants will
combine their expertise and technology and tools, like
Website Optimizer. But opinions alone are just opinions, even expert opinions. The other think to keep in mind
is that what is popular today may not really
work tomorrow. So look at this button:
gorgeous, beveled edges, shiny, glassy. Just makes you want to
press it, doesn’t it? But that may not be that
attractive a few months from now as more and more people use
button styles like this. So in a way, what works today
may not really be the most successful way of designing and delivering content tomorrow. So the system we have today is
good, but it’s not great for these reasons. One, is we spend a lot of time
picking the right traffic and looking at what happens, but we
don’t really focus on where they end up. We also do our best to make the
right decisions about what information to show customers,
but we don’t know for sure if that’s really the best
way to do it. There are lots of really,
really well trained and experienced people out there,
but they’re not fortune tellers, either. And sometimes even the highest
paid experts can be wrong, if they only use their opinion. And lastly, content that is
fresh today is stale tomorrow. What other challenges
are ahead of us? The personalities
in our company. The hippo. The highest paid person
in the organization. This might be your
VP of Marketing. He could be the CEO. You might be that person. This is a very smart person,
very well educated, years of experience, knows the business
inside and out and has said, you know what, you want
to sell more product, here’s what you do. You’ve got to add these 10
bullet points; these are bullet points I use in every
sales call, when I started this company. And you’ve got to put
this table here. This is going to show how we are
positioned relative to our competitors. That’s all people are going to
need to buy the product, or fill out the lead generation
form, or download the white paper, whatever. Whatever the goal
of that page is. And then you have the
creative genius. This is a person probably not
a big fan of the hippo. May not sit near that person at
lunch, reluctantly comes to work and goes to meetings with
the hippo, but thrives in a very creative environment. Fresh thinking, went to
a great design school. He or she said, “you
know what? I know he is the owner of the
company, or he’s the senior corporate group VP, but let
me tell you something. All these bullets, that matrix
that he has, that there’s your problem. What you need to do is put this
gorgeous flash movie spot in the center that page. I’ve put all this animation
in there. I got this music. I got all these great colors. Once someone watches this five
minute movie they’re just going to be blown away.” That might be true, but
we don’t know, right? Only your customers know. What’s going to happen when the
customers see either of those two pages? They may like the more
fact-oriented, robust, matrix and strategic description of
whatever it is you’re selling. They may respond well to
the emotional buy. They may have preferred some
third idea that you had but that was overruled by these
two powerhouses. But ultimately our customers,
lots of them are going to hit the back button. They’re going to say that’s not
what I was looking for. This isn’t the information
that I wanted to see, it wasn’t presented to me in the
way that I wanted to see it. I’m going to hit the back button
and look at the next search result which is
probably one of your competitors, and they’re
gone forever. So, have no fear. I know it sounds bleak, but
Website Optimizer is going to change everything for you. Before we get into that, let’s
talk about the dollars and cents, the economics of website
marketing and website performance. How many of us spend hours, if
not weeks, driving efforts to increase our search engine
optimization? To make sure that we have all of
the right key words on our pages so we show up certain and
hide in the results, and we have our site map set up
perfectly, and all of the common practices for improving
search engine optimization. And how many of us have teams of
people every day adjusting bids and finding new key
words, and changing punctuations, and creatives. And we get all these people
to our site, right? Great job! Most of them will hit that
site and bounce. It’s not uncommon to see some
pages have 90% bounce rates. And of those who don’t bounce,
that may go on to the next page in your site, most of
them will abandon, right? And only this little slither,
which is probably 2% or 3%, will actually convert and become
one of your customers. So testing is all about
saying thank you. Thank you, web marketing team. Thank you, search engine
optimizers. Thank you, bid professionals
for getting this traffic to my site. It’s time now to convert them. It’s time not to have all of
these people abandoning the process either on the first or
the second or the third page. So the economics of testing
are irrefutable. For zero spend on your part, you
can increase your sales, because if you’re only
converting five out of a hundred now, and after testing
and finding better ways to display information and give
customers a better experience, you convert 10 out of a hundred,
that’s doubling your conversion rate, doubling
your sales. It is like Google saying to you,
you know what, for every dollar you spend on average,
we’re going to give you an extra dollar. In a way, a free testing tool
effectively does that. So, the economics of managing
traffic to your site are pretty compelling to improve the
performance of the pages and think about marketing
