Videos: Actionable Web Analytics Tips
Through the years Dr. Stephen Turner and "not Dr only Mr." John Marshall have had a deep influence on my evolution as a Web Analytics professional.
My first foray, all those years ago, into a "big boy" web analytics tool was using ClickTracks (after having dabbled in build your own log parsing ecosystem that dumps data in a massive Data Warehouse that fed a business intelligence tool to do web analytics!!). Over the years that early experience with ClickTracks and conversations with Dr. Turner and John helped shape a lot of my thinking.
John Marshall as you all know was the ex-CEO of ClickTracks, and now along with me is the co-founder of MarketMotive Inc which focus on online marketing education and certification. Dr. Stephen Turner wrote what is perhaps still the most widely used logfile web analytics application in the world: Analog (in 1995!). He is also the ex-CTO of ClickTracks.
Dr. Turner was in the Silicon Valley recently and I could not resist having him and John come over to do a quick video.
Two reasons actually.
One they have spent a very long number of years in the field contributing in so many ways, true web analtyics legends to the rest of us. Two, as you'll see soon, we just have so much fun when we hang out.
The question was simple:
"If you could reach deep into your experience and share one awesome actionable tip with Analysts & Marketers what would it be?"
Here are three answers that you'll surely find helpful. . . .
A bit of history + Guru Dr. Turner shares his tip: Segmentation!
(This video is 7 mins long.)
Key lessons you'll learn. . . .
Specific examples of what segments you could start with.
How Segmentation could help you understand visitor persona.
Learn how you can customize landing pages for higher conversions.
Dr. Turner's top secret tip! [Ok, ok you got it out of me: "spend" time with your data!]
There are some links at the bottom of this post if you want to learn more about how to use segmentation. In particular I recommend the post about making your reports "connectable", it is segmentation and also data communication. Let me know what you think of it.
Guru Marshall shares his actionable analytics tip: Internal Site Search!
(This video is 5 mins long.)
Key lessons you'll learn. . . .
How SEO influences what Visitors search on your site.
Why gods gift to humanity (ok Analysts :)) is intent.
Why it makes sense to mix Guru Turner and Guru Marshall's tips!
See with all search data out there the sexy smoldering one in the room was search data you already have? Now that you know that. . . .
My tip: Quantifying Economic Value!
(This video is 7 mins long.)
Key lessons you'll learn. . . .
Why this is the next step in the evolution: HITS -> Page Views -> Visits -> Quantified Goal Values.
How can you charm the pants off your boss/'s using this method.
How do you actually quantify goal values and compute your economic value.
Why it changes every conversation you will ever have about your website.
Ok what do you think? Did you get that? I hope you did, or I'll have to write a long blog post explaining it! : )
All kidding aside I firmly believe that the difference between Analysis Ninjas and Reporting Squirrels is the fact the latter have spent no time quantifying the economic value of their websites (*especially* if your site is a non-ecommerce site, a blog, a lead gen site, a tech support site, a ….).
So did you notice how much fun we had? Web Analytics is such a fun amazing thing.
Now your turn.
Dr. Turner, John and I would love to get your thoughts on our three tips. Please share your feedback using the comment form below.
Oh and if you could give one, just one, solitary piece of advice, what would it be?
Thank you.
PS:
Couple other related posts you might find interesting:
- Excellent Analytics Tip#4: Make Your Analysis/Reports “Connectable”
- Excellent Analytics Tip#2: Segment Absolutely Everything
- Ten Minutes With….. Dr. Stephen Turner
- Kick Butt With Internal Site Search Analytics
- Excellent Analytics Tip #13: Measure Macro AND Micro Conversions
Subscribe 
January 13th, 2009 at 03:10
I feel that a lot of marketers are always obsessed with getting the first purchase/conversion, while ignoring the second purchase and the behavior of a returning user.
In today’s economy, or any economy, I believe a lot of companies would like to get more out of their existing client base. Usually, the plan is to use an automated email system and periodic newsletters with special offers.
My suggestion is to segment first time and returning visitors/buyers. Try to spend more time with your data to learn how do visitors return to you site and work to optimize\create funnels that will yield returning purchases.
Example: If you find out that most of your returning visitors are going to:
Google > Search for “your web site’s name” > Reach your homepage > Search > Bounce/ Leave/etc…
Maybe by creating new sub-domains, that will appear in SERPs below your homepage, you will direct your returning organic traffic to better landing pages.
January 13th, 2009 at 04:39
Very nice post and videos Avinash! Same for your previous one – it inspired me to start experimenting with Goals, assigning values etc. And it also pushed me to finally explore the Segments in Google Analytics – I just didn't spare the time to look at all those features before.
January 13th, 2009 at 07:04
These are great videos on your discussions Avinash. Although this is potentially adding a layer of the unknown into site search, I am a huge advocate of enhancing site search to provide content/product suggestions and predictions based upon what the visitor begins typing.
I know you'll be familiar with this, but for some of your readers the following few benefits might be useful..
For site owners and your visitors suggestive search can:
expose content/products on your site that the visitor may not have known existed
providing your visitors with a quicker and more streamlined search facility, enhancing their user experience with your brand
reduce the number of 'no results found' pages
Dynamic search suggestions can also bring back information on sale values and other persuasive trigger content, therefore increasing the likelihood of a visitor converting.
In reference back to your video on internal site search and intent, this type of enhanced site search can affect a visitors intent as they may choose to view one the suggestions that has been presented, rather than what they were intending to search for.
I'd be interested to hear thoughts on how analytic experts report on sites providing suggestive/predictive search functionality. In fact I wonder how Google report on their suggestive search toolbar – Avinash, any insights?!
January 13th, 2009 at 07:55
These videos are great and help to drive home a lot of important points.
The one piece of advice that I'd give is to take the guesswork out of quantifying the value of your goals. Things get a lot more complicated when trying to determine lifetime values of people who do things like pay subscriptions, buy repeatedly, or go through a trial period. This is another case where you can look at the average, but segmentation helps a lot.
One of the top priorities of my business intelligence software company right now (shameless plug – http://www.rjmetrics.com) is to integrate with the new Google Analytics API so that we can compute segmented lifetime values for goals in GA with a minimum amount of work for customers. There are, of course, lots of other good BI tools out there as well. Additionally, depending on the size of your business and your data analysis skills, you often can get a good snapshot-in-time approximation by dumping everything into excel and working away for a few hours.
January 13th, 2009 at 08:41
Hi Avinash,
Thank you so much for these awesome videos. It is really nice to see experts break down seemingly complex concepts to simple terms.
My single 2-cent piece also has to do with Site Search.. but with a twist: Null Searches. In addition to knowing my site visitors' intent, knowing which searches came up with empty results helps us identify (or reject) business opportunities.
Another piece – not original, but worth mentioning nonetheless: Benchmark + Segment + Trend + Test = BESTEST for short :)
January 13th, 2009 at 09:26
Hi