Michael Whitaker's web analytics blog

Dead links to shopping cart page still on many Yahoo! Stores

Michael Whitaker — 2009-05-27T15:32:29-07:00

Don't mean to sound alarmist, but according to Yahoo!'s very cool Site Explorer there are literally 1000s of Yahoo! Stores that still have a bad link to their shopping cart page.

A while ago, Yahoo! made a change to their checkout URLs - going from order.store.yahoo.com to order.store.yahoo.net. The actual add to cart functionality was automatically updated so no issues there, and visitors can still put items in the cart and then place an order.

The problem is if you have hardcoded a "View Cart" link in your navigation that links to order.store.yahoo.com/cgi-bin/wg-order?xyz. In this case, if a visitor puts an item in the cart and then continues browsing, the only way to go back to the cart page is by hitting the back button. Perhaps not a huge issue, but why risk annoying your visitors with a dead "View Cart" link?

In most cases, the fix should be easy even if your store was developed by someone else. Edit the Variables page and search (CTRL-F or Command-F in Firefox or Safari) for "order.store.yahoo.com". If it is found on the page just replace the .com with .net. If the View Cart link is not on the Variables page then ask the person who developed the site where the main navigation content is pulled from.

Remember to periodically put yourself in the shoes of your customers: browse your own site, place test orders, and ask your friends to do the same......because this issue would have been very hard to spot in web analytics reporting.

Reducing checkout errors using event tracking

Michael Whitaker — 2009-05-20T17:10:48-07:00

Big hat tip to Joe Megibow for providing the inspiration for this post. At the recent eMetrics conference Joe talked about how Expedia is listening to their customers and using web analytics and monitoring techniques to fix problems with their site, particularly in the checkout process. Especially for online retailers fixing and reducing checkout problems means higher conversion rates and money in the bank.

Of course the first step to fixing something is knowing that there is a problem in the first place, and customers may not always be kind enough to tell you about it. As it happens, event tracking is ideally suited to tackle this issue and track error messages that customers see as they progress through the checkout pages. All we have to do is fire off events when these various errors messages are displayed on the page. Event tracking is very flexible in that I don't have to know the actual error message ahead of time; we just send the error message itself as an action. The additional benefit of event tracking is that it is completely separate from the traditional web analytics pageview model. Just say no to fake pageviews!

Anyway, here is an example of a Yahoo! Store checkout page, but you should be able to do this in other platforms.

(Click for larger)

Not only are we seeing the various error messages (there are probably more than a dozen) that can happen, but two of them are already standing out. I will not attempt an analysis here, but the point is that we can start asking the right questions by seeing this type of data.

Since event tracking is still in beta and implementation of this technique depends on how error messages are displayed on the page I won't try to detail the steps involved, but anyone interested to learn more, feel free to contact me or ping me on Twitter/monitus and I will send you in the right direction.

Great time to have a Yahoo! Store

Michael Whitaker — 2009-05-04T15:57:38-07:00

This probably applies to small biz online retailers in general, but if you have a Yahoo! Store here are some of the things you can do at low or no cost today:

Use advanced web analytics like Yahoo! Web Analytics or Google Analytics.
Do multivariate and AB testing.
Run segmented email campaigns.
Develop multi-touch campaign attribution models.
Personalize and target your content for different visitor segments.
Tap into vast social media networks.
You can even create your own advanced voice apps.
Extract and mash up data via open API to suit your needs.

If you asked me about being able to do this a handful of years ago I would have said that only the big boys could afford to do this type of stuff. These powerful tools come with strings attached however. You have to raise your own game and data literacy to take full advantage of them. Find the people and resources who can help you learn all this.

Speed and ease of implementation of your campaigns will be key and you'll have to learn to make decisions in the face of too little or too much and often ambiguous data.

Website Optimizer and Event Tracking for e-commerce

Michael Whitaker — 2009-02-27T12:57:59-08:00

Website Optimizer is great, but Website Optimizer with e-commerce data would be better. In general, increasing conversion rates is what we are trying to do, but it has to be in context, right baby? For online retailers, a good context would be revenue. I want to know that a wining combination according to Website Optimizer actually makes money. Perhaps a "winning" combination appeals to fewer visitors, but these visitors have a higher average order size. So, a better conversion rate may not always be in line with financial goals (do I sound too serious?).

Website Optimizer will show the same combination to a visitor for up to two years. That's ok, but lets say I see combination 1 today, don't convert, come back a full year later (and my cookies are still there), don't see the test page and then convert. Website Optimizer will credit the conversion to combination 1. Is this justified? Perhaps, but I will just say that latent sales and revenue attribution are tough nuts to crack and that there may be no one right approach.

