AB Analytics : Web Analytics and Optimisation

Far from average ecommerce conversion rate analysis - part one

Re-evaluate eCommerce Success

Like this article?  a) tell us  b) share it c) download it and d) keep updated

online ecommerce retail shopping
Evidence suggests that increasing numbers of people are venturing online to purchase goods particularly within the higher income and “silver surfer” households.  So why do so many e-Retailer continue to witness low single digit conversion rates and high cart abandon rates?

In the first of a three-part series, we explore how to re-evaluate eRetail success and identify common points of failure.

Your biggest competitor is you

Segment or repent - Conversion calculation methods used by companies are never identical so, don’t benchmark performance against your neighbour.  The best indicator for comparison is historic data from within the business.  Instead, start by asking how can we double or, triple our existing conversion rate? 

Analysis: It might be fine to use an average site-wide conversion rate for board meetings but, that figure is useless when it comes to finding real insight.  As not all visitors are equal, segment the different types to find the dynamite that will lead to an explosion in conversion performance.  Start by using the following conversion rate reporting filters:

  • Marketing Channels: {Direct, Email, Paid Search, etc}
  • Internal Promotions: {Sale offers, Shipping discounts, etc}
  • Product: {Business specific}
  • Seasonality: {Industry specific}
  • Technology: {Device & Browser}

average ecommerce conversion rate graph

Of course that doesn’t mean to say, we take our eye off the ball either.  Keep tabs on how the competition are doing by reviewing industry trends using free and paid solutions. A comprehensive list of sources is summarised at Smart Insights and a few more include FireClick Index, Alexa, Think and Compete


Watch out for ROPO cop

Add offline data - If call-center and in-store sales data isn’t integrated with web eCommerce data, any conversion rate number reported to the business is going to under value web influence.  Market trends indicate that a significant number of people (+40%) research online and purchase offline (ROPO).  Find out more useful stats over at consumer barometer.com

google consumer barometer ropo statistics

Measuring Call Centre conversion - Visitor transfer from online shop to call-center can be easily monitored using call tracking technologies. Be certain to select a provider that integrates seamlessly with your existing web analytics solution. Here are a few well known providers to consider, Response Tap, Mongoose Metrics, Marchex, IfByPhone and PowerMyAnalytics.

Manually capture offline orders - Use market research and/or in-store surveys to provide a proxy for the ROPO % in your business.  This method requires relatively little investment and will allow for aggregated offline sales (ePOS) data to be reports outside of the web analytics tool.  Apply a manual calculation using the proxy to acknowledge web influence and visualise for business users by combining the data in Microsoft Excel. 

Including offline orders automatically - Successful web analytics solutions in the e-Retail sector like IBM Coremetrics have long since provided a means by which offline data integration processing (DIP) can be achieved.  However, many platforms now also offer the ability to collect data in an automated way via an Application Programming Interface (API).  Leveraging this data transfer technique allows businesses to measure web influence more accurately and reliably. The growth API technology adoption amongst online eCommerce companies has been well documented. However, the benefits of a merged online and offline data ecosystem come with a challenge - additional transaction volumes and the visualisation of new data types.
 
Enhanced visualisation - Traditional web analytics tools offer standardised web reports that are pre-canned for speed optimisation but, hold limited value when it comes to customised conversion analysis between online and offline data sets.  To handle the vast volume and iterative data interrogation an alternative approach is recommended (that is unless you’re prepared to re-mortgage the house for Adobe (Omniture) Insight).

