Senin, 31 Agustus 2009

The Value Of Landing Pages

Imagine that we're launching a brand new advertising campaign for our new e-commerce website that sells Empanadas, my favorite food. The structure of the website is simple. We have a homepage, a few category pages that lists empanadas by type (baked, fried, etc), and hundreds of individual pages for each type of empanada (ham and cheese, steak, chicken, veggie, etc.).

Website structure

(click to enlarge)

Given this site design and our goal to sell as many empanadas as possible, let's look at this question:

Which type of landing page (home, category, or product) leads people to purchase more empanadas?

To answer it, we'll use two Google Analytics features, Custom Reports and Advanced Segments, to find out exactly, in dollars, which is the best type of page. And to perform this analysis we need one of two things: 1. e-commerce or 2. goals with a goal value.

Searching for the answer in Landing Pages
First go to the Content > Landing Pages.

(click to enlarge)

This report is naturally a good place to start but it only gives us three metrics: Entrances, Bounces and Bounce Rate. I want to know dollar amount, not bounce rate. To get the value of each landing page we have to create a custom report.

Step 1) Create the Custom Report
Go to Custom Reporting and create the following report:

Dimension: Landing Page
Metrics: Entrances, Abandonment Rate, Goal Completed and Value per visitor

(click to enlarge)

Great. Now I know the average value for any visitor that starts on these pages. On average the value per landing pages is $0.07. This means for all people who arrive at my webpage, on average each person will buy $0.07 worth of empanadas. Not much huh? However, as you can see some pages have a consistently much better conversion rate than others. For example, my home page -- /home.html -- gives me a per visit value of $0.10. I'd like to compare that with my other two page types: product and categories. We could go through this list and pick out one by one which is better, or write a regular expression in the search filter box, but an easier and more flexible way to identify these page is via Advanced Segments.

Step 2) Create the Advanced Segment
Take a minute to think about the layout of your website. Is there a unique identifier that let's you segment your landing page types? If there isn't then ask your Webmaster what you can do to get around this problem. In our example, remember that our website is very simple. Every empanada page contains the word empanada.html, every category page contains category.html, and the home page is home.html. To begin with, let's create a category segment.

Create the "Category" Advanced Segment
1. Go to Advanced Segments>Create New.
2. Dimension: Landing Page
3. Contains "category.html"
4. Name it "Visits that land on Category."
5. Save and Apply to report

Ouch! Visitors that land on my category pages spend an average of $0.04. Much worse than the average of $0.07. Now let's compare with what happens when a user lands on a page of an individual empanada product page. It's the same process as above except we use Landing Page Contains "empanada.html."

Create the "Empanada" Advanced Segment
1. Go to Advanced Segments>Create New.
2. Dimension: Landing Page
3. Contains "empanada.html"
4. Name it "Visits that land on empanada."
5. Save and Apply to report

Here is what we get:

(click to enlarge)

Wow! Visits that see a product page before anything else spend $0.30 on average. That's over 7 times more than the value of the category landing pages. Which pages should we use? Our empanada pages of course! We no longer have to guess which page is best. Even if we have hundreds of different types of empanadas we can calculate to the penny the potential value of focusing our advertisements on products.

Yeah, that's nice but how do I do the same for my website?

The above is a great example of full circle analytics. Set up goals, then create the reports and segments you best need to analyze the success of the goals. We chose to look at Landing Pages, but after you have goals, reports and segments in place, you can do most analyses.

Here are the key takeaways:

1. Most importantly your URLs must have a unique identifier (like our ?type=empanadas) so you can segment by page type AND either e-commerce implementation or a goal value.

2. Instead of thinking home, category, and product think home, broad, or specific. Usually, the more specific and focused the landing pages the better.