after the click and not just before. From a big picture point of
view, what are we trying to get at here? What we’re trying to talk about
is, yes, you must, of course, drive the right
traffic to your site. You absolutely must measure
and analyze so you know what’s going on. But after that, you have to test
potential improvements to your site, implement the
winners, and then continue. And if you do this, if you can
do this, you’re going to have an unfair advantage over
your competition. You’re going to have much better
customer experiences for future leads, and you’re
going to be rewarded internally. Because what you’re effectively
doing is you’re going to squeeze out
dramatically more revenue out of every person that hits your
site, or if it’s not sales, it’s leads, it’s whatever it is,
awareness, whatever you’re trying to accomplish
with your website. For zero spend, and you’re
going to learn a lot more about your customers and
what works and what doesn’t work, as well. And it’s lot of fun. At the end of the day, we think
of these three steps as a continuous process. And we think the companies that
do this well are going to be the companies that have
the most success online. So, no surprise. Google does have quite a number
of tools available and these three are pretty
important. These are, in a way, that
three legs of the Google Adwords tool. We have Adwords, itself, the
flagship product, which is great if you want to drive
traffic to your site. If you want to find more people
that are interested in what you have to offer, great
opportunity there. We have Google Analytics,
which is an amazing web analytics tool, free. Paste some snippet at the bottom
of each page of your site or on some template, and
you’ll start seeing data and reports and analysis that will
just blow your mind. And now here we talk about
website optimization, which is to say, I’ve driven some traffic
I look, I see the web pages are not doing well, let’s
make some improvements. Let’s increase the
conversion rate. Let’s get more people through
the funnel, and let’s give people a better customer
experience. Let’s, by the way, do all of
that using this free tool that can be set up in minutes. So here’s a typical Google
Analytics report. This is a landing page report. We look at this and say OK, not
surprisingly, most of our customers enter on
the home page. A lot of people are hitting the
contact page and the news alerts page, the email
sign-up page. Look at this. More than half will bounce. In fact, 70% of the people
that hit this newsletter sign-up page will leave. So
the question is, at this point, are you looking
at this report? What are you going
to do about it? Are you going to say, well
that’s great now I know that this page is not doing well. I guess I should run
another report. In a way, this is like Google
Analytics telling you, hey there’s a fire in your house,
and it’s burning. People leaving your site because
they don’t like the way they see whenever
they see here. So, are you going to improve it,
and save your house, and build a nice, beautiful, big,
strong house, or are you going to let this hemorrhaging
continue? Hopefully, it’s the former, or
at least by the end of the webinar you will decide
it’s the former. But let’s say you agree,
you say, yeah, I could improve that page. Then you have a meeting. And you have a conference room
full of people and someone says this page is
too darn long. It needs to be shorter. And then that person’s
nemesis, from another department, says,
“no, no, no, no. We need to provide more
information.” People aren’t seeing enough to know whether
or not they want to give up their business. Then someone else says, “ah,
no, no, no, we need testimonials that tout all of
the benefits to the user, or maybe features, or maybe how
long we’ve been in business.” And then someone else says,
“ah, no, no, this picture, this hero image is wrong.”
Alright, so you get the picture, that lots and
lots of people will have lots of opinions. Even you, yourself, have
lots of opinions on how to improve the page. Most of those ideas are
probably not right. Some of them may be good, but
it’s really hard to know which is right and which isn’t. And if it were easy, then you’d
be having a hundred percent conversion rate, which
you’re probably not. So how do I address
this problem? How do I deal with the
fact that I could make all these changes? I don’t have time to make each
one, independently, and see what happens. I also don’t believe that
that’s scientifically accurate, because I have no idea
if it has to do with some other external factor like the
time of the year that I may change 27 versus 26, right? So let’s look at the
example here. Here is the Google
Picasa home page. We drive a ton of
traffic here. We obviously advertise. A lot of people blog
about this feature. It’s a great product. Organizes all of your photos. It’s a quick download. If you haven’t tried it, I
strongly recommend it. But at any rate, the page looks
pretty good, doesn’t it? It uses the word, “free,”
always a crowd pleaser. Has a picture, so you can see,
it does all these things. A little piece of software. Very action oriented. Find, organize, share. This is a pretty good
page, right? Maybe we could have a page like
this, which has a button instead of the link. Instead of saying, “find,
organize, and share,” it says, “the easy way to share,” using
the word, “easy.” Maybe we don’t show the image at all. And as you can see, this is
page, when we tested it, increased downloads by 30%. From here to here 30% increases
in downloads. This is a good example of how
it’s hard to know that this page really is better
than this page. This could easily have been my
original, and this could have been my new page. I could have actually hurt the
conversion rate, and you don’t really know until you test it. Who would know that this one
would be 30% better than that one, but it is. The only way to know is to have
a bunch of ideas and use tools like Website Optimizer
to see which ones are the most promising. So what exactly does Website
Optimizer do? Well, here’s how it works. The visitor hits your site. There’s a bunch of people. We will then split them up and
show some portion of them one version of your page, another
portion a second version, and a third. And if you use multivariate
testing, you can have literally thousands of versions
of your page. If you use A-B, you
could have dozens. Or you could have two versions
of your page, really. Could be as small or
as big as you want. What we’ll then do is, the
visitor that is shown this page, if they then reach your
thank you page, your conversion page, your goal
page, whatever it is. Thank you for downloading, thank
you for giving us your contact info, thank you
for watching our demo, whatever it is. We will then record that and
say, “OK, Bob, saw the original, he converted.” Say,
“Jane” she saw combination one, she converted.” “Sarah saw
combination two, let’s say she didn’t convert.” And
we keep doing that. And we keep tabs. And we track, and track,
and track. And finally we say, “Wow! You know what? We showed 500 people each of
these versions of the page. A total of 1500 people visiting
your site over the last few weeks. Turns out that 20% that are
shown this version will convert and only 2% that are
shown this version will convert.” And so now you no
longer have to debate, “Oh well, is the picture with the
model behind the flowers better than the picture without
the model but the white background? Or does the green buttons for
the call to action is it better than silver links? You use data to make
your decisions. So that’s a big overview of how
Website Optimizer works. Now, enough with the theory. Let’s say you feel like,
man, I’ve been sitting through this webinar. Of course, we should
be doing testing. Of course, we should be using
science and letting our customers tell us what to do as
opposed to making decisions based on opinion. Of course, increasing conversion
rates will increase my sales and return
on investment. How I actually do it? You’ve come to the
right place. First thing you do is you go
to your Adwords account. Does this require you to
spend money in Adwords? No. Does this only work for
Adwords traffic? No. It’ll work for all traffic
that hits that page. You do have to have an Adwords
login and password, but you can set up an account just
for Website Optimizer. You don’t even have to give us
your billing information. If later on you decide that
because your pages are performing so well that you want
to consider search engine marketing, obviously we’d love
to have your business. But that’s not the point
of this webinar. The point of this webinar
is how to increase your conversion rates using
Website Optimizer. So, let’s go on to the demo. When you want to set up your
first test, the main thing you need to do is pick a page that
you want to test, tag it, and launch it. So how does that work? Before we go into the actual set
up, what’s going on here? In this example, we’re going to
do a simple A/B test. We’re going to have this
page, the A page. This is the B page. And this is my thank
you, or goal, page. You actually don’t even
need a goal page. You could make the thank you
some activity on one of these pages, like clicking
on a button. For this example we’re
going to have a proper thank you page. We have our control script that
you paste on the top. What this does is it decides
what content the visitor sees. Everybody gets directed to the
A page from your ad, or from your homepage, or however
they get to this page. This control script says, should
I put them on the rest of this page, or should I
send them to the B page? This happens in milliseconds. Then these tracking scripts
down here, they record the page views. Then the gold script, over here,
says if someone hits that thank you page, we’re going
to record that view. And then we’re going to look
and see what version of the page he or she was shown prior
to this conversion event. They could have been seeing this
page a few hours before, a few minutes before,
a few days before. As long as they saw one of these
pages and then they hit this page we can record
a proper conversion. If they hit this goal page
independent, without ever seeing any of these test pages,
that’s no problem. We just won’t count them in the
experiment and we won’t record them because they weren’t
participating in the experiment. Now let’s go for
the live demo. All right. Let me show you an interesting
example here. I have two pages. I have a
WebsiteOptimizerDemo.com/a.htm. This is obviously not a real
Website Optimizer page, but just for the purposes
of this demo. And I have Try This Free
Testing Tool, some information, some selling
points, and some call to action down here. But let’s say we thought, well
maybe we should change the color and instead of saying Try
It For Free, we need to be a little bit more thought
provoking. Should you be using
Website Optimizer? Put those selling
points up here. I think we increase the spacing
in between them. Change the image. Have a different
call to action. Instead of Get Your Account,
