Archive for Web Analytics

The Balance Between Visits And Conversion Rate

See if this scenario sounds familiar:

This report shows visits are up year over year. Yea! Well done everyone, the markting budget is paying off! Let’s keep the momentum going and see if we can continue to drive even more visits to the site!


This report shows conversion rate is down year over year. Boo! What the heck is going on everyone! Let’s optimize our accounts, cut cost per acquisition in all channels and limit non-qualified visitors entering the site!


The truth lies when you look at conversion rate on top of visits. Obviously, if you’re spending more on reaching out to drive more visitors to the site, chances are they are new visitors and not very low in the conversion funnel, causing a drop in conversion rate. Do you ask someone to marry you on the first date? No of course not, yet that is what too many companies expect when they pay to drive new traffic and expect them to purchase after their first interaction with the site.
What is a “good” conversion rate? Is less than 1% bad? Is greater than 7% amazing? Neither. It all depends on volume of traffic, average order value and margin. If you sell a product for $50 online, would you rather have a site that gets 100,000 daily visits and have a dismal looking conversion rate of .9%, or a site with an amazing conversion rate of 6% but gets 1,000 visits? Option one would be making $45,000 while option two makes only $3,000.
It gets dangerous when an organization has its mind set on a specific conversion rate and makes short term changes to maintain it – like managing their marketing channels so that only the most qualified and interested visitors come to the site. These visitors tend to be repeat customers already far down the conversion funnel – converting the converted. A healthy business will continue to invest in adding more new customers into top of the the bucket. If not, and the tyranny of maintaing conversion rate runs rampant, the amount of people leaking out the bottom of the bucket will exceed the amount going in the top. A target conversion rate should allow for a healthy amount of new visitors who are an investment in the long term. Today’s expensive, non-converting visitors are tomorrows cheap, high-converting visitors.

Leave a Comment

Don’t Write Checks With Your Ads That Your Site Can’t Cash

Landing pages are underestimated in the role they play in paid search. I think the reason is because it’s a lot easier to tweak adtext, bids and keywords than the landing page. For the query “front load stacked washer and dryer” this is the ad and landing page Sears gives me:

Can you imagine walking into a sears and telling the rep that you are looking for a front loading stacked washer and dryer and in response he says, “sure, all washers and dryers are in the back, go ahead and find it yourself.” Yet that is what this landing page is telling me – “here’s everything we sell related to washers and dryers, figure it out.”

You can discover where your ads are writing checks that your landing page can’t cash in Google Analytics. Pull up your keyword report under Advertising > AdWords and add the secondary dimension of Destination URL (or make a custom report like mine below). Sort by bounce rate and here you will see all the keywords that aren’t matching very well with their chosen landing pages.

In my example below, line 2 has spent $335 and has a bounce rate of 90% and -40% ROI – ouch. Time to rethink the quality of this keyword and the quality of this landing page.

Leave a Comment

Analyze Search Terms On Paid Search Landing Pages

Is the keyword you’re bidding on in AdWords a good match for the landing page you have chosen for that keyword? And if not, (after looking at low conversion rates or high bounce rates) what landing page would be better? One way to find out is to see if the visitor does an internal search after landing on that search page.
Navigate to the Content > Site Search > Pages in Google Analytics. Use an advanced segment to show only visitors from your Google AdWords ads. Click on one of the landing pages and then add a secondary dimension of keyword.
In the second column is the paid keyword someone used to get to this landing page. In the first column is the keyword the person used in your internal site search after landing on the page. Essentially visitors clicking on your AdWords ads are telling you with their search term what they want to see after clicking on your ad that you’re not showing them. This can give you insight into changing the landing pages that you have set up with your keywords. Pair keywords that you are bidding on with landing pages that people find after doing a search.