3. If you don't use an e-commerce website don't worry, you can do the same analysis. For e-commerce websites its much easier for us to calculate exact dollar return -- but! we can also use goal value to calculate user value. So, if you don't sell a product, your goal might be to have the users fill out a contact form. If for every 100 users that fill the form you can gain 5 leads that over a month spend an average of $100 each then the value of your form is 5x$100=$500/100=$5 per form completed. This goal value can also be used to calculate landing page value.

Now that you know exactly how to use Google Analytics to identify the value of your landing pages it's time to apply the lessons to your website. How much money do your landing pages bring you?


Jumat, 28 Agustus 2009

An API Integration To Measure Significant Change

Sophisticated, useful and cool applications are being developed everyday through the open Google Analytics API. We're loving what we're seeing. Basically, developers are grabbing their data from Google Analytics and slicing and dicing it, mixing it and mashing it with other data and applications, creating dashboards and widgets, and innovating some of the coolest stuff a data driven person could hope for. For example, we're really impressed with an app called Trendly which makes it easier to find important movers and shakers among your data via an innovative new interface, cutting down on the time you need to monitor your profiles. The team who built Trendly is using it as their one stop Google Analytics dashboard. We asked the team to share how this application came about, and here's what they wrote:
How many of you can afford to pay someone to monitor your analytics full time? We can't. We're a small startup, and we just don't have the resources to make that happen.

We use Google Analytics to track visits to our website, www.dabbledb.com. We'd love to have someone watching the hundreds of keywords, referrers, and campaigns that drive traffic to our site, someone who would send us a quick email whenever something really interesting happened: "Hey guys, thought you'd like to know that your average visitors from 'online database' doubled last week, and it's staying there - guess that SEO is working!"
So, using the Google Analytics API, we created Trendly, a monitoring and visualization tool which you can look at anytime and easily see what's changed. In short, Trendly uses mathematical models to take noisy data like this:

and figure out when significant changes have happened, marking it like this:

According to Trendly, our average daily visitors from the search words "online database" went up from 18 to 32 in mid-January, and then up again to 50 in early February. Also, Trendly sends us periodic emails to let us know about changes like these, saving us a lot of time. It also prepares a news feed with attractive charts that put the changes into perspective relative to everything else that's going on. Take a look at this - it calls out significant changes and makes them easy to notice with a timeline on the right.


When we first built Trendly for our internal use, we cobbled it together by screen-scraping and downloading exports from Google Analytics. But part of what made this tool exciting to us is that it solves a pretty universal problem. Trendly is your analyst until you can afford to hire a full time analyst. Heck, it probably keeps a clearer log of important changes than an analyst would! And with Trendly, you can delay this much longer since it cuts down your worflow by hours per week.

The new GA Data API allowed us to share it! With no signup and a couple of clicks, anyone can authenticate with Google and authorize us to grab their data and generate the reports. Suddenly our internal tool became a new product offering which can help any Google Analytics user. Give it a try and see for yourself.

What the guys at DabbleDB built is amazing.
If you have developed a useful new tool or integration on top of Google Analytics, drop us an email at analytics-api@google.com. If it's innovative and useful we'll highlight it to our readers on this blog.


Rabu, 26 Agustus 2009

New DDD Series: Bottlenecks To Implementation



Last October, Google Analytics introduced a handful of powerful new features that enterprise-class organizations had been asking for. It was a major upgrade that made Google Analytics even more powerful while remaining easy to use and free.

However, no matter what the analytics tool, there are still bottlenecks, often at an organizational level, that can prevent a company from even getting started using any type of web analytics. Nick Mihailovski, our Google Analytics Developer Relations Manager, and Avinash Kaushik, Analytics Evangelist at Google, both expert web analytics practitioners, have worked in the trenches, consulted with and had implementation discussions with scores of companies. Each website has different requirements, and each company has a different culture.

In this 3 part series of "Data Driven Discussion" videos, Nick and Avinash spend a few minutes talking about bottlenecks to implementing analytics. This first video is specifically about the obstacles that they see enterprise-class organizations confronting on the way to creating an analytics-driven online presence.