we say, Click Here To Learn More. So this is maybe a
softer sell page. This one is more of, hey, use
it, it’s free, get it now. Unclear which would do better. So we could debate that, or we
could set up a test. Let’s say, let’s keep our A page. This is probably our
original page. But let’s show half of
the people that try and reach this page. Let’s show them this
page instead. And then, let’s say this were
our thank you page. This is a page that we might
send users to after they create an account. You go into your Adwords
account in campaign management. You click on the Website
Optimizer link, right here. Then you say, Create
New Experiment. That will then take you to this
first page and it’ll say OK, before you proceed we want
to make sure you got all your ducks in a row. Do you have your test page? You an alternate version, and
you have a conversion page. Actually, before this step, we
would have also asked you if you want to do an A/B or
multivaried experiment. Since this is a basic webinar,
we’re going to do A/B. We’ll do another webinar later
to show you how to do multivariate. I’m good. I’ve done all this. Let me proceed. Then we say we’ll give your
experiment a name. In this case I’m just
calling it A/B demo. And then we say give us the URLs
of your original page and any alternate pages. And if I wanted to I could add
additional ones by just saying add another version add B, C,
D, as many as I wanted to. So that’s A and that’s
B. So far, so good. Then we say give us your goal
page, the conversion page. Put that there, the
thank you dot HTM. Remember that one? And we say continue. And what we’ll do is we’ll look
to make sure those pages are accessible. And then we say, OK, now
it’s time to put the tags on your page. And we give you an option. We say, do you want your
IT team to do it? If you do, you select this radio
button, and we give you this little handy link. You email that to the IT team,
give them the tags, and all the instructions. They’ll know to do. Or you can say, “No, no, no, I’m
going to do this myself.” Since this is a hands-on
demo, I’m going to be doing the self install. I’m going to say continue. And now I get the tags. So, it says OK, first thing,
for the original page, put this control at the top. Put this tracker
at the bottom. So I’m going to take control. I’m going to copy that. Here is an example of the source
code of my A page, looks familiar. Try the website tool call
to action, well picture. I’m just going to paste the
control at the top here. And for the purposes of the
webinar, I’m just going to highlight that so you can
see it more easily. Now I’m going back to
Website Optimizer. And then they say put this at
the bottom of your original. Fine. Copy that. Go here. Go to the bottom. The very bottom is fine. Again, highlighting for
your convenience. You do not need to highlight. I don’t even think you can if
you wanted to, if this were really the source code
of your page. And that’s it for the
original page. And we say, alright,
I’m good with that. What else do you
want me to do? And we say, OK, put this at
the bottom of all of your variations. In this case, the B page. If I had a C and D, I’d put it
at the bottom of those, too. So here’s a B page. Remember this one? This one says should you be
using Website Optimizer? If you don’t like the code,
don’t even worry. Just go to the bottom
and paste. Again, highlighting for
your convenience. Very simple. All I did was I put the control
on the original page at the top, tracker at the
bottom on the B, track record at the bottom. And finally we say put a tracker
on the goal page. No problem. Copy. This is the goal page. This is the thank you for
signing up, yada, yada. Go to the bottom. Paste that. That’s all. So now I’ve done that. That took me all
of two minutes. Let’s go back. You upload that to your server
so that the live pages reflect those new tags. And then you say validate. It will go and fetch all of them
and say we looked at the original control at the top
tracker at the bottom. We looked at the variation page,
tracker at the bottom. We looked at the conversion
page, tracker at the bottom. The right tags are on there. You didn’t mix up the tags
for your experiment. Everybody’s happy. You say, right on, let’s roll. Then you say continue. Now here we have a little
confirmation page, and we say all right. Here are your pages. The A, the B, here’s
a thank you. You want to preview? Sure, let’s preview B. Let’s
make sure that we’re going to look at the right stuff. Here’s my original. Looks familiar? There’s the first combination,
and I can do this for as many combinations as I have.
We say fine. Everything looks good. Now you could say start
experiment. Before we do that I want to
show you the A page live. I’m going to refresh this one. Everything looks good. Now I’m going to start
the experiment. It says OK, you’ve launched
this experiment. Everything’s good. Now, if I go back to this A page
here, live on the web, and I hit refresh, half of the
time we’re going to allocate this visitor to the A page
and half to the B page. So there’s a chance I’m going
to get assigned to the A as part of the experiment. Let’s see. Ah, look at that. Got redirected to the B page. Didn’t feel like I
got redirected. We have pretty fast
servers over here. But I did. And then you may be saying,
well, what if the person goes back to that page again? Is he going to get assigned to
some random combination? Nope. I can hit refresh all day long
because there’s a cookie on this browser now that says when
this guy loads this URL, and even if he tried to go to
