<?xml version="1.0" encoding="utf-8"?>
<feed xmlns="http://www.w3.org/2005/Atom">
  <title>Case Studies</title>
  <subtitle>Any real life case study on creating, improving or promoting websites.</subtitle>
  <link rel="alternate" type="text/html" href="http://www.improvetheweb.com/topics/case-studies"/>
  <link rel="self" type="application/atom+xml" href="http://www.improvetheweb.com/taxonomy/term/22/atom/feed"/>
  <id>http://www.improvetheweb.com/taxonomy/term/22/atom/feed</id>
  <updated>2007-10-04T21:09:05-05:00</updated>
  <entry>
    <title>7 Shopping Cart Tips That Even Amazon Could Use</title>
    <link rel="alternate" type="text/html" href="http://www.improvetheweb.com/7-shopping-cart-tips-even-amazon-could-use" />
    <id>http://www.improvetheweb.com/7-shopping-cart-tips-even-amazon-could-use</id>
    <published>2007-10-04T08:20:19-05:00</published>
    <updated>2007-10-04T23:22:54-05:00</updated>
    <author>
      <name>Yuri</name>
    </author>
    <category term="Case Studies" />
    <summary type="html"><![CDATA[<p>How many times have you started shopping, put a great deal of effort in entering all the required information, only to see the "You can't buy now" message. Read where Amazon has failed and how you can improve your shopping cart (scroll to the bottom, if in a hurry).</p>
<h2>The story</h2>
    ]]></summary>
    <content type="html"><![CDATA[<p>How many times have you started shopping, put a great deal of effort in entering all the required information, only to see the "You can't buy now" message. Read where Amazon has failed and how you can improve your shopping cart (scroll to the bottom, if in a hurry).</p>
<h2>The story</h2>
<p>So, I wanted to buy <a href="http://www.amazon.com/exec/obidos/ASIN/B00005R2F8/permissionmarket">a Seth Godin's book, The Bootstrapper's Bible: Volume 1</a> (Amazon fails to ship to me, so I'd eventually buy Seth's other books for a paperback price, if he had them in a digital format, too). For such a price, lower than any e-book I know - and of much greater value, I thought this was worth having.</p>
<p>(Btw, if you haven't read Seth's free e-book "Unleashing the Ideavirus", <a href="http://www.improvetheweb.com/8-ways-improve-your-e-commerce-product-page">you must, ought to, have to and are absolutely obliged to read it</a>. It is about why ideas rule the world and how you can make them spread for your benefit.</p>
<p>You can also find his other free e-books, including <a href="http://www.churchmarketingsucks.com/graphics/8.01.BootstrappersBible.pdf">a free Bootstrapping Guide Manifesto</a> (PDF, 1.2 Mb), in <a href="http://www.squidoo.com/seth">his lens</a>. Thanks, <a href="http://sethgodin.typepad.com">Seth</a>.)</p>
<p>But back to the story.</p>
<p>The button on the right, aptly named "Buy now with 1-Click&reg;", was convincing enough.</p>
<p>So, I clicked and a window appeared, asking for registering or logging in (Bad) (why not do it using the info I use in buying the book, anyway?).</p>
<div style="float:right;clear:left;">
<script type="text/javascript">submit_url = 'http://www.improvetheweb.com/7-shopping-cart-tips-even-amazon-could-use';</script><script type="text/javascript" src="http://sphinn.com/evb/button.php"></script></div>
<p>The form had my email address entered and simply wanted my password (Good - they remembered me and my email address).</p>
<p>The next window wanted my payment method, name, credit card number and card expiration date. So far so good, but it wasn't the entire "1-Click&reg;" purchase, though.</p>
<p>After clicking continue, I was hoping to get my e-book. Not so fast. It wanted my new billing information, prompting for no country, but absolutely requiring me to select a state (from the US states) (Bad)</p>
<p>When I did, it needed a zipcode, matching a state. Fun. Luckily, I remembered a 95001 zip code for California (CA), so I went through this hurdle, too. What would happen to some foreigner not familiar with the zipcodes of the US states I don't know (a lost sale, maybe?).</p>
<p>After clicking "Continue" for the 4th time (apart from clicking on the "Buy now" button), I got the following message:</p>
<blockquote><p>All Amazon Digital transactions are completed with 1-Click® to provide you with a quick and easy purchase process.</p>
<p>Your e-Document purchase will be completed when you click "Continue" and your credit card will be charged. Your purchase will be available for you to download immediately.
</p></blockquote>
<p>Talking about the easy process after 4-5 clicks seems like an attempt to make fun of me. Don't make fun of your customers (please, Amazon). (To be fair, the same message does appear when you do have the "1-Click&reg;" option enabled, after clicking on it).</p>
<p>While the usefulness of the above passage is disputable (I'd say that in order to reduce the amount of clicks, this step must have been spotlessly eliminated), I had to click the "Continue" button for the 5th time of 1 promised by the luring button. (Bad, yet again)</p>
<p>And, being pumped up during the buying/ordering/hacking process and that passage, I get the following message:</p>
<blockquote><p>
We could not process your order because of geographical restrictions on the product which you were attempting to purchase. Please refer to the terms of use for this product to determine the geographical restrictions.</p>
<p>We apologize for any inconvenience this may have caused you.
</p></blockquote>
<p>Sorry, Amazon. Five clicks instead of one I can forgive (and Yoda speak I too can), but not letting me download an e-book, is out of place. You've just lost my loyalty and joined the ranks of "Will use as the last choice only" companies that I appear to have now. (Need I repeat myself?)</p>
<h2>What's wrong</h2>
<p>Here's what is wrong:</p>
<ul>
<li>asking me to register before purchasing is no good (asking me to log in to use my already entered information is good, but you have to really use the information)</li>
<li>asking for my billing address, while having my credit card details, seems out of place, as I seem to have some info in my account (it appears the "1-Click" option needs to be turned on, but why not turn it on automatically, when the button is pressed, if the required information is present?)</li>
<li>if you are selling internationally (digital products sell this way, mostly), allow the form to accept any information (<a href="http://www.cre8asiteforums.com/forums/index.php?showtopic=45332">avoid things that annoy international shoppers</a>)</li>
<li>restricting the purchase to the USA is the worst mistake you can make. In the spirit of Seth Godin (who inspires people to share stuff), a thought to limit distribution should not exist at all</li>
<li>telling me to go back to check the terms of use for the product is one thing, but not giving me a direct link is a crime. Give me a link, don't make me go back 6 pages back and try to find it.</li>
<li>and the funniest thing is, there was not only a not obvious "terms of use for this product", but there is no such a link at all on the long, detailed page. </li>
</ul>
<h2>How to do it right</h2>
<p>So, if you want to be better than Amazon, remember this:</p>
<ol>
<li>[if you do promise a way to buy in one click] do strive to make the process obvious (for people to use it) and reduce the number of clicks/characters entered ruthlessly (such as making fewer ordering steps and required forms to be filled)</li>
<li>make your button labels obvious and true (show links how to them, too)</li>
<li>if you want your visitors to register, try to use their ordering information and make this process as easy as possible</li>
<li>if the item has "geographical restrictions", do everything to remove them and make sure the customer can get what he/she wants</li>
<li>if you still have them, make them obvious on the product page, not after 5 (or more, if you are a new customer) pages of entering information and clicking through</li>
<li>in forms, <a href="http://www.uxmatters.com/MT/archives/000107.php">place form labels above the fields</a>, not on the side</li>
<li>on the error page, provide clear ways to solve it (a link to a page with useful info will help a lot)</li>
</ol>