Desktop installed software like Omniscope, Tableau and QlikView are great options for visualisation that integrate with many analytics platforms.   There are hosted solutions too, like SumAll, Sweet Spot or iJentoAnd finally, for e-Retailer who want more control in a blizzard of data there’s Snowplow.

sumall data visualisation of external data sets

Before starting the next section, let's agree that web analytics tools should not be used in place of financial reporting tools.  E commerce order processing platforms will always be more accurate and reported numbers between the different counting technologies will never match.  However, the gap can be mitigated…



Accounting for change (revenue)

Currency exchange rate conversion - Accounting principles dictate that real-world events like currency fluctuation be incorporated when reporting financial performance so, why not reflect that in the web analytics data too.  e-Retailers operating across multiple regions already perform some form of currency conversion.  Let’s elaborate on the choices:

google analytics currency exchange rate

  1. Use an exchange rate feed provided by a third party that belongs to or, integrates with web analytics solution - daily refresh (Google, ECB, OER)

    + Low maintenance, closely tracks Global market fluctuations, more “accurate”
    - Not aligned to businesses internal rates, requires FX neutral reporting effort

  2. Use an internal FX rate manually refreshed - daily/monthly (Web Dev Team)

    + Aligned with Finance reporting
    - Human error factor more likely

  3. Use a fixed rate of exchange that is hard-coded (Clifton)

    + Easy to maintain, consistent across
    - Data mismatch with financial reporting
Clearly with so many options available, eRetailers must decide carefully on the most appropriate method.


Reconciling changes (orders)

Fulfilling orders - When items purchased online are cancelled or, modified life becomes more problematic for eCommerce site owners that are purely reliant on site metrics for conversion analysis.  Order management systems don’t match web analytics tools leading to further suspicion about data accuracy of the latter. 

To align web analytics tools more closely, consider two common methods:

reconcile transactional orders with web analytics tools
  1. Sending negative order transactions via API

  2. Uploading order modification files via FTP Server
Ask your existing vendor about what available choices as more options may be available.



Conversion metric selection

[ f = Orders / Visitors ] vs [ f = Orders / Visits ]


Re-open denominator debate - The ‘unique visitors’ metric is far from what many preconceive to be a unique person or, individual and yet it is championed as the denominator metric in every eCommerce conversion beauty pageant.  Web analytics solutions were designed to use a combination of visitor device (cookies) and browser (sessions) to determine uniqueness which is fundamentally inaccurate.  Fast-forward to today’s online shopping behaviour and that becomes even more obsolete with individuals’ using multiple devices (desktops, laptops, mobiles and tablet) and browsers (Chrome, Firefox, Internet Explorer) to browse and purchase.  Even when fingerprint authenticated internet browsing is enforced, tracking individuals’ using web technology will not be accurate and if it’s “all about the trends” as our industry rhetoric would suggest, why not just use ‘visits’ as the denominator for e-commerce conversion measurement? 

ecommerce conversion using a new scalesThis idea is certainly not an original one (Belkin) and there is prominent opposition to it too (Kaushik) but, let’s agree to disagree and consider how a business might adapt to using visits.  You’ve got a hard sell within the business to change from ‘unique visitors’ as people are familiar with the web terminology, comfortably ignorant about what it actually means and unhappy to see conversion rates plummet from 3% to 0.03%.  However, think about how much easier eCommerce conversion investigation would become.  Analysis by marketing channel or, visit and device type would reflect tracking technology counting methodology and comparing different tools or time periods would be less problematic (Longden).  Change is never easy but, if you built it, they will come.

A measure of honesty - A few years ago, whilst working at Skype, the metric used to report the health of the business to investors was the number of ‘registered’ users.  A hollow and vain metric on account of the fact that there were registered users that never bothered to actually log-in and use the product and also some people tended created multiple accounts.  Internally however, the use of ‘active users’ was championed to measure conversion - The number of people who logged in during a given period after all was a better reflection of the VOIP providers’ peer-to-peer network strength and community.  These days the Skype CEO talks openly about ‘connected’ (rebranded active) users externally and sure this metric isn’t bullet-proof either but, is at least an honest account of performance.  At the last count, Skype has +600 million registered or, +250 connected users and continues to grow at pace, albeit with a different baseline for conversion.  And finally...

Not all new visitors are welcome

Exclude the unwanted - Not everyone that visits your online store wants to buy something. Visitors may have arrived at your site by mistake or, might simply be killing time browsing the internet aimlessly.  Don’t beat yourself up too much and focus on the visitors with the most potential to convert.


uninvited ecommerce site visitors How about excluding unwanted visitors from the conversion rate calculation entirely to report a more realistic number back to the business?

Tracking single page visits or, time-on-page metrics can be misleading due to the underlying way that analytics tools track.  create a new metric to segment with…

Create a ‘timing event’ using javascript that is triggered after a visitor has spent a specified amount on time on site.  Each web analytics tool captures information differently and as an example, here’s how Google Analytics does it:

setTimeout("_gaq.push(['_trackEvent', 'read', 'ten'])",10000);

After the site code is added the “adjusted” conversion rate analysis can begin by excluding sessions by New Visitors that don’t include the timing event.

*Small print - Adjusted tracking involves collecting additional requests or ‘hits’ so it is wise to calculate any increase before implementing. If you’re an Adobe (Omniture) SiteCatalyst user, check the secondary server call quota and for Google Analytics Free users, be sure to come in under 10million/mth.

That completes the first instalment on eCommerce site conversion success. Pop back again soon to read about tracking enhancements and lost opportunities in part two.

Like this article?  a) tell us  b) share it c) download it and d) keep updated


How are you measuring eCommerce site success?

We would love to hear your opinion in the comments section below so don't be shy. 

 top 

About the Author

Alex Brown is a Digital Analytics and Site Optimisation expert who works as an independent freelance consultant.  The opinions shared in this blog are based on personal experiences gathered over a decade of data crunching and technology evaluation.  The author makes no attempt to be grammatically, politically (or otherwise) correct.  Spelling was never a strong point and for practical reasons, not all vendors in the market are referenced in the article - no hard feelings.

Related Products & Services

With in-depth knowledge of both the market place and major technology vendors, allow us to help select, negotiate, support and manage web analytics implementation for your business.  Contact us for a confidential and informal discussion to see how our solutions can improve your business today.

Resources
(In Alphabetical Order)
Adobe Omniture Insights
Alexa Competitor Site Information
Avinash Kaushik Unique Visitor Conversion Rates
Brian Clifton Converting Foreign Currency
Compete Competitor Information
Coremetrics IBM Analytics
Engaget API Growth Infographic
European Central Bank Daily FX feed
Fireclick Index Digital River Conversion Rates Information
Google Analytics Adjusting for Site Bounce
Google FX Ecommerce Currency Conversion
Google Research Consumer Barometer Information
Google Think Other Consumer Insights
IfbyPhone Phone Call Tracking
iJento Data Integration and Reporting Services
Jonny Longden Unique Visitor Conversion Issues
Marchex Call Tracking Solution
Matt Belkin Use Visits for Conversion not Unique Visitors
Monetate API Growth Infographic
Mongoose Metrics Call Phone Measurement
Open Exchange Rate FX Currency Conversion Solution
Power (formerly Pimp) My Analytics Visit to Call Monitoring Solution
Qlikview Data Management and Visualisation Software
Response Tap Website to Phone Measurement
Smart Insights Useful Sources of Market Statistics
Snowplow Versatile Digital Analytics Solution
SumAll Data Integration and Visualisation Service
Sweet Spot Information Integration SaaS Solution
Tableau Data Visualisation Software
Timing Events HTML School
Visokio Omnisciope Powerful Data ETL and Visualisation


Like this article?  a) tell us  b) share it c) download it and d) keep updated

Ecommerce image courtesy of Digitalart / Freedigitalphotos.net
Reject stamp courtesty of Stuart Miles / Freedigitalphotos.net
Spirit level courtesy of Arvind Balaraman / Freedigitalphotos.net
Alien image courtesy of Simon Howden / Freedigitalphotos.net


Comments: 1

  • Martin Livingstone May 20

    How do you see it? What do you really think makes visitors convert? That part of their brain that tells them, "ok its time to make that purchase". What triggers that offline interactive that a potential buyer wants with a company? He wants more assurance than what he's seen online. He wants something more real. And since, we companies prove to be real humans we just have to relate with them on that level. "Be Human and natural". Using call tracking services like you've mentioned (though you might need to add Ringostat - another great service) you can rightly know what particular page of your website, or Ad that triggered that inbound call and can measure your ROI with that.

Add a comment

Email again: