Segmentation and personas are areas of interest to traditional marketing departments. But now with web analytics, consumer segments can be set up and analysed real time and/or historically with many of the standard web analytics tools such as Google Analytics, Omniture’s Site Catalyst or Discover, Clicktracks or Webtrends.

If you look at the average conversion rate for your whole website, on it’s own although useful, would considerably more insightful if you could filter the effectiveness of different segments / people / campaigns – see the conversion rate of .
What do I mean by a consumer segment?

It could be PPC (pay per click) campaign traffic, or seo traffic, or email campaign traffic. Or a segment where people clicked on a banner or landing page on the website.

You can then easily compare the average conversion rates of these different segments and also bear in mind actual amount of traffic from each segment. For example, email campaigns might have a conversion rate of 4.2% compared to PPC traffic with a 2.6% but there might be significantly more PPC traffic than email traffic in which case, better to spend one’s time and efforts optimising PPC than email campaigns.

I’ll post some more on segmentation later this week. Thanks so much for reading!

Today at econsultancy’s digital cream, digital marketers, social media – ites, search – ites, and web analytics ites thronged together and debated. It was actually fascinating and informative so a big thank you to econsultancy for organising it and to all the staff for being so helpful. ( I just got back from just over two weeks holiday in Playacar and in the yucatan in Mexico – the photo is Tulum).

I went to the web analytics roundtable in the morning and the social media roundtable in the afternoon. At each, there were 7 senior marketers and a moderator all talking out their strategy, sharing ideas and revealing what they want to do. To the extent that competitors were even opening up and talking quite honestly about plans and site features that hadn’t even been launched yet.

As web analytics continues to shift into marketing optimisation, the big solutions like omniture and web trends differentiate by offering integrated multi-variate testing, behavioural targeting and integration in and out with external email solutions or databases so you only need one place to your fully integrated picture of what’s going on with your company,
Buy a web analytics tool first or hire the skilled web analytics person? Better to hire the right person first and use a free tool than spend the money on the tool without having the right person to get the analysis and insights out of it.

Social media is so fast moving. Yet for many companies there is still caution and worries about security, brand awareness and even how to measure it’s impact.

I know I’ve missed things out so get in touch if you have any thoughts about the event.

8 conversion rate tactics below to help increase conversion rates on your website.
1. People Click On What They Want

People navigate the web by “scent”, Byran Eisenberg, conversion guru and persuasion architect, tells us. Scent was first described by Xerox PARC to describe the parallels between a human’s information-gathering techniques on the Web and an animal’s food-gathering techniques in the wild. People seek information through the “scent” given off by their trigger words.

According to research performed by usability guru Jared Spool, when visitors found their trigger words on a landing page, they were successful at completing their task 72% of the time; if the trigger word wasn’t on the page, they were only successful 6% of the time. The scent of the keywords kept them on the right path; lacking that scent, they stopped searching that particular “trail”. One tip to make sure you have your visitors’ trigger words covered is to make sure each major button or link:

completes this sentence: “I want to _____”.
includes trigger words / strong scent

2. Start Using Persuasive Call To Action Words

Impotent call to action hyperlinks like “read more” and “submit” sometimes make me feel embarrassed for website owners. They should know better.

Persuasive call to action hyperlinks should include an imperative verb and a benefit. For example, which hyperlink is more persuasive: A or B?

George found an investment secret that changed his life. Read More
George found an investment secret that changed his life. See how George doubled his income in one year.

You can see from this comparison why the second example is more likely to induce action.

3. Better Product Images Are Worth A Thousand Calls to Action

Having better-looking product images than other sellers will do wonders. If research is any indication, product images are a major factor in converting visitors. In fact, 83 percent of eBay shoppers skip listings without images, while sites with galleries get 15% more activity and those with so-called super-size photos show a 24 percent spike in sales. The better photo wins every time. Many people skimp on the quality of their product images and use manufacturer-supplied images which is a mistake.