<p>As you can see, this list is pretty general what people say you can do <a href="http://www.improvetheweb.com/8-ways-improve-your-e-commerce-product-page">to improve your shopping cart</a>, so putting things into practice should clearly help you reduce cart abandonment rate.</p>
<script type="text/javascript" src="http://sphinn.com/evb/button.php"></script>    ]]></content>
  </entry>
  <entry>
    <title>Increase Your AdWords Conversion Rate by 126% in a Few Hours</title>
    <link rel="alternate" type="text/html" href="http://www.improvetheweb.com/how-increase-adwords-conversions" />
    <id>http://www.improvetheweb.com/how-increase-adwords-conversions</id>
    <published>2007-09-25T12:02:29-05:00</published>
    <updated>2007-10-04T21:09:05-05:00</updated>
    <author>
      <name>Yuri</name>
    </author>
    <category term="Case Studies" />
    <summary type="html"><![CDATA[<style type="text/css">
tr.search td, th.search {background:skyblue;}
.adwords tr.content td, .adwords .content th {background:lightgreen;}
.diff td {background:#ff9;}
.adwords td, .adwords th, .adwords tr {border:1px solid black; margin:0px; padding:3px}
table.adwords {border:1px solid black; text-align:center; font-size:80%;}
.adwords th, tr.total {background-color: #FFFF99;}
.adwords caption {text-align:left; font-size:120%;}
</style><p>
Though I've had the data for this post for a while, I wanted to share it along with the results of the landing page contest that ran at SEOmoz. It turned out that my page was filtered, because "it got filtered out in the process."</p>
<p>At any rate, I want to share my experience from another project and some knowledge with you.</p>
<p>Earlier, I was working on optimizing Google AdWords performance to increase software download conversions. After I had tweaked the landing page, split the ad group in two and rewritten the ads, the conversions from search ads have more than doubled. So I thought I'd share the case with you.</p>
    ]]></summary>
    <content type="html"><![CDATA[<style type="text/css">
tr.search td, th.search {background:skyblue;}
.adwords tr.content td, .adwords .content th {background:lightgreen;}
.diff td {background:#ff9;}
.adwords td, .adwords th, .adwords tr {border:1px solid black; margin:0px; padding:3px}
table.adwords {border:1px solid black; text-align:center; font-size:80%;}
.adwords th, tr.total {background-color: #FFFF99;}
.adwords caption {text-align:left; font-size:120%;}
</style><p>
Though I've had the data for this post for a while, I wanted to share it along with the results of the landing page contest that ran at SEOmoz. It turned out that my page was filtered, because "it got filtered out in the process."</p>
<p>At any rate, I want to share my experience from another project and some knowledge with you.</p>
<p>Earlier, I was working on optimizing Google AdWords performance to increase software download conversions. After I had tweaked the landing page, split the ad group in two and rewritten the ads, the conversions from search ads have more than doubled. So I thought I'd share the case with you.</p>
<p>&lt;!--break--></p>
<div style="float:right; clear:left;">
<script type="text/javascript" src="http://sphinn.com/evb/button.php"></script></div>
<p>Below you'll find a table, where the data is presented (the numbers are real, that's why the subject/client/page isn't identified). The data was gathered during two weeks (one week one group, another two groups). Below the table, you'll find details on what was changed and comments on the results.</p>
<p>Enjoy.</p>
<table class="adwords" summary="The table shows Google AdWords click, conversion and costs data. Each row shows data for either search ads, content ads or total (both) data.">
<caption>
Table: AdWords conversion increase and cost reduction due to optimization.<br />
</caption>
<thead>
<tr>
<td colspan="3">
</td>
<th>Clicks
</th>
<th abbr="clickthrough">CTR</th>