the A page, he is assigned to B, to protect the integrity
of the experiment. I can open up another browser,
and I can go to WebsiteOptimizerA.htm. [TYPING] Now he got assigned to
the B page as well. Let’s do this. Let’s clear all our cookies. And then let’s try and go
back to the A page. Again, there’s a 50% chance I
could get assigned to B. In this case, I got assigned
to A. I hit refresh. You notice I’m not getting
assigned to B anymore. I’m always going to be on A. I
could clear cookies again, and there’s a 50% chance I might get
assigned to B and I’ll be contained to the B experiment. Hopefully what you just saw was
that I set up an A/B test in probably five minutes. I was going on a little bit
on the slower side. Let’s go back and say,
OK, that’s great. You guys at Google, you
split the traffic. Is that all you do? No. Give you these great reports. So in this case we would show
you, well your original was converting at this
percentage rate. We showed it to this many
people, and this many people out of that pool converted. But this other version converted
at a higher rate and had this percent improvement
over the original. And we’ll do that for all of
the versions of the page. And so now, you don’t have to
hire a team of statisticians to figure out what is the
confidence interval? What is the probability of
one beating the other? Is the experiment done yet? Does it need more time? We do all of that for you. And so when you think about
testing, what this tool does is it lets you alternate
content on your site. It stores the locations, in
this case, for A/B, the locations of that alternate
content. It assigns visitors to specific
treatments in the experiment. It then does all of cookie
management to make sure that they don’t see many different
versions and also that if they convert that we record it. And then it crunches all the
numbers and tells you which one is doing better, and which
ones are doing worse. Now isn’t that a better way to
make decisions than the hippo and the creative genius
attacking each other in the conference room? I think so. And isn’t it fun? Yeah, I think so, too. So, we talked about that
A/B test. Great for first time testers. Great for experienced
testers that want to do different layout. Great for pages that don’t get
a whole lot of traffic. Also, complements the Google
Analytics page report, so in addition to the conversion
rate, you could look in Analytics. And you could say,
well be that HTM. What was the average
time on page? What was the bounce rate? What was the average page views
per visit for people that entered from that page? How did that compare to A? And all I did was I added
control on top, tracker on bottom, tracker on bottom,
tracker on bottom of the goal page. What about multivariate? We’ll cover that in
a future webinar. That’s really fancy. Very powerful. Very robust. It requires one
more tagging step, but allows you to only work off of one page
and create hundreds of versions of it. Most people create dozens
of versions. But, you could have create
hundreds, if not thousands. You get richer reports because
in addition to which version the page will be best, we’ll
also tell you what change. We may say, well you know what,
this headline, whatever was up here was influencing
conversion rates the most. Whatever was down here had
almost nothing, no matter what you tested, down here didn’t
move the needle one bit. So you get a little bit
more precision. And so you trade off a little
bit, extra tagging step and a little bit more traffic
requirement for a little bit more precision. So, both are easily done
with Website Optimizer. So what did we talk about
so far for this webinar? I think I really want to come
back to that main point of driving traffic to your site,
seeing what’s going on, testing changes, finding
winners, and constantly redoing that until your site
just knocks the doors off the competition. If there’s one slide to
remember, if you are thinking about how to improve marketing
at your company, particularly in the online arena,
this is the one. So, some questions
to consider. One of them may be where are we
losing the most customers? Where are you driving the
traffic to your site? It may not be your home page. Maybe it’s mostly from some
advertising campaign that brings in some other page. Maybe there is a particular page
that appears highly in some organic searches. And everybody goes through that
page, and it’s not even your home page. Well, find it, and find out
what the bounce rate is. Then you think about what we
could do to make that page even better? Then think about whether your
designs are being outperformed by your competition. Is your competition testing
every page on their site, knowing what works, throwing
away what doesn’t work, while you continue to make
your best guess? Or, put it the other way, is
your competition not really aware of the power of testing? Maybe they don’t even
know that there’s a free tool from Google. Maybe you’d be the first one
in your industry or in your region to use testing and really
optimize it and get those conversions. Last thing to think about is,
do my web content tools support simple testing? Website Optimizer is designed to
work with any content tool. It could be something on your
machine, like FrontPage, Dreamweaver. It could be some very high-end
content management system that you host on your own servers. It could be some software as a
service hosted version of some content management system. As long as our tags end up on
the page when the browser renders it, everything
is good. Could be because you put it on