4. B2B Products or Services Need Merchandising, Too

The same holds true if you are in B2B: Better product images are worth a thousand calls to action. Many B2B sites offer downloads of whitepapers or demos in exchange for completing a form, but fail to make the most basic of efforts to persuade visitors to do so. Don’t just tell them about your whitepaper… merchandise it. Show a cover, show them how easy it is to read with all your pretty charts. Test to see which pieces matter the most.

5. Headlines Must be Made to Stick

Most headlines (and copy for that matter) suffer from what Chip and Dan Heath refer to in their book Made to Stick as a curse of knowledge: Once you know something, it’s difficult to imagine what it is like to not know it. The headline on your page is the one thing that about 80% of your visitors will read. But while headlines are often crafted for their persuasive abilities, they often assume too much prior knowledge on the part of the reader. Make sure that everybody understands what your headline is about, even if they have no reference to understand it. Then invest as much time as possible testing your headline’s abilities to both (1) gather attention and (2) entice visitors to invest the next 30 seconds on your page by explaining what’s in it for them — in language they can understand!

6. Always be Testing

Doing A/B or multivariate testing used to require some in-house programming expertise or expensive third-party software. Thankfully, Google has provided us with a free alternative in the form of Google Website Optimizer. While it may not offer every feature some of the other solutions provide, it is quite an elegant solution and getting better all the time. I actually prefer that people don’t spend their money on a tool, but focus those resources on better copy and imagery instead. There are no more excuses for not testing regularly. Remember what Claude Hopkins wisely said in 1923: “Almost any question can be answered cheaply, quickly and finally, by a test campaign. And that’s the way to answer them – not by arguments around a table. Go to the court of last resort – buyers of your products.”

7. Should we be testing hundreds of thousands of variations?

This question illustrates the market’s misunderstanding of testing. For the vast majority of businesses, this is more like random testing. You can test thousands of combinations in a multivariate test, but being able to doesn’t mean you should. Let’s focus on this example. I’ve kept the numbers simple for clarity’s sake, but let’s assume:

Example I (not recommended):

1,000 = Test combinations (the number of page sections and variations in the test)
10,000 = Page views per day
100% = Visitors in experiment (we’ll run the experiment with all our traffic)
2.4% = Current conversion rate (average conversion rate)
20% = Expected improvement
The duration for this test: 34.9 days. (More than a month!)

Example II (recommended):

20 = Test combinations (focused on key drivers)
10,000 = Page Views per day
100% = Visitors in experiment
2.4% = Current conversion rate
20% = Expected improvement (focus on key drivers in the hierarchy of optimization rather than just random elements, and your expectations should be higher)
The duration for this test: 0.698 days. (Under a day!)

Under the guise of being “scientific”, the companies that originally offered these tools charged on a monthly basis. While they had plenty of experience in managing their software, they had little experience in identifying valuable tests. Plus, they had zero incentive to get quick results while customers paid a monthly fee.

Multivariate testing for the sake of conversion rate optimization should be scientific. However, testing is about improving your business results, not scientific experimentation. Unless you’re running a lab, you’re testing for profit. (No offense, non-profits… yes, you should be testing too.) Testing only what matters is how to recover opportunity cost. Time is money. Don’t waste it by testing which variables matter; rather, invest your time in improving those variables and your understanding of them. Fix the things that hurt your conversions as fast as possible, and make more money today.

8. Read the Reviews on Conversion

Reviews have been all the buzz the past couple of years. If you recently purchased something online, has a review influenced your purchase decision?