Leave a Comment

Conversion Funnel Web Analytics Measurement Model

Instead of using margin and volume to make your web analytics framework, It may be easier to build it out using a conversion funnel. The different steps of the funnel give you ideas and starting points for brainstorming the functions your site has. It also forces you to think through your marketing and web analytics strategy for each step of the customer acquisition cycle. The steps are awareness, consideration (desire, interest), conversion, loyalty and advocacy.
Each step has Key Performance Indicators and their associated targets and goal values. For some of the goal values you’ll have to determine internally, (Avinash’s post on this is rad) and other goal values are the Per Visit Values calculated by Google Analytics (dividing revenue by visits).

I like this kind of view because it puts into prospective where the majority of your advertising efforts are going. It’s easy to heavy up on everything leading up to the conversion and then forget about loyalty and advocacy, or vice-versa.

Comments (1)

What Does Web Analytics Tell You About How You Are Training Your Customers?

After reading a post by Seth Godin the other day called Train Your Customers, I asked myself, “how could you measure in web analytics the ways in which you are training your customers on your website?” So I took his list and thought up how to measure them using different metrics for each. I think if you can first tie a behavior to a metric, you can start to discover ways to increase or decrease that behavior, once you start measuring it.

  • Be respectful – You got me on this one. Maybe measure the amount of positive vs. negative comments on a blog? (Crap, blowing it on the first one, keep reading they get better I swear)
  • Be patient – In terms of how long it takes the visitor to find what they are looking for: measure Length of Visit under Visitors > Visitor loyalty > Length of Visit. Then use a Visits With Conversions advanced segment to see how patient those people who buy are. Is there a threshold of patience where people give up? How many steps does it take to convert? In terms of page load time use Google’s Page Speed to measure page load times. Another one: bounce rate – this will show how many people are patient enough to read the stuff you put up on your site.
  • Keep their satisfaction to themselves – Just like spread the word below except this one is more subjective. You would need to measure the amount of positive feeback compared to negative. How many opportunities to you give people to share with others?
  • Be selfish – If the site had any kind of donation aspect you could look at the conversion rate of donations. Or selfish in terms of the kind of content that most interests people. Is  product-centric or customer-centric content more popular? Look under Content > Top Content, use advanced filters.
  • Be focused on a superstar – I think measuring the conversion rate of traffic from social media could work for this one. Maybe people are really focused on all the noise you make on Twitter and the amount of visits from those sources turn out to add very little to the bottom line.  Look at Traffic Sources > Referring Sites > flilter for Twitter and then look under the Ecommerce or Goal tab to see conversion rate.
  • Demand personal service – Amount of inquiries to customer service. The amount of browsing between different categories could show that amount of personalization someone would want to help them shop.
  • Be calm – Pageviews per visit? Bounce rate? Depth of visit? Some visitors can be more click-happy thank others.
  • Never settle for the current iteration – kind of like Demand for personal service.
  • Be cheap – Ecommerce > Average Order Value. How small of purchases are people making and how are you aquiring for those kinds of visitors?
  • Embrace acceptance – how much traffic comes form comparison and review sites? Do people need to be reassured by others or their social circle that purchasing from you is the right decision?
  • Spread the word – Amount of clicks on the social “share this” buttons and the amount of inbound links and referring traffic. This can be done in Google Analytics using onclick events.
  • Expect pampering – Goal Abandoned Funnels under the Goals tab. Does the lack of free shipping make someone abandon? What kind of pampering is needed to keep people from abandoning the funnel?
  • Demand free – How many blog posts, ebooks, free consultations and touches with the customer does it take before they buy? Ecommerce > Visits to Purchase.
  • Be eager to switch brands to save a buck – Make a custom segment of Return Visitors and apply it to the transactions report under the Ecommerce tab. If the line is going down returning visitors aren’t coming back to buy. To be able to see the number of purchases from people who previously bought use the User Defined Report.
  • Value and honor long-term loyalty – Visitors > Visitor Loyalty > Loyalty. This will show you how many visits were the visitor’s nth visit. The more visits, the more loyal. Also do this test with traffic that comes from paid channels. How many people who come from banener ads end up coming back? Set up an advanced segment from the Campaign Dimension and then put that over the Loyalty report.
  • Be skeptical – Amount of visits to purchase. If someone visits the site multiple times before they purchase, chances are they are skeptical. This metric is found under the Ecommerce tab > visits to Purchase. What is causing the skepticism? Make a custom segment  with Count of Visits to a Transaction as the dimension and set it to greater than 3 and see what content these people look at that makes them so skeptical.