In other words Website Optimizer uses visitors as the basis whereas web analytics packages in general (Yahoo! Web Analytics and Google Analytics) use visits. (side note: the Website Optimizer reports show Conv/Visitors, but not according to this help file. Maybe just an outdated file).

So, 2 years vs 30 minutes. Where am I going with this? Well, simply put I still like visits as a basis for calculating conversion rates (as do others).

We started using Event Tracking to display Website Optimizer experiments in Google Analytics.

Events are tied to 30-minute visits, so we are no longer using visitors as the basis, but the more I think about it the more convinced I am that it is a valid approach. The advantage of event tracking is that it is independent of pageviews, so no need to worry about fake pageviews. The only thing to pay attention to is bounce rate, so we just send events to a separate profile.

This is what my experiment looks like in Website Optimizer:

(Click for larger)

The same experiment in GA:

(Click for larger)

I think I can work with this. I am not sure if GA and GWO data is updated at the same time, so this would be an obvious source of discrepancy. As you'd expect visits would be higher than visitors over time, but otherwise the numbers seem to be consistent and both combinations get the same number of visits. The conversion rates are also consistent.

This example also exemplifies my point about taking into account revenue. Combination 1 has a lower conversion rate, but a higher average value. At the time of this writing no winning combination was declared.

In summary, I think it is a valid approach to analyze website experiments in terms of visits and that Event Tracking is a great vehicle to do that. Agree/Disagree?

Happy analyzing!

Michael

P.S. We use a JS library to make event tracking calls a bit easier. We'd be happy to distribute it if there is any interest.

Ad tracking for Yahoo! Store

Michael Whitaker — 2008-10-30T12:14:04-07:00

Using search marketing ROI data as the basis for making smart business decisions (as I recommend you should) obviously requires that your conversion tracking is done properly. If you don't have the data you can't make decisions. Sounds like common sense, but it is not common practice from what I see.

If you run Yahoo! Search Marketing campaigns or Google Adwords campaigns, follow their respective instructions or wizards to grab the conversion tracking script and paste it in the Page Message field of the Order Confirmation tab of the checkout manager.

This is a the good first step and most of you do that. However, out of the box, these scripts only count conversions, not revenue. Or you can specify a fixed average order value for all conversions, which is obviously not very accurate.

Here is a tip that every Yahoo! Store merchant can use right away to not only count conversions, but also grab the actual revenue amount for each order. It takes 5 seconds to implement and is easy to do.

A) For Yahoo! Search Marketing, your conversion script looks something like this.

Just make this simple change, highlighted for clarity:

B) For Google Adwords, the script looks something like this (this example assumes that you have specified a fixed conversion value of $20):

<script language="JavaScript" type="text/javascript">  </script> <script language="JavaScript" src="https://www.googleadservices.com/pagead/conversion.js"> </script>

Just make this simple change, highlighted for clarity:

<script language="JavaScript" type="text/javascript">  </script> <script language="JavaScript" src="https://www.googleadservices.com/pagead/conversion.js"> </script>

That's it. These scripts will now use the actual revenue amount from each transaction.

If you use Adcenter, I am sure the change would be simple too. Also, instead of orderSubTotal, you could use orderTotal. I like orderSubTotal because it excludes shipping and tax, but you may prefer the total order amount.

And finally, since every Yahoo! Store uses the checkout manager it would be nice to have a simple wizard in YSM that gives you the enhanced conversion tracking script out of the box, but until then just follow these simple instructions.

Smush it! image optimization

Michael Whitaker — 2008-10-04T12:20:58-07:00

I have had the pleasure of listening to Nicole Sullivan recently about website performance optimization and the impact it has on the bottom line. Read her blog.

Nicole and her teammate Stoyan have now released a very cool Firefox extension called smush! it that downloads the images on a webpage, smushes them (ie reduces their file size) and zips them all up. Just upload those optimized images back to your website or Yahoo! Store.

Even if you think that everyone has broadband, the lighter you can make your pages the better. Some of the biggest Yahoo! Stores have very heavy page weights (we are taking in some case megabytes, not kilobytes).