New research further illustrates the value of reviews:

77% of online shoppers use reviews and ratings when purchasing (Jupiter Research, August 2006)
63% of consumers indicate they are more likely to purchase from a site if it has product ratings and reviews. (CompUSA & iPerceptions study)
86.9% of respondents said they would trust a friend’s recommendation over a review by a critic, while 83.8% said they would trust user reviews over a critic. (MarketingSherpa)
Most people don’t seem to focus on all the factors involved in implementing reviews to enhance conversion. It’s important that you test and optimize for conversion and persuasion by focusing on the following areas:

Placement for Visibility
Above the fold
Size
Stars or other graphic
Near point of attention or action
Review Interaction

Ease of reading
Sorting
Rating Distribution
Use across the site
Single Dimension versus Multi-Dimension Reviews

What are the key attributes across different categories
Can review content influence purchase decision
Credibility Factors

Negative and positive reviews
Review approval policy
Reviewer characteristics
What Does a Review Mean

Number of reviews
What questions are you asking
Qualitative versus quantitative
Reviews are just one example of the market trend demanding more authenticity and transparency, and they are key factors in getting your visitors to take action. Any time you have a choice between opening up more or less, always opt for giving your customers more.

What do you think? Do you have any ideas on how to make your website perform better?

If you are a company with a sizeable email marketing budget and list you will be interested in increasing sales or increasing leads. But what should you key performance indicators be?
Some choices for a ecommerce-driven business are:

Sales (total sales, sales by campaign/type/segment/cell, per e-mail delivered)
Profit/margin per e-mail sent and campaign level
List growth/churn (size of database, growth month over month, churn rate month/aggregate)

Some companies are more about lead generation than pure sales. They want to get a lead into the hands of a partner or call centre or salesperson–so they are about efficiencies, rather than sales and/or profit margin. There choice could be:

1. Cost per lead
2. Funnel movement (how leads migrate through the funnel stages)
3. Total reach (how much of the prospect database can be reached at any given time). This includes churn metrics as well.

Notice that the primary KPIs for the two instances above are not about click-through rate or open rate.

When the real focus of the program is purely branding, the metrics seem simpler, but with a deeper definition:

1. Response rate (both open and click). Brand marketers are increasingly looking at the total reach of their base and less at campaign-over-campaign results. They want to count how many eyeballs were exposed to the brand each month, the resulting interactions and incremental Web exposures.
2. Site visits and site journeys. Site traffic statistics driven from e-mail are becoming increasingly valuable to see what content and interests are driven from each campaign/program and business division. Depth of site exploration… List growth. See how email traffic performs against searches made, types of products bought so you can define your personas against email marketing.

All brands place increased value on the size and relative growth of the database since it represents a potential share of market and voice. It is essentially the lifeline to reach a mass of customers in an efficient manner. This, combined with qualitative feedback, is quite useful in measuring brand attitude, awareness and level of involvement with the brand.
So also for example in this context thinking about how to encourage higher subscription rates to an email newsletter via personalised call to actions based on site activity or cookie id to move more people from your bottom left to top right (great customers).
Here are some standard KPIs but obviously every company’s KPIs will be different and will need to evolve over time as well.

E-mail KPIs:

E-mail sent/delivered
Campaign over campaign impressions/opens (total/unique/% rate)
Campaign over campaign clicks (aggregate/unique /% rate)
Click to open (% who open the message and click through)
Churn rate (% unsubscribe/opt out/undeliverable)
Send to a friend (viral rate)
ISP domain response (open/clicks)

Commerce-Driven Program KPIs:

Total sales (campaign, month, quarter, segment)
Profit/margin per e-mail sent / source
Sales per e-mail sent
Cost per e-mail sent
Average Order Value (AOV)
Number of orders
Conversion rate (to open/click through)
Number of site visits (page rank)

Demand Generation

Number of leads (by product/client type)
Number of downloads
Site visits
Page Rank
Cost per lead
Conversion rate
Membership/subscription/database growth

Web Focused

Cost per visitor
Number of leads by entry page
Heavy user share
Average page views
Stacking effect
Length of site visit by source

What do you think? Have I missed some that you think glaringly obvious…
Thanks so much for reading

Say for example, we would like to compare the performance of three different marketing campaigns against one another, by which I mean organic, ppc, online banner advertising, we would want to analyse the conversion rate of the campaigns over time.

