 True story, my very first week at Google, that song you just heard, Size Gangam Style, hit one billion views on YouTube. First video to ever break a billion. Every single company meeting for the next month ended in everybody doing the little horsey dance. So welcome to Google. So today I'm going to talk to you about a theme of fives. We're going to start with five tips for using analytics to drive CRO. Then we're going to look at five ways to drive conversion with personalization. And I'm going to finish with leaving you with five things that you can start doing today. I was listening to a random talk about a month ago, maybe. And I heard him say that you can't do SEO without CRO. And there's been a lot of speakers so far at Moz that have talked about search or intent and meeting that search or intent. And that's what this is about. So I want to take this one step farther and say that you can't do CRO without analytics because you need to know what's going on on your site when you're doing all of these optimizations. So by transitive property, you cannot do SEO without analytics. Sweet. We have many factors that contribute to low conversion rates on our site. So first up, user experience. Are you delivering a good user experience to people when they land on your websites? Is your site content personalized? Does it speak to the users when they come to your website? I think this goes back to that meeting your search or intent. Do you have actionable web analytics? Does your data that you're collecting actually help you to understand what is going on on your website? And finally, do you have the development resources that you need to do all of this? I know that for me personally, in my many years of actually doing analytics and optimization, this has been a real challenge to the point where I've gone on long endeavors to build server-side testing tools with engineering resources that I scrapped together to be able to get around having web dev resources. It's a problem, and it's something that you can solve with tools often, but you can also solve with grit. So the first thing I want to talk about is understanding your customer behavior. And we're going to do that with our tips for using analytics to drive optimization. So tip number one, implementation to track conversion rate optimization. If you are using a tag management system, a TMS, you likely have a data layer. And you can use that data layer to collect custom information. So let's look at an example of what this might look like. Now I'm home in my hometown of San Francisco, and I'm feeling a little lazy. I'm craving some Italian food. I go ahead and I do a search for Italian food delivery, specifically in this case from Postmates. And I click on that first result for Tommaso's restaurant. And I get taken to the Postmates website. Now this so far is a good experience for me, the end user. But for Postmates, if you look at that URL, you can see that it's just an individual URL for this restaurant. They don't have any indication of a category like Italian food or a city like San Francisco. And so in their analytics data, they're not actually able to drill in and understand these characteristics about what users are searching for and arriving to. Now I've been given permission to pick on Postmates a little bit here. My fiance actually was recently hired to drive growth and SEO there. So thank you so much for letting me pick on you. But here's something they could do. Again, they can use this data layer to collect custom information. They can have a page subcategory of Italian. And they can have a city of San Francisco. And they can take this information from the data layer and send it to a custom dimension in Google Analytics. And then they can slice and dice all of that page information, that landing page data in Google Analytics by these custom dimensions to know so much more about what their users are looking for and how they can then optimize that process for them. Another way that you want to use implementation to track what you're doing is to use campaign tagging to distinguish for variations. So you guys, if you're familiar with Google Analytics, you might have heard of UTM tracking. That's what I'm talking about here. So first case, email testing. You want to tag all of the different CTA buttons in your emails with UTM parameters. And specifically, you want to tag them with that additional parameter of UTM content to know the difference between these buttons. Let's say you have two different emails that you're sending to test headline copy. You want to make sure that you're tagging them individually to know that one was from headline A and the other was from headline B so that you can track that traffic all the way through your website. Similarly, if you're doing PPC, you're doing ad testing, use that UTM content field. I like to use it to track my headlines with my ads. That way, very clearly in Analytics, I can see which headlines were working well for my business. And finally, social media. If you're doing social media, if you're posting on many different channels, you want to have individual tracking links for each post, for each channel. Here's an example from my own blog. This is a little old, it's back in the day when Google Plus was still kind of a thing. So you'll see some posts on there about Google Plus. But what I did here was for every post, I tagged it for either being from Twitter or from G+. And in the content field, I used the content field to track the headline of the article, the name of the blog post that I was driving traffic to. And by doing so, I'm actually able to then do an analysis. Where I see that when I blog about Google Analytics, I tend to get a lot more traffic from Twitter. But when I blog about Google Tag Manager, I actually get a lot more of my traffic coming from Google Plus. I did a little more digging, and what I figured out was there's actually a very strong Google Tag Manager community on Google Plus. And my articles would be shared in this community, and that's where all that traffic was coming from. I wouldn't have been able to do this level of analysis so I would figure that out and then to start sharing my own content there had I not tagged it properly from the get-go. So tip number two is analytics goals to track CRO success. So if you use Google Analytics or really any analytics tool, you can set up goals in that tool. So prior to my role as the evangelist for Google Analytics, I spent a lot of years as a practitioner. And right before I moved to the GA team, I spent two years actually building and running an analytics and optimization program internally at Google for what was then the Google Apps for Business group, what is now the G Suite group, if you guys are familiar with that. And they had a three-step sign-up process to sign up for Google Apps for Business. You can see it here in this funnel. Step one, two, three, and the goal is that they completed this form. And I tracked every step of this funnel with goals, and here we have a goal flow. And what I love so much about this report is I can actually track my drop-off. I can see that I have huge drop-off from step one to step two here. And when I went in and I kind of dug around and tried to figure out what that might be, I realized that the message that we had on our home page when people went into step one didn't necessarily resonate from them to what they actually saw on that step one page. So we experimented with putting different messaging on that page and trying to make that a less jarring experience, and it definitely helped to improve this drop-off. But this was a good starting point for knowing where we should look to optimize first. And this is a custom report in Google Analytics. I'm looking at just some custom variables, custom dimensions down the side. Across the top, though, I have all the goals for this web business. I have goal one, that they step one of that form, step two, step three, all across the top. And what I like about this is that I actually am tracking micro-conversions, that they went from step one to step two of that sign-up funnel, and I'm tracking macro-conversion, so that they actually signed up and completed that form. And this is a great way to, again, look at that funnel in a different format, in a table format, with much more granularity. So this is the next one. They are customer errors. They're a large publisher out of Europe. And they are doing a lot of testing. And they actually count on a lot of their traffic coming from SEO. That's the majority of their traffic that they get is organic. So they really care that their user is getting that searcher intent. They're meeting the goals when they search and land on their website. And one of the ways that they're tracking that is they actually have their analytics goals set up to match what they're testing. So they're using Google Optimize, which is our A-B testing software. And what you can see here is the integration between Google Analytics and Optimize actually uses the same goals from GA that you have set up to track your business goals right to your optimization goals. So they have things that are very publisher-centric. Somebody went two pages deep on their website or three pages deep, or they clicked on a sidebar post. And these are the things that they are optimizing against. Tip number three. Your site can tell you what's important. So I think many of you are probably familiar with this report in Google Analytics. This is a site search report. You see some heads nodding. Great. So who has a search box on your websites? Are you using site search? Yes? Good. If you're not, you should. This is one of my very favorite reports in Google Analytics because it is actually your users telling you exactly what they want on your website when they're there that they can't find right off the bat. So in this case, we're looking at my blog again. I see a lot of search terms for Tag Manager and Google Tag Manager. So clearly, people are looking for content on GTM. I might want to consider blogging more about this. But I think this is especially relevant for SEO because these are the words that you should be optimizing that content towards to meet that search or intent all the way up to funnel. So this is another tool. It's called Crazy Egg. It's one of my very favorite tools outside of Google Analytics for heat mapping specifically. Don't use Google Analytics for heat mapping in page analytics. That thing sucked. Crazy Egg, though, doesn't suck. Don't tweet that, please. I might get fired. But what I really like about Crazy Egg is it actually tracks all of the different clicks on your website. And you can do a lot of different things with that information. So prior to my Google days, I worked at a company called the Apollo Group. And I was specifically working on a career portal. We spent a lot of time developing this new web page that brought together a lot of different resources for our students to help them build resumes and find jobs and find counselors and all of these different things that somebody would need as their completing a degree and getting ready to go into the workforce. So after spending so much time developing the site, we were really excited when it launched. And we started to look at the data. We had three main calls to action on this page, three big boxes right there above the fold. And we were really, really surprised to see that our three main calls to action weren't getting any clicks. We were like, what's going on? We spent all of this time developing this content. We see that we have a lot of people coming to this page, but they're not actually clicking on the things we want them to click on. Why is that? We put Crazy Egg on there. And what we found out was, in fact, people were clicking there. They were clicking a lot on these boxes, these images, of what we wanted people to go to. But my developers hadn't actually made the images clickable. The only thing that was clickable was a tiny link in the description below. And that was easily missed. So that was a huge missed opportunity that we were able to find with Crazy Egg, fix it. And we saw a huge spike in people actually then going to those resources that we really wanted to drive them to. Tip number four, key analytics reports to pay attention to. So normally when you hear about a talk about analytics to drive your optimization, you think, great. Somebody's going to tell me about Google Analytics reports the whole time. I felt like I had to tell you about a few. So that's this section. So first is our devices report. And this is just a report looking at all of your different traffic by desktop, mobile, and tablet. And that's cool, but that's not the thing I really like about it. What I like, if you look all the way to the right, is the goals. So in the conversion section here, you can see that you can actually change. You can toggle between all of the different goals you have set up in Google Analytics. And I can see that some of my channels or some of my devices are performing poorly. In this case, tablet and mobile are significantly underperforming web. I happen to know that in this case, the mobile sign up flow was pretty awful. It wasn't optimized. But I have that data now that I could take back to my teams and say, hey, we're getting a lot of people coming through here. They want to complete this funnel. They can't, because it's such a crappy experience. Let's optimize this. And we have the browser report. So here you can see internet explorer significantly underperforming our average. Now you might say, why do we care? Who's using internet explorer anymore? There are a number of people still doing that, though. If we click in, we can see that it's actually a few browser versions that are the culprit here. I happen to know that the business had stopped supporting IE8 some time before, but there are still a significant number of people that we're actually trying to go through our sign up funnel with IE8. So maybe I should use that information, take it back to my team, and say, we might want to support this a little longer. OK, site search. I already covered this, but this is a little different angle here. So this is our search term report. And across the top, we have all of our goals. And what I'm doing here is I'm actually looking at which search terms are driving conversions all the way across my funnel. So I like this. This is a custom report. You can easily build this with any of the different goals that you have set up in Google Analytics. Finally, and especially relevant for this audience, our search console integration. So here we're looking at landing pages as well as queries. And this is the integration to hook up Google Analytics to Search Console to be able to get all of that Search Console data within GA. OK, last tip in this section, qualitative surveys as zero tie breakers. So I'm a huge fan of qualitative surveys. I can't tell you how many times I have run a quantitative test and it came up flat. And I needed an answer. So specifically, back in my Google Analytics, sorry, Google Apps for business days, we worked for months on a whole site revamp. Lots and lots of people spending all day, all night thinking about this website redesign, doing a lot of work. Finally, we're ready. We test it. What do you think happens? It fails. It failed by a couple of percentage points. So we dug into our analytics data. We figured out what were some of the triggers that we could or the levers that we could pull and try to optimize this. And we did that. But also, on each version of our test pages, on the tests in the original, we put the same qualitative survey. And what we found at the end of all of this was after we made these additional optimizations, our results were pretty flat between the old site and the new site. But of course, as any marketer knows, you spent all of this time, your CEOs, your executives, they want you to go ahead and push out this new site after such an investment. And if there's not really a big difference, what's the harm? Maybe it's good, maybe it's bad. That qualitative survey saved us. It actually told us that people overwhelmingly preferred the new website. They found it easier to use. They were able to find the content that they wanted more. They liked the look and feel. It was just more intuitive to them. So we're able to confidently say, even though the quantitative numbers are flat, qualitatively, we feel good about this. And I'm not the only one to say this. I know several people have already touched on surveys being super important here. And so I just want to add on to that and say, this is something that you should definitely be considering. It's the voice of customer telling you how they are using your product. I'm going to switch gears a little bit now and move into driving conversion with personalization by providing a personalized experience to your users. So Gartner did a little study. They found that 89% of companies expected to compete mostly on the basis of customer experience, which means if you're not trying to deliver that personalized experience, you are not going to win. So again, this is a five ways section. I'm going to go through five ways of using personalization to drive conversion. So the first one is understanding the market with personalization at mass scale. And in this case, what I'm going to say is that you don't actually have to do anything to personalize. This is the Google Apps for Business website early on in my days there. We had a headline that said, Discover a Better Way of Working. And our marketing team loved this headline. They thought it was this holistic message that spoke to the suite of software tools that we were selling. We were selling Gmail and Docs and Drives to businesses. We were going to help them discover a better way of working. And engineering took a look at the data. And they said, that's a load of crap. We know that 80% of people who sign up for Google Apps for Business only ever use Gmail. So let's change the headline. Thankfully, we said, oh, let's test it first. OK, great. So this is what we tested. We had our headline of Discover a Better Way of Working against a whole bunch about Gmail. Engineering, especially like this last one, get email for your domain. Marketing really hated it. You can imagine that is not a marketing friendly message to put out in market. But marketing also said, there's no way that's going to win. That's laughable. I don't care. Test it. Put them at ease. What do you think happened? That's right. Get email for your domain. Not only one. It blew it away. 20% increase in sign ups for Google Apps for Business with this headline. I happened to be on the other side of the world for a conference when this test finished. Two weeks later, I was getting ready to go to bed. And I started getting frantic pings from my VP of engineering saying, you better launch that site. And then I got frantic pings a few minutes later from my VP of marketing saying, don't you dare launch that new headline. I was like, crap, I'm stuck in the middle. I'm not even there. I can't really talk to them real time. After several hours of negotiation, well into the night, what happened was we decided that we would go ahead and launch this new headline. Mainly because engineering said, you know that testing tool that we built for you? Well, if you don't launch our headline, we're going to stop supporting that. So we launched it. But I spent the next several hours of my night building some follow up test plans that we ended up launching within the next two days to again test a more marketing friendly version of this headline. And after a couple more weeks, what we ended up with was get Gmail for your business with subtext of and Docs Drive and Hangouts and all these other things that you can do with the suite. Couple lessons there. One, again, you don't have to do anything, but understand who your users are to personalize. And two, when you're going to test, make sure that you have sign off from all of the different parties involved. Otherwise, you too may lose many hours of sleep trying to get everybody on the same page. Understanding the market with personalization at a local level. Again, Google Apps for Business. This is our US website. And I don't know what your business is, but at Google, it's very common that you will have one marketing template that you translate for all of the locales that you want to have a website in. So this was our Japanese website. US, Japan, they look pretty similar, right? Have you ever seen a Japanese website? They don't look like this. In fact, they are busy. They are so busy. They have confetti flying all over the screen and widgets everywhere. And at me as an American looking at this, I'm like, oh, it's overwhelming. I can't even understand what the message is. But this is how they do business. This is what they like. This is what speaks to them. And so as you can imagine, this website did not speak to them. And we heard this over and over and over again from our Japanese team. And finally, we said, OK, let's test something a little bit more local. We didn't go all the way to crazy widgets all over the screen quite yet. But one of the things that they said was, hey. You see that gray button in the upper right-hand corner? Let's go back to the English version. It says contact sales. Nobody in Japan is going to click that button and submit a lead form online. That is not how you do business in Japan. We want you to, they want a hand-held process. They want to meet with the sales team and understand a walkthrough and have printouts to take back to their executive teams. And this doesn't work. So let's get rid of that button. Let's replace it with a guide. So we did. We tested this. We created this guide. It was a three-step flow, poorly translated to flow of available. They called it a guided flow of what you had to do during this 30-day trial process. And it was printable, a big print button on it. They could print it out, give it to their executives. It was much more of a Japanese way of doing business. And it killed it. 693% increase in clicks on that guide and an 8% increase in people actually signing up for a trial of Google Apps for business, hugely successful. And because we didn't learn that we needed to do things at a local level, we tried to test this in the UK and the US and it failed miserably because that's not how we buy enterprise software. We took this a few steps further and we tested many more things in a more Japanese style. So on the left was a mock that we made of various things that we wanted to test. In the middle, we wanted to do a much more local hero image. And on the right, about a year later, this is the website that we ended up launching. And it was the first time for that business that we broke the template at Google and gave a unique website to a unique audience. And it performed so much better for them. So the third level of personalization. Understanding the referral source with personalization at the traffic level. Slightly newer version of the Google Apps for business website. We had a Gmail landing page. If you clicked on an ad that led to a landing page for Gmail, you would land here. And it just had a headline of Gmail. We had an inclination that if we tried to bring over the headline of the ad onto the landing page, it might help with conversion. So we tested a headline that said, get Gmail for your business or for work. And we made the landing page say that to you. It worked. It had a very nice increase in continuation through our trial sign up flow process. And that's because we were using consistency of message. Number four. Understanding the customer with personalization at the action level. So last year, I had the opportunity to speak at the Wharton Customer Analytics Initiative conference. Really cool conference if you've ever had a chance to check it out. The best of academia with the best of analytics loved it. This was their event website. And I had gone to this website to sign up for the event as a speaker to register. And a couple of days before the conference, I came back because I wanted to check the agenda. And this is what I saw. The event is sold out. But they knew that I had actually already been here, already signed up. And so instead of sending me this kind of bland experience, they could have tried something like this. Just by knowing that I had already been there before, I'd fill it out their form, they could say, we're looking forward to seeing you there. And that would go a long way for me to feel like it was a personalized experience. Just by knowing that I had already taken action on that website. So the last level of personalization is understanding the customer with personalization in real time. So this is Light in the Box. It's an online vendor, kind of like an Amazon. And I'm on their website, and I'm searching for digital cameras. And I do a lot of research. I may even add one to my cart, but I don't check out. Now, as a good marketer, if I put on my marketer analyst hat, I'm going to go into Google Analytics. This is a report in Google Analytics 360. It's called a Custom Funnels Report. And I'm going to build this funnel. I want to understand what's going on in this process. So I see that a lot of people have visited this website. Many have looked at that digital camera page. Some may have even added to their cart, but very, very few have checked out. Pretty bad checkout conversion rate here. What's cool about this report is if I click that red arrow pointing down from any of those steps, I'm going to get a box. In this case, it says, hey, we know that 2,346 people fell off at this stage. They didn't check out. Would you like to create a segment and remark it to these people? Of course I do. I'm a great marketer. I want to bring them back to my site. So I'm going to go ahead and do that, create an audience, and remark it to those users. I can also share that same audience with something like Google Optimize. And I can make sure that not only am I giving them a message in the ad I'm remarketing to, to them with something like get free shipping on your next order. But when they come back to the site, I can then serve them a personalized experience with that same message to that same audience that says, welcome back. Get free shipping on your next order. So my last section, taking action. Five things that you can start doing today. First, make sure that you implement to track your relevant data. Remember that postmates example. Second, explore your analytics data for optimization opportunities. Third, identify your key metrics and your secondary objectives. Make sure that you're tracking them in goals so that you can use that information across all sorts of reporting. Fourth, build personalized examples, experiences. I give you five examples of how you can do this all the way from doing literally nothing but understanding who your customer is all the way down to providing a very personalized experience based on the action somebody has taken on your site. And finally, measure, analyze, and repeat. So we had a theme of five today, five tips for using analytics to drive optimization, five ways to drive conversion with personalization, and five things you can start doing today. And remember, you can't do SEO without analytics. And one last thing I want to leave you with, this is a little bit of a side note. I am very, very passionate about the women in tech, women in analytics community. I've spent the last few years of my career really promoting this. And so I just wanted to leave you with a few resources. If you're interested in analytics or data of any sorts, the Digital Analytics Association has a women in analytics group. It's an amazing community to have conversations and meetups. If you're on Measure Slack, there's a women in analytics channel. If you're not on Measure Slack, you should join it. You can join at join.measure.chat. It's a wonderful place to talk about anything in everything, measure, measurement, data, digital, related. You can start a group within your own company. I know so many people have done this. At Google, we started a women at analytics group. And we have meetups within our analytics group to get our women on the marketing side and the sales side to meet up with our engineering women and hear that tech versus business perspective and really promote each other. And we have men involved in this organization too. And it's really been a very fruitful effort to get people excited and to really build allies within our own company. The Google Analytics Handle on Twitter has created a women in analytics list. This is a list of well over 100 really excellent women in the data and measurement space. Great list to follow. Finally, I blog about this topic fairly often. I've got several posts up on my blog. And this sticker, I actually have a few of these with me. If you track me down, if you want a women in analytics sticker for your laptop, go ahead and find me. I've got a few to hand out. So thank you very much. It was really a pleasure to be here. Thank you. Thank you. Thank you. Thank you.