<th>Conversion<br /> Rate
</th>
<th>Conversions
</th>
<th>Cost per <br />Conversion
</th>
<th abbr="CPC">Average CPC</th>
<th>Cost
</th>
</tr>
</thead>
<tbody>
<tr class="search">
<th rowspan="3" colspan="2">Before
</th>
<th class="search">Search
</th>
<td>1022
</td>
<td>2.84%</td>
<td>15.85%
</td>
<td>162
</td>
<td>$1.30
</td>
<td>$0.21</td>
<td>$210.49
</td>
</tr>
<tr class="content">
<th>Content
</th>
<td>1567
</td>
<td>0.14%</td>
<td>8.58%
</td>
<td>129
</td>
<td>$0.48
</td>
<td>$0.04</td>
<td>$51.36</td>
</tr>
<tr class="total">
<th>Total
</th>
<td>2589
</td>
<td>0.23%</td>
<td>11.24%
</td>
<td>291
</td>
<td>$0.94
</td>
<td>$0.11</td>
<td>$261.86</td>
</tr>
<tr class="search">
<th rowspan="5">After
</th>
<th abbr="first group"rowspan="2">Ad Group 1
</th>
<th class="search">Search
</th>
<td>612
</td>
<td>3.12%</td>
<td>31.86%
</td>
<td>195
</td>
<td>$0.64
</td>
<td>$0.20</td>
<td>$124.33
</td>
</tr>
<tr class="content">
<th>Content
</th>
<td>0
</td>
<td>0</td>
<td>0
</td>
<td>0
</td>
<td>0
</td>
<td>0
</td>
<td>0</td>
</tr>
<tr  class="search">
<th abbr="second group" rowspan="2">Ad Group 2
</th>
<th class="search">Search
</th>
<td>489
</td>
<td>4.13%</td>
<td>40.90%
</td>
<td>200
</td>
<td>$0.46
</td>
<td>$0.20</td>
<td>$91.00
</td>
</tr>
<tr class="content">
<th>Content
</th>
<td>0
</td>
<td>0</td>
<td>0
</td>
<td>0
</td>
<td>0
</td>
<td>0</td>
<td>0</td>
</tr>
<tr class="total">
<th colspan="2" rowspan="1">Total
</th>
<td>1101
</td>
<td>3.60%</td>
<td>35.88%
</td>
<td>395
</td>
<td>$0.54
</td>
<td>$0.20</td>
<td>$215.33
</td>
</tr>
<tr class="search">
<th colspan="2" rowspan="3">Difference
</th>
<th class="search">Search
</th>
<td>+79<br />(+7.7%)
</td>
<td>+0.76% <br /> (+26.76%)</td>
<td>+20.03%<br /> (+126%)
</td>
<td>+233 <br /> (+162.35%)
</td>
<td>-0.66 <br /> (-50.77%)
</td>
<td>-$0.01 <br /> (-5%)</td>
<td>+$4.84 <br /> (+2.3%)
</td>
</tr>
<tr class="content">
<th>Content
</th>
<td>-1567 <br /> (-100%)</td>
<td>-0.14% <br /> (-100%)</td>
<td>-8.58% <br /> (-100%)</td>
<td>-129 <br /> (-100%)</td>
<td>-$0.48 <br /> (-100%)</td>
<td>-$0.04 <br /> (-100%)</td>
<td>-$51.36 <br /> (-100%)</td>
</tr>
<tr class="diff">
<th>Total
</th>
<td>-1488 <br /> (-57.47%)
</td>
<td>+3.37% <br /> (+1465.22%)</td>
<td>+24.64% <br /> (+219.21%)
</td>
<td>+104 <br /> (+35.74%)
</td>
<td>-$0.40 <br /> (-42.55%)
</td>
<td>+$0.09 <br /> (+81.81%)</td>
<td>-$46.53 <br /> (-17.77%)
</td>
</tr>
</tbody>
<tfoot>
<tr>
<td colspan="3">
</td>
<th>Clicks
</th>
<th abbr="clickthrough">CTR</th>
<th>Conversion<br /> Rate
</th>
<th>Conversions
</th>
<th>Cost per <br />Conversion
</th>
<th abbr="CPC">Av. CPC</th>
<th>Cost
</th>
</tr>
</tfoot>
</table>
<p>(the increase percentage did not include the initial 100%. Thus, +20 from 100 is +20%, not +120%)</p>
<h2>Details</h2>
<h3>Before</h3>
<ul>
<li>the "Before" data is taken from the week before the test was started</li>
<li>the conversion was a software download</li>
<li>the initial landing page has a 3 field form</li>
<li>the page listed features, along with a thumbnail of a screenshot, system requirements, etc (no benefits, though)</li>
</ul>
<h3>After</h3>
<ul>
<li>the "After" data is taken after a month has passed, during which the tweaks were done to the AdWords campaign</li>
<li>the landing page listed the the benefits the owner of the software would get</li>
<li>beneath them, shown were the ways how the software can help the owner and his customers to get the benefits</li>
<li>to download, the visitor would need to click on a <span style="font: 14pt Verdana; text-decoration:underline; color:blue;">Verdana 14pt blue underlined link</span> (did you hover over or click it? it isn't a live link)</li>
<li>after the link, file size was shown (11Mb)</li>
<li>becides the header, the above described contents and the footer, the page didn't have any other elements (a more obvious click path)</li>