the template of a CMS system. Could be because you’ve pasted
it and it’s part of the source code of the page. Could be some fancy server
script that you wrote that inserts the tags as it’s
constructing the page. It really doesn’t matter. As long as the tags been up
there, you’re good to go. There are some web content tools
and content management systems that have invested some
resources in making it even easier to use Website
Optimizer. Where they will actually
do the tagging for you. You probably should just ask
your content management or your content authoring tool
provider what kind of support they have for easier tagging
using Website Optimizer. So before we wrap up some
FAQs about Optimizer. If I need extra consulting
I’m on my own, right? I mean, I must be. It’s a free tool from Google and
I know there’s no Google consulting services. No. Not true. Actually, there are Website
Optimizer authorized consultants, and I will show you
where you can find them. So, let’s go here. and let’s look at the Website
Optimizer home, this is a real home page. This isn’t the demo page,
although we do run tests on this page, obviously. But if you go to
Google.com/WebsiteOptimizer, got that?, you’ll end up here. Why is this a good page? A lot of people hit this page
when they first hear about Website Optimizer. Someone blogged about us,
writes a kind review. And they may not
come back here. But actually, if I were to
recommend a bookmark, this would be one of them. Because in addition to the demo,
the quick start guide, viewing a sample report, we also
have the testimonials, case studies, and we have
a link to our authorized consultants. And here you’ll find a number
of companies who we have certified to work with
Website Optimizer. All of these guys have run
multiple Website Optimizer experiments and they get special
training from Google and special technical support. So these guys are the Website
Optimizer gurus. You can look down this list and
you can see, you can click on here, get a little bit
more information. You can view all of them. Quickly, Conversion
Rate Experts. These guys, based in the UK,
have a great article called “101 Ways To Use To Website
Optimizer.” They have worked on some projects with really
extraordinary, off the chart, results. EpikOne. These guys are also fantastic. Remember the case study where
I showed the pictures of the flowers and things. They actually did that project
and they integrated Website Optimizer with a fairly
sophisticated e-commerce tool not provided by Google, a third
party, and had great results with the dynamically
generated content. Future Now. Literally wrote the book on
persuasion marketing. So if you feel like, “I know
how to add the tags, I just don’t know to test.” These guys
will tell you exactly what to test and they’ll help
you come up with the best ideas to test, as opposed to
just coming up a thousand random ideas. Optimost. One of the earliest
testing companies. Great company. ROI Revolution. These guys have a great article
about how to use Website Optimizer and Analytics
together, and Google Analytics together. Experts in analytics and
website optimization. Very strong team there. Site Tuners. One of the pioneers in
testing, as well. They also offer a really
interesting pricing, where you can pay for performance. They’ve probably run hundreds
if not thousands of tests in their experience. Web Share. Another great company. They recently did a case study
about how testing testimonials and images by the testimonials
can have a major impact. They really know the math
behind testing. And also experts in
web analytics. WiderFunnel, these guys up
in British Columbia. Hard core conversion
optimization testers. They really, really focus on all
kinds of advanced tests to improve conversion rates. So, you could throw a rock at
this page and any of these companies would be great
to help you out. But, it is a self-service
tool, so a lot of people prefer to do it themselves. And that’s fine. No problem. We just want to have that page
and those companies is available for those of you who
prefer to get a little bit of extra help or who say, “I get
it, but I’d rather have someone else do it for
me.” Fair enough. Another FAQ. This only works for
Adwords traffic because it’s in Adwords. Not true. We talked about that. It will work for any visitor
wherever they come from: organic, Google, Yahoo,
Ask, MSN, whatever. It works for all your traffic. “Oh, but I need to spend
money in Adwords to use it.” Also not true. You could not spend a single
penny and we’ll let you use Website Optimizer,
free of charge. All you need is an Adwords
username and password, and you just create an account
just for Website Optimizer if you want. Or if you use Adwords today,
just click on the link, like I showed you before. “I need to use Google
Analytics. I got to pare out whatever other
web analytics program I have and use Analytics. Not true. Certainly Website Optimizer
shares some of the back-end of Analytics. But, if you don’t want to
use Google Analytics, you don’t have to. You could use either no web
analytics, or somebody else’s web analytics program. Website Optimizer will happily
work in any of those situations. “I need tons of traffic to run a
test. Testing is really only for the Fortune 10,
right? ” Not true. You can do an A/B test between
two pages, and if you get a couple hundred conversions over
the life of the test, you should be fine. Another way to think about
conversion is wherever you put that goal script is
a conversion. It doesn’t have to be someone