Senin, 24 Agustus 2009

Back to Basics: Tip for exporting rows


As more and more people use Google Analytics to run reports for their advertising campaigns, we've had to come up with faster and easier ways for people to use Analytics for their everyday needs. One request that comes up quite often is to do away with 500 row export limitation for reports. Understandably, it's annoying to repeat your steps when you're exporting a report that has more than 500 rows.
We've found a helpful workaround that lets you export any number of rows in one go. We've listed the steps below so that you can bookmark this page whenever you need to export all the data listed in your reports.

Instructions

In our example Keyword report, you can see that there are over 3,000 keywords to export. This would mean that we'd have to hit 'Export' over six times!


To avoid the manual labor of exporting and then consolidating all your CSV reports into one, follow the steps below:

1. Go the report that contains the data you want to export.
2. Append "&limit=5000" (or however rows you need) to the URL displayed in your browser URL window, and hit enter to reload the report.

For example:


Before: https://www.google.com/analytics/reporting/keywords....2311
After : https://www.google.com/analytics/reporting/keywords?.................2311&limit=5000


3. After you've clicked 'Enter,' visually confirm that the URL displayed in your browser has the "limit" parameter appended to it. While there won't be any visible difference in the user interface, exporting will now yield more rows.


4. Select the Export tab, and click 'CSV' (not the option that says 'CSV for Excel').



5. The exported data should contain all the rows from your Analytics table.


We hope this added some precious time back to your Monday!

Jumat, 21 Agustus 2009

Getting Started Guides For The Google Analytics API

If you're ready to try your hand and using the Data Export API, we've created some new guides to help you get started quickly and easily.

First, for the JavaScript library, here's our new super-simple getting started guide. It leads you through creating a sample application step-by-step. With this guide, you can have a working HTML page that pulls in your Analytics data in minutes (really)! Once you're done, you'll be familiar with all the key elements you need to create a basic application. The guide also shows you where to go next, since you'll be poised to customize your app.

In addition, we also have a new Authentication Guide. Authentication is one of the more difficult aspects of getting started with Google Data, and for the JavaScript and Java guides, much of the authorization complexity is handled by library methods. For that reason, we've pulled the auth stuff out of all the language guides and put it in one central doc, with relevant samples beneath each authentication methodology. We hope this makes it easier for you to:

A. Get started quickly without being confused by authentication.
B. Have a source of authentication details when you are ready.
C. Understand which authentication method to use and when.

We'd love to hear your feedback on these guides through our Developer Group. Stay tuned for a similar guide for the Java client library.

Kamis, 20 Agustus 2009

GAAC Program Goes Even More Global

It’s been a while since we’ve updated you on the phenomenal growth of our Google Analytics Authorized Consultant (GAAC) network. Over the last year, we are delighted to have added the Czech Republic, Switzerland, Romania, South Africa, Brazil, Argentina, Poland, India and Russia to the countries serviced "in-house" (by local companies) via the GAAC network. Their local business experience and of course their ability to speak the language are hugely important to helping their customers successfully deploy Google's analytics-related products.

Working with Google's worldwide network of Authorized Consultants is one of the most cost-effective investments you can make as an online business. Our "AC" partners are carefully vetted by our partner team and meet rigorous qualification standards, whether you need assistance with Google Analytics, Website Optimizer, or Urchin Software. Each offers a range of services including most or all of the following:
  • Technical implementation
  • Configuration/customization
  • Consultation/optimization
  • Training & seminars
  • Paid support with SLAs

Our global network now offers Authorized Consultants in these regions:
  • North America: USA, Canada, Mexico
  • South America: Argentina, Brazil
  • Europe: Austria, Belgium, Czech Republic, Denmark, Finland, France, Germany, Ireland, Italy, Netherlands, Norway, Poland, Portugal, Romania, Russia, Spain, Sweden, Switzerland, Turkey, UK
  • Middle East & Africa: Israel, South Africa
  • Asia Pacific: Australia, China, India, Japan, New Zealand, Singapore, Sri Lanka

Contact one of our partners at http://www.google.com/analytics/authorized_consultants.html


Rabu, 19 Agustus 2009

Does Google Website Optimizer Work On Large Sites?

You bet it does! With a resounding yes, we're proud to give a shout out to our sibling product, Google Website Optimizer, which was used successfully to run a huge, and we mean huuuuuuge, multivariate test on the YouTube homepage. Take a look at what happened on the YouTube blog. Over 1000 different recipes were tested on all US homepage visits, with great results - the new page performed 15% better than the original page.

The YouTube blog post is fascinating reading, showing screenshots of the different variables on the homepage that were tested. And for those of you working on high traffic, enterprise-level sites, you know that making small, proven improvements - thereby moving the needle by small percentages - can mean huge wins for your bottom line. It's a must-read showing the practicality of multivariate testing.

Senin, 17 Agustus 2009

Back to Basics: Troubleshooting goals

One of the easiest ways to make sense of your data and measure business objectives for your website (and even assign a monetary value to them) is to create goals. However, once you've correctly implemented your tracking code and identified the pages you want to create goals for, you may run into some common goal set-up problems. Below are some tips to help you solve some common issues first-time Analytics users run into so that you can get going on the road to taking useful action with your Analytics goals.

Incorrect goal URL
One of the most common goal set-up errors is to enter the incorrect URL. To confirm that you're entering the URL correctly, try reaching the goal page on your own site and copying and pasting the URL into the 'Goal URL' field.

You can remove the domain name entirely from the field and Google Analytics will still track the page as a goal. For example, if your goal is http://www.mysite.com/1/thanks.html, you can safely use '/1/thanks.html' as your Goal URL.

Duplicate final goal step
When you have steps leading up to your goal, the last step you specify in the Goal Funnel Settings should be the final step prior to reaching your goal. The goal itself should be placed in the 'Goal URL' field and does not need to also be included as a last step.

Goal value error
It is possible that the Goal value entered includes non-numerical characters. If the goal value is set to $2,560, the error message is displayed because the value contains the characters "$" and ",". In this case, only the value 2560 will suffice.

Historical data not relevant
If you've recently applied a new goal, it will only apply to data processed after the goal was implemented, and not to historical data in your account. This is why you might not see data for your website if you haven't had anyone convert on a goal after you created one.

To read the full list of reasons why your goal creation failed, read this Google Analytics Help Article. For an in-depth explanation about goals and how to set them up, please read this previous post.

Rabu, 12 Agustus 2009

Analytics Data In Excel Through Our API

Many of our clients use Excel to manage their ad campaigns, visualize marketing performance, and perform complex data analysis. Most analysts use the Google Analytics Export feature to manually export their report data to CSV files. Then they import the CSV file into Excel. No longer! Now, with the Google Analytics API you can bypass this manual step and export Google Analytics data directly from within Excel! Once you've set it up, there's no need to visit the Analytics reports to view data in Excel.

Thanks to a variety of developers, here are four solutions that can transform you from a reporting monkey to an analysis ninja (as Avinash would say).
  • VBA Macros - The simplest solution of them all. Mikael Thuneberg's page explains how to make API requests directly from Excel using VBA Scripts and includes a pre-built Excel worksheet to get you started.

  • The Tatvic Excel Plugin - Another easy-to-use plugin for Windows users that supports both Excel 2003 and Excel 2007. To get started you download the plugin then register to use the tool. Its simple UI helps you build complex queries and get data from Analytics right into your Excel worksheet.

    Tatvic's Plugin Query Builder

  • Excellent Analytics - Is an open-source initiative by Mark Red and Dropit. This Excel 2007 plugin works on Vista/XP and comes with a query builder to help you create Google Analytics queries and pull data right into Excel. Webanalytics.info put together a great step by step tutorial to get started using this plugin.