I cannot imagine the increase in business they'll get if they use this tool to optimize their images. Here is an example of the homepage analysis of an Internet Retailer 500 store:

Get the extention at www.SmushIt.com

Analytics KPI gadget for iGoogle RIP and Debrief

Michael Whitaker — 2008-09-19T12:48:44-07:00

Sadly, our Analytics KPI gadget has joined the deadpool, as Techcrunch would put it. Our gadget required the use of inlining, which has been deprecated. Inlining essentially means that the gadget becomes part of the iGoogle page. This allowed us to not have to ask for usernames and passwords to access Google Analytics reports, and all you had to do was log in to GA in a separate screen. On the downside, inlined gadgets can potentially change and affect the display of the iGoogle page, so I understand the policy change. The alternative would be perhaps to create some kind of proxy server to access GA, but then usernames and passwords would have to be stored and sent in the clear. That's a big red flag for us (side note: I am amazed at how comfortable many people are entering their credentials on 3rd party sites, but that's another matter).

This was a great project to work on and learn. Widgets and gadgets are such great distribution mechanisms of content. Imagine that: with one click a gadget can be added to the personalized Google homepage, potentially reaching millions of people. The use of our gadget was of course just a drop in the ocean, but I am very grateful to everyone who used the gadget.

Here are some usage details as Google to their great credit allows tracking of gadgets via GA.

We reached approximately 10,000 visits a day (and going up if I may add), comprising of 66 different languages. Web analytics is indeed international. I was hoping we'd be able to use event tracking to see how people interacted with the gadget, but that's for another day as it's still in beta. Event tracking will be great for getting specific insights about your site, but have you thought about the possibilities for gadgets?

One of the main goals of the gadget was to keep things simple. Too much information and limited screen real estate require you to think about the metrics that matter to you. Read Jeremy's post (via Andrew Chen). Scan for abnormalities and outliers and then do a deep dive into the data to see what the causes are.

What you may not know is that we were seriously considering using a humorous approach. If you are reminded that your conversion rate SUCKS rather than down by 1.2%, maybe it will prod you to do something about it. Hat tip goes to Avinash for providing the inspiration in one of his presentations. Here are some crazy ideas we tossed around:

I wouldn't expect to see this type of approach in any official web application, but maybe this highlights the need for more API access to present data in slightly more unconventional ways.

Technorati data in Google Analytics

Michael Whitaker — 2008-05-23T16:43:50-07:00

We have updated our Analytics Fox extension for Firefox, which can now query the Technorati API and return relevant blogs for your keywords. The Technorati data integrates into the Google Analytics Keywords report.

Find out which blogs use the same keywords that drive traffic to your site. I should add that you don't have to have a blog yourself. You can find out who is talking about your organization or mentioning your products.

The extension is free of course and you don't need a Yahoo! Store or a Monitus Tools account. You do need to get a free Technorati account and a free API key, which you can get here.

For more information and to install the extension, please visit analyticsindex.com.

New radio program for Yahoo! Store merchants

Michael Whitaker — 2008-04-21T14:15:05-07:00

Shawna Fennell of 1Choice4YourStore.com has just started a new radio show and has kindly invited me to participate as a guest speaker. I will chime in regularly with web analytics tips and tricks.

Shawna sounds like a natural on radio and I wish her lots of success. The show is live every Monday at 3pm PT, but you can also get podcasts. Here is the link: wsradio.com

Cheers,
Michael

Visit history for latent transactions in Google Analytics

Michael Whitaker — 2008-04-18T15:03:06-07:00

I have been wondering about latent sales for quite some time, and how giving credit to only the last referrer leading up to a transaction could hide valuable information. While many visitors will convert during the first visit there are many visitors that take much longer to convert, interacting with your site across many visits and days. If you only give credit to the last referrer you may oversee some earlier referrers that may have at least assisted in the sales cycle.

We have updated our Analytics Fox extension to allow Monitus Tools users to view the referrer information of the visits leading up to a transaction right inside the Google Analytics e-commerce transaction report. We are also launching a Transaction Assist email report that will give merchants this information on a daily basis.

Here is an actual example.

1) The is the Ecommerce > Product Performance > Product Overview report for a particular product, segmented by keyword. The Google search "Samsonite..." (full keyword hidden for privacy) got the credit for the sale.

2) The corresponding transaction can be seen in the transactions report.

3) The extension displays the number of visits (6) and a blue icon next to the transaction number. When you hover over the icon the visit history is displayed.

This yields real insight. Not only was this visitor initially referred from another site, but the visitor also went to a comparison shopping engine before making a number of different Google searches. The last visit corresponds to the keyword that gets the credit in GA.

This example also highlights the apparent disconnect between different tracking systems. You can be sure that the shopping comparison engine will show a sale in its own reporting tool.

Hours of endless fun! At least for me...

Cheers,
Michael