<li>during the test, content network was disabled, because it yielded absolutely no sales</li>
<li>in the AdWords campaign, an ad group was split in two to focus on two distinct kinds of visitors and keyword groups</li>
<li>along with the landing page, the ads were changed to reflect the benefits and the relevant keywords</li>
</ul>
<h2>Commentary</h2>
<ul>
<li>search clicks have increased due to a slight increase in CTR, most likely, since the search audience was the same</li>
<li>removal of content network resulted in:
<ul>
<li>less clicks (sales only went from search clicks, though)</li>
<li>relatively low total conversion increase, due to downloads from content network before the test</li>
<li>drastic total increase in CTR, since people don't click on content network ads a lot</li>
<li>increase of total CPC even though search PPC dropped by 5%</li>
<li>a total decrease in cost by 17.77% even though search costs upped by 2.3%</li>
</ul>
</li>
<li>conversions still increased by 35.74% during removing content network, which means how <strong>powerful were the search clicks</strong> (+162.35% extra conversions from search)</li>
<li>overall, the pattern is this:
<ul>
<li>search clicks remained the same</li>
<li>conversion rate more than doubled</li>
<li>optimized search beat old search and content performance</li>
<li>cost per conversion was reduced almost by half (-50.77% in search)</li>
<li>search CPC even got lower, reinforced by increased CTR and conversions rate</li>
<li>even though search cost has slightly increased (2.3%), total costs went down by 17.77%</li>
</ul>
</li>
</ul>
<p>So, to me, this means that the generally known tips to increase conversions and reduce costs work:</p>
<ul>
<li>make it easy to convert is the best way to increase conversions (who knew?)</li>
<li>focus ad groups, keywords and ad groups around very specific subjects/keyphrases</li>
<li>us keywords in ads, too, along with call to action</li>
<li>explain the benefits of the conversion (a page with plain added info was tested and didn't perform as well as a page w/o it)</li>
<li>test, test and test some more to increase maximum conversions</li>
</ul>
<p>Read more about <a href="http://www.improvetheweb.com/eight-tips-on-how-to-improve-your-adwords-conversions-and-bid-costs-at-a-time">increasing AdWords conversions and reducing costs</a>.</p>
<p>However, here's the thing.</p>
<p>The amount of attempted orders stayed the same or even dropped by a small margin (less than 10%).</p>
<p>This means that just tweaking the landing page and the ads won't help you sell if the intermediate conversion isn't a sale, but it will help people to go <a href="http://www.improvetheweb.com/how-visitors-view-websites-design-information-scent">the conversion path</a>.</p>
<p>To sell, you need to have all the information a customer might need on the website and an easy way to order the product, coupled with things that <a href="http://www.improvetheweb.com/trust-as-the-most-important-online-value">increase trust</a>, such as customer testimonials, trust signs (there are many for various industries) and so on.</p>
<h2>Rounding up</h2>
<p>Increasing intermediate conversions may not help you with sales, if you don't make it easy and convenient to learn about your product and buy from you.</p>
<p>However, if your conversion is what you need and the visitors can get it from the page (free book download, for example), then even just landing page and ad optimization work like a charm.</p>
<p>Read the <a href="http://www.searchenginejournal.com/5-steps-to-increase-the-efficiency-of-your-adwords-ppc-budget/5378/">tips to improve AdWords conversions and costs</a> from Search Engine Journal.</p>
<script type="text/javascript" src="http://sphinn.com/evb/button.php"></script>    ]]></content>
  </entry>
</feed>