    Excellent Analytic's Query Builder Interface

  • ShufflePoint - Works somewhat differently than the solutions above. ShufflePoint has developed a query language that works with the Google Analytics API to achieve common tasks, such as defining the last 30 days as a date range. One then uses this language to construct an Analytics Data Export API query either by navigating to a URL within Excel, or by using their web-based query builder, then importing this data into Excel. This process allows the ShufflePoint solution to work across most versions of Excel, as well as Powerpoint, and iGoogle gadgets.

    Shufflepoint's Web Based Query Builder
We continue to be impressed by the new solutions developers are bringing to market by leveraging the Google Analytics Platform. If you have developed a useful new tool or integration on top of Google Analytics, drop us an email at analytics-api@google.com. If it's innovative and useful we'll highlight it to our readers on this blog.


Senin, 10 Agustus 2009

Back to Basics: Direct, referral or organic - definitions straight from the source

In your Analytics reports, you'll see some of the same entries come up again and again in your data tables. In the last Back to Basics post, we learned about 'not set' entries -- this week we'll learn what it means when you see 'direct,' 'referral' and 'organic' under the Sources column in your reports.

  • (direct)[(none)] - Visitors who visited the site by typing the URL directly into their browser. 'Direct' can also refer to the visitors who clicked on the links from their bookmarks/favorites, untagged links within emails, or links from documents that don't include tracking variables (such as PDFs or Word documents).

  • [referral] - Visitors referred by links on other websites. (Links that have been tagged with campaign variables won't show up as [referral] unless they happen to have been tagged with utm_medium=referral. )

  • [organic] - Visitors referred by an unpaid search engine listing, e.g. a Google.com search.

Once you learn where the traffic to your site is coming from, you can start analyzing the information to make intelligent decisions for your website. For example, the Referring Sites report shows you which websites have been most effective at driving people to your site -- and which ones haven't been effective. Furthermore, if you have defined as goals the key pages you want visitors to see, you can see the percentage of visits from each referral during which the visitor saw these pages. (Just click Goals tab to see your conversion rates for each goal.)

To learn more about how to spot quality traffic from your Goals tab, please refer to this earlier Back to Basics post.

Jumat, 07 Agustus 2009

SES San Jose Next Week!

The Search Engine Strategies San Jose conference is next week! This conference has become probably the most well-attended conference for anything related to doing business on the web. You'll see everything - booths representing cutting edge, fledgling technologies as well as the traditional online companies - and everyone, from experts in almost every facet of the web, to analysts, media and bloggers, to ad execs.

There's a ton of great sessions to attend and networking to be done, and a bunch of Googlers will be there including many of us on the Google Analytics team. We'll also have a booth where we will be demoing and answering your questions, along with Google Website Optimizer, AdWords, AdSense, Webmaster Tools, YouTube, and Doubleclick. You can register now and save 20% with the code SJ20SES.

Here are some sessions that might be of interest for you, oh ye of the analytical leanings:

Monday, August 10th, 2009
All day Web Analytics Association Base Camp - training on ways to optimize your website and what tools to use.

Tuesday, August 11th, 2009
10:00-11:30 - Always Be Testing - Bryan Eisenberg
11:45-12:45 - Creating a Web Analytics Culture - Feras Alhlou
1:45-2:45 - How to Turn Your Web Analytics into a Money Making Machine - Avinash Kaushik, Bryan Eisenberg, Jim Sterne
3:00 - 4:00 - Meaningful SEO Metrics: Going Beyond the Numbers
4:30-5:30 - Extreme Makeover: Conversion Edition - Bryan Eisenberg