finally purchasing something. It could have been someone starting to purchase something. Or it could be someone
going to some other page on your site. “So it’s free, so there’s
no customer support, right?” Also not true. You just go into the help
section and there is contact information there. And you fill out a little email
form and we provide free email support. “This will increase my rank
on search results.” No. This will improve customer
experiences. So this tool is all about
testing content and finding what content works best
for your customers. It is not designed, nor will it
have any special impact, on your rank on search results any
different from any other change on your site. “This is expensive. It must be. There’s all this math and all
this tagging, and cookies, and serving, and reporting, and
yada, yada.” Also not true. It’s pretty happy
to provide it. In the grand scheme of things,
it’s a win-win situation because customers get pages that
perform better, that give them what they want. Advertisers, website
owners get websites that are more efficient. And if at the end of the day
that means that some portion will consider spending money on
advertising, that’s great. Certainly Google stands to
benefit from that, but that’s not required. “Well, this isn’t for Fortune
100 sites then; it’s free. How can it handle the traffic
that such and such company has?” It absolutely is. There are a number of major
companies using Website Optimizer already. None that we can publicly
disclose for privacy reasons. But, I can tell you that at
Google will use it on the Gmail homepage. We’ve used it on the Google
Earth homepage. We’ve used it on Google Pack. We’ve used it on Picasa. We used it on Adwords
homepage. We use it on pages that get lots
and lots of traffic, and it works like a charm. Obviously we have the benefit
of plugging into a pretty robust infrastructure here and
this tool is tested for the big leagues. “I can run a test in a day,
and then I can end all discussions.” Well, I may have
been a little bit enthusiastic when I was talking about all
those conference room sessions and ending in scientific
results. We don’t recommend implementing
any results until the test is run for at least
two weeks internally. When we do our own tests,
we do at least a month. Reasons for that is you want
to let the data normalize. You don’t want to jump
to conclusions. You want to have full
day of week effects kind of smoothed out. That kind of thing. “I can just add tags and it will
tell me what to test.” Well, you got the adding
tags part right. But we won’t tell you what to
test. You come up with what to test. And then we’ll tell you
which one performs better based on customer behavior. So you have to come to the table
with ideas on– well maybe make it shorter and
longer, change this word, change that image. And then we’ll tell you
if those ideas were impactful or not. And again, if you don’t know
where to start, again, there are those authorized
consultants who– they are paid services– but they will help you a great
deal if you so choose. This only works on
static pages. I think I mentioned
this before. As long as the tags end up on
the page, could be static, dynamic, partly dynamic. It doesn’t really matter. “I need a conversion or
goal page.” Not true. You could actually, as long as
that goal script gets fired, it can be fired because you– as an example, we have a help
topic about a timer test where as long as the visitor doesn’t
leave that page within three seconds or whatever number of
seconds you set, we’ll say, that’s a goal. That’s done. So you don’t really
need a goal page. Most people use one. But it could be an action. It could be clicking on the
link on the test page. That could be the goal. Or it could be going to any
number of follow-on pages. It doesn’t have to
be one goal page. You could have an experiment on
one test and have it go to many pages. “Well, this won’t work because
my goal is on another domain.” This is called a cross
domain goal. We have a help topic on
how to handle that. The easiest way is, of course,
to pick a goal on the same domain as deep into that
domain as you can. So let’s say you have a
third-party shopping cart. Maybe make the goal the process
of clicking on the shopping cart link or putting
a product into the cart, but not necessarily being on
the cart’s domain. But if you insist on that, we’re
happy to help out in that situation, and we have
help topics about that. “This won’t work with
third-parties CMS templates.” Absolutely. Again, as long as the tags fit
on the page, it works fine. Again, there are some CMS
providers who support Website Optimizer integration
even deeper than manually adding the tags. And they’ll add the tags for
you, where you just paste that IT guy link that
I talked about. And they’ll grab the tags and
stick it on the page for you. And then it’s essentially
codeless testing. We’re almost to the end
of this webinar. Congratulations on staying
through this. I guess you’re benefiting
from a pause button. But let’s talk about some best
practices in testing. Some common mistakes I see
people make when they do their first test is they go know
really hard core, and they’re like, “You know, Tom. I got the religion. I’m going to test everything. I’m going to test 1,000 versions
of this page.” And that’s possible. Certainly we support that
scenario technically. Not advisable, though. Because in order to really
see any differences in performance, you’re going
to want about a hundred conversions per combination or
per version of the page. So if you want to do five