First of all, we would take the average conversion rate on a daily basis, by average I mean the central tendency for our data to centre around a particular value, be that mode, median or mode. And then compare these average conversion rates over time for each individual campaign against one another.

However, when we are looking at how the data is dispersed, we may see some points that either stick out too high or too low. So how can we tell whether these are significant data points. If we look at all of the raw data that makes up the averages, it may be that there is one number that is so off the charts that it is completely skewing the whole average – these are called outliers.

It is important to get rid of outliers, otherwise you would have no idea whether or not one campaign is performing better than others.

I will post more about this later in terms of my key take aways overall. But for now, thanks so much to Rene and Aurelie for their interesting event Web Analytics day and their hospitality. Here are some take aways from the event.

Ian Thomas from Microsoft showed us some slides from Gatineau which will be launching in a couple of weeks. Click here if you would like to sign for Microsoft’s beta invite request.

The key USP of gatineau is the ability to segment your visitors by age and sex demographic from Live passport holders who visit your site (there are 30 million live passport holders so for your website this is likely to be a statistically significant segment). This demographic data is apparently completely anonymous. Plus some nice zooming in. As to whether this become a significant player in the market, like Google analytics, only time will tell. I do have some nice photos of the screen shots but will have to put those up later.

Actual web analytics client casestudies – I always think that actual results from actual clients are the meat on the bone, for example Eric Petersen’s presentation, Web Analytics is hard.

1. Top online retailer improved site design as a result of web analytics and multi-variate testing.
Analyst finds that visitors using site search are better customers.
Analyst recommend improving visibility of internal site search.
Controlled testing of different visibility and position of internal search within site
One successful experiment returns a 2% increase in visitors using site search.
Result: a six figure lift in weekly revenue, equal to a nearly 1% increaase in total company revenue over a year.

2. A well known brand analysed website to isolate a specific group of users on the web site.
Analyst shows that “product compare” functionality is under utilised.
Visitor segmentation is used to isolate visitors using “product compare” functionality.
Analyst discovers this segment has a 33% higher average order value.
Changes made bsaed on this analysis increase traffic to “compare product” functionality by 11%, increase purchases by 16% and reduce exits from these pages by 56%.
Result: an annual increase in revenue of $2.2 million

3. A Fortune 500 travel company used web analytics too and made millions by improving their messaging.
Analyst speculates that prospects were not seeing “best price guarantee” information.
Concerns were confounded by multiple messages and placements throughout site.
Multi-variate testing was used to test different messages and treatments.
The winning format was shown to convert at a rate of 0.6 points better than the control.
Result: an estimated life in online bookings of $30 million annually.

Sometimes in web analytics, we get far too concerned with which tool should we use, which theory or thought leader are we following. Ultimately, we are in the business of making more money for our clients or our own website, be that in increased sales or increase lead generation – which is why I gain so much from hearing exactly how and why people are getting the results they are for their websites.

Thanks for reading and please do comment if you have thoughts for web analytics princess

This is a different type of post from normal because there will be lots of graphics instead of words. I am also ill at the moment which is really annoying. The question being, which online marketing banner is more effective, from a web analyst’s perspective of course, and which other bits of data or charts would you have liked to have had?

Bubble charts key (I like bubble charts because you can compare 3 datasets at the same time):

Size of bubble determines number of visits in (000s) driven to website.
Bounce rate is the number of visitors who leave site after less than 10 seconds and only visit 1 page in the site.
Conversion rate is total number of conversions (eg items sold) divided by total unique visits to the site.
Message effectiveness chart where each different coloured bubble represents a completely distinct marketing message and design on logo.
CTR (click through rate) is the total number of clicks on the banners divided by total number of impressions on media/publications/other websites where banners were placed.
Cost represents the total ad spend in (000s) with the various publications.

 

 

As usual, I very much welcome your feedback and thoughts and please do let me know which online banner you think is more effective and why and which other charts or data would you would have used?