Wednesday, August 12th, 2009
10:45am - 12:00pm - Convert Your Visitors to Customers (Google Site Search) - Nitin Mangtani
10:45-12:00 - Landing Page Testing & Tuning - Tim Ash
1:00pm - 2:00pm - Conference Keynote - Nick Fox, Google
2:30pm - 3:45pm - Google AdWords, Analytics & Website Optimizer Secrets Revealed + Google Science Fair
4:00pm - 5:15pm - Real World Multivariate Testing - Trevor Claiborne from Google Website Optimizer, Jim McDonald, Ayat Shukairy, David Sprinkle

Thursday, August 13
10:30-11:45 - Advanced Paid Search Techniques
2:15-3:30 - Extreme Makeover: Live Landing Page Clinic - Tim Ash


We hope to see you there!


Rabu, 05 Agustus 2009

Back to Basics: "not set" Entries

From time to time you might see a "not set" entry in your Top Content or Keywords reports. Hopefully at this point you go to the Google Analytics Help Center to do a search for "not set" to find the definition. :) If you don't, not to worry - this blog post sums up why this entry appears in your report and what you can do to prevent it from happening in the future.

"not set"
Any direct visit or referral visit will be shown under "not set" because it does not have a keyword, ad content, or any other campaign information associated with the visit. The explanations below can help you figure out why a campaign attribute wasn't collected along with the visit.

Re-tagging your AdWords destination URLs
If you are seeing a significant amount of "not set" or '(not set)' entries in your AdWords-related reports, you may wish to disable auto-tagging and instead use the URL Builder to tag your destination URLs so that they're set to the specific campaign variables you want to appear in your reports.

gclid redirection for keywords
Sometimes Analytics users are master URL taggers and still see a "not set" keyword entry. Usually this happens when there is some kind of redirection and the gclid (which is the magic that makes autotagging happen) doesn't work the way it's supposed to. At this point, we recommend you do some detective work going backwards from what you know. Start by clicking on the "not set" entry, and from the Dimension menu, select 'Source' and then 'Medium.' You can also select 'Landing Page' in the Dimension menu if 'Source' or 'Medium' doesn't give you any clues about the missing keyword. These segmenting options should help you narrow down the source of this keyword so that you can pinpoint which keyword's gclid isn't behaving properly.

For more information, search the Google Analytics Help Center (it's important to include the quotation marks!) by entering "not set" in the search box. If you know of any other tips or tricks about "not set" entries, please feel free to post a comment.


Senin, 03 Agustus 2009

The API: Sometimes You Just Gotta Explore

You may have heard about the Google Analytics Data Export API and be wondering, "What do I do with it?" Well, you may already know that you can pull most of your Analytics profile data using the Google Analytics Data Export API just by creating the right query. And then from there, the sky is the limit. Isolate, integrate, and share the data that you want to see in a huge variety of ways. Still, it's this first step that can be overwhelming. How to build a query from scratch, especially when you're used to the point-and-click experience of the your Google Analytics account's reports? To help you out, we built a visual aid to make query construction a snap: the Data Feed Query Explorer.

Query Explorer tool found in Google Analytics developer docs feed reference.

The Query Explorer is the perfect place to start exploring the Data Export API. You can use the tool to quickly get data from your Analytics account before you even start to write your first line of code. Log into the tool, select a profile, and get a display of data in seconds. From there, you can figure out just what data you want, while at the same time learning how to use the API.

By using the Query Explorer you can:
  • See the data feed request constructed for you as you select different dimensions and metrics

  • Figure out exactly which metric/dimension combination works

  • Dial in the exact sorting and filtering that you need

  • Use the permalink for any query that you build
As a first step, check out the Popular Queries tab to see what a full query looks like, such as:
We've included interactive help and links to our documentation so you can learn how all the query parameters work together. This beta version of the Query Explorer is a developer tool that we wanted to release as soon as possible to help you explore the Data Export API and troubleshoot your app.

Using the API just got a lot easier with the Query Explorer. We hope you enjoy it!