Leave a Comment

Using Goal Compleation In Advanced Segments

I really like using goals as advanced segments to find out what is working on the site.

Say your Goal #1 is a lead generation goal to get email sign-ups from people who want you to contact them later for more information. Set up an advanced segment for Goal1 Completions and then add it, and only it, to your report.

Now you’re looking at everything that went just right with these visitors: 1. Their expectations of what they were going to see before they clicked matched what they saw. 2. What they saw was engaging. 3. They trusted you enough to give you a chance and then they converted. Here’s some ideas on how to figure out what those 3 things were so that you can make it happen more often:

1. What kept these visitors from bouncing was that what they were expecting, by clicking on the search result or link in a referring site, is what they got. This can be found under Traffic Sources. What keywords and referring sites are driving these visitors who convert? I’m going to try to maximize the traffic from these sources especially to the page that they landed on. Under Content > Top Landing pages I can find out which landing pages worked. Make more content like that with those same keywords. Include those keywords in your SEO efforts.

2. How engaged does someone need to be before they convert? Back to Visitors > Visitor Loyalty > Depth of Visit you can see how many pageviews it it takes on average before someone converts. If the sweet spot is between three and four pages then I can start trying out strategies for getting more pageviews per visit. More up-sells (people who like this also like this) and more links to similar content to keep people on the site.

3. The amount of trust it takes can be found under Visitors > Visitor Loyalty > Loyalty. Here you can see how many visits on average it takes for someone to convert. Once I see how many visits it takes before someone trusts me enough to convert I can set a goal to get those repeat visitors. Creating content more frequently and maximizing the ways people can get alerted to new content (Twitter, RSS, Facebook) can help get repeat visitors.

Leave a Comment

An Easy Example Of How To Use Google Analytics To Improve Your Site

This is one of the first things I look at when I analyze any website. Under Content in the left column click on Top Landing Pages. These are the most frequently landed on pages that your visitors see when they first come to your website. These are your “head” pages, which means that small improvements to these pages can quickly yield high ROI. In the furthest right column in the report is the bounce rate for each of these pages. % Bounce Rate means the % of visitors that came to your site and then left instantly, or in other words, that landing page was not compelling enough for them to read more than that one page. Bounce rate is a great metric to measure the quality of the traffic you are acquiring. It helps you hone in on where and how your website is failing your visitors.

So now you can see what is failing, in one more step you can get an idea of why it is failing: Click on one of those poor performing pages (I put little cash signs next to my contenders) with a high bounce rate to analyze it on it’s own. Then click on the drop down that says Content Detail on it and select Entrance Keywords. Do the keywords people use to get to this page align with the content of the page? Looking at their keywords you can get a sense of what their intent was and why it doesn’t match up with what the page is delivering. Now you have something to work with and fix. Now make the changes to that page so that it better matches what your visitors want. Bounce rate will go down, and positive outcomes will increase.

Leave a Comment

Metrics Determine Methods

Choose wisely the metrics that you use to define success.

Pick a bad metric to measure and you’ll adopt poor methods. For example, if high average time on site or page views/visitor is the metric you choose for gauging the success of your site you may end up with a bunch of visitors who are annoyed with how long it takes them to find anything.

Or if having a huge amount of visits is your metric of choice you may be tempted to use spammy techniques to get traffic; is it better to get a click and then annoy someone, or better to only reach the people who care?

I think a good question to ask is, therefore what? “I want to double the amount of email subscribers to my site,” therefore what? “So that I can get more sales.” Howabout seeing if you can improve your conversion rate from 1% to 2% on those currently in the subscriber list, which also doubles the amount of sales your email subscribers generate, and in the process creates loyal customers that will have a desire to generate word of mouth for you and  improve their lifetime value?

Comments (2)

Why Doesen’t Hulu Show Ratings?

There are reports on how many videos Hulu streams but how come they don’t tell us how many views individual shows are getting? Usually TV shows are quick to point out which ones are getting the highest ratings on TVs even though the way ratings are derived is anything but exact:

Nielsen is making an assumption using a sampling statistic based on 5,000 homes what the approximately 113 million U.S. television-viewing homes are watching.

Yet online, exact amounts of viewership is much more possible. Hulu knows exactly (almost exactly depending on the constraints of their web analytic providers) how many people are watching which shows, how many people drop out and watch shows only half way and they also know the mix of shows people are watching. For example, they would know that a high percentage of people who watch the Simpsons also watch Family Guy, etc.. Sure, Hulu has their “most popular videos” category but they don’t show how many views to substantiate their claim of what is most popular.

You would think they would advertise things like, “Come see the most viewed show on Hulu!” but they don’t, why not? They are hiding something. I bet there is some conflicting data between what the Nielson ratings show and what online shows and they don’t want their advertisers to know about it. And their “most popular videos” category is probably anything but the most popular. I think they cherry pick which clips they want people to watch more of based on which shows demand the highest costing CPMs.

What if Arrested Development is the most popular? But since that show is not airing on TV they don’t want people to like it more, they want people to like The Office more so they can get those people to tune in on Thursdays to sell more advertising. Is their new show Community, which is on the top row for most popular, among the most viewed? Doubtful, I bet they want more people to be exposed to the show since they have a lot riding on it becoming a success. Does Hulu take stocking fees like in supermarkets where networks pay them to put their show on the homepage? Maybe.

For sure they have some good reasons why they don’t reveal which shows get the most views.

Leave a Comment

Super Crunchers

I finished reading the book Super Crunchers by Ian Ayres. I thought it was good. I liked his explanation of randomized a/b or multi variant testing done online and off.

He explains that randomly dividing prospects into two groups and seeing which approach has the highest rate is one of the most powerful super crunching techniques ever devised.

When you rely on historical data, it is much harder to tease out causation. The sample size is key. If we get a large enough sample, we can be pretty sure that the group coming up heads will be statistically identical to the group coming up tails. If we then intervene to treat the heads differently, we can measure the pure effect of the intervention…after randomization makes the two groups identical on every other dimension, we can be confident that any change in the two groups outcome was caused by their different treatment.

Of course, randomization doesn’t mean that those who were treated differently are exactly the same as those who were not treated differently. If we looked at the heights of people in one group, we would see a bell curve of heights.  The point is that we would see the same bell curve of heights for those for those in the other group. Since the distribution of both groups becomes increasingly identical as the sample size increases, then we can attribute any differences in the average group response to the difference in treatment.

In lab experiments, researches create data by carefully controlling for everything to create matched pairs that are identical except for the thing being tested. Outside of the lab, it’s sometimes simply impossible to create pairs that are the same on all peripheral dimensions. Randomization is how businesses can create data without creating perfectly matched distributions.

The power behind randomized testing is undeniable. So should we just have computers make all our decisions for us? With that question in mind is were he goes throughout the majority of the book.

Randomized trials require firms to hypothesize in advance before the test starts. Historical data lets the researcher sit back and decide what to test after the fact. Randomizers need to take more initiative than people who run after the fact regressions.

The most important thing that is left to humans is to use our minds and our intuition to guess at what veriables should and should not be included in the statistical analyisis. The regressions can test whether there is a casual effect and estimate the size of the causal impact, but somebody (some body, some human) needs to specify the test itself.

So then the question becomes what do we test, and after we test the question becomes, what are the results telling us?

Leave a Comment