different pages for an A/B test, and you only had 20
conversions for the life of the test, it’s going to be hard
for us to say with any confidence which page
is doing better. Because when you say this
one had more conversions technically, but what’s the
difference between three conversions and two
conversions. It’s really not that
easy to say. So I would say start small. Rule of thumb is 100 conversions
per page. If you find that you finish the
test fast and you know a week later or two weeks later,
you’re like, “Oh man, I can test more than this.”
More power to you. Just don’t go the other
way around. Don’t start of at a thousand
and then have to go to two. Secondly, test big changes. In my example, here. Let’s go back to my example. The A and the B. Of course my
experiment’s live, I think. Here’s the A, right? Look at that. And now, let’s look at the B.
Clearly different pages. Different colors. Different image. Different layout. If you can’t see the difference
between two pages in a few seconds, your visitors
won’t either. And likely at the end of the
tests, we’re going to tell you scientifically, well,
you know what? There really isn’t a whole lot
of difference between the performance of these
two pages. So go big. Don’t try semicolon
versus comma. Try big picture of product
versus no picture of product. Or big picture of product with
10 models versus small picture of product with an arrow. That kind of thing. Consider early indicators
if you don’t have enough conversions. I’ve talked to some people that
say, “Hey, I’m selling this $10,000 software package. We sell two or three a month. And I’ve been running a test
that I’m not getting any results yet.” Well, you’re
only getting two or three conversions a month. It’s going to be hard for us
to know which page is doing better even if page A had three
conversions this month and page B had one. So instead, if you’re in that
situation, why don’t you make the goal viewing the product
details page? Or maybe make the goal remaining
on the landing page more than a number of seconds. Or maybe make the goal
requesting information, filling out a form, where you’re
going to get a lot higher data rate so that
we can actually see the difference between the
performance, if there is any. The other thing is, don’t
jump to conclusions. A lot of people get really
excited and they see one version outperform the others. And they say, “Oh, man,
I’ve got a winner. This one had 10 conversions. The other one had eight. We’re done.” That’s wrong
for a lot of reasons. But the main thing
is, a natural experiment will have noise. Things will be bumping around
up and down in what may be first place today may not
be first place tomorrow. But you know what? After two weeks, and we look
at all of the data, what in totality is first place is
probably much like more likely to be your winner. And we’ll run the numbers and
let you know if it is. So don’t be too quick to call
the election before the votes are all counted. Lastly, I showed you
this Website Optimizer homepage already. We have a link to
our user forum. Great way to help each
other and get help. Case studies. Support documentation. Demos. And again our, consulting
services. So, what did we cover today? We talked about what’s wrong,
what’s broken with how you design pages. Why you got to look before
and after the click. It’s drive, measure,
and convert. How to think about that process
of constantly getting more targeted traffic, measuring
what’s going on, and then trying ways to improve
that experience. We talked about the
tools from Google, specifically, Website Optimizer. We talked about some tests
that we’ve seen. We did a demo, showed you how
I just pasted control top tracker, bottom tracker,
bottom tracker, bottom. Right? Very simple. Did it in a few minutes. Wizard will tell you
exactly what to do. And then we went over some myths
and misconceptions as well best practices
in testing. So, appreciate you visiting. Checking us out. Hopefully I was able
to convert you. Go ahead. Go set up your first test. Do
an A/B. Just do two very simple pages. I guarantee you, you will be
very pleased with the results. And I look forward to talking
to you again soon.

7 thoughts on “Introduction to Website Optimizer Webinar

  1. Allegra Bennett Post author

    This webinar is a great idea with good information. Only drawback for me is the lengthy professorial type delivery which seems out of sync with what the web is about. Mind you, I do appreciate these tools and ideas that you give out for FREE! Love you for that. Just shorter, more to the point would be terrific.

    Reply
  2. Gail Mills Post author

    the 101 video stopped at 21 minutes. good content wish it had continued.

    Reply
  3. Marvin Franklin Post author

    You have a great Job,I do this for a living using all free sites!!! I have to pay myself everyday and I live in Waikiki. You need to make your Vids shorter and add Music. Aloha…

    Reply
  4. kryps1 Post author

    Why are you using Microsoft Word to work with your HTML? Shouldn't you be using an authentic text-editor?

    Every time I see a cool company doing cool things, I end up seeing some newb methods being performed by their staff.

    Reply
  5. Luke Stratton Post author

    Funny how a computer expert doesn't know how to make it full screen.

    Reply
  6. Judy He Post author

    this is supposte to be an intoduction but you're already talking about marketing.  What the hell is a landing page?  Why dont you start with the basics, like how to get listed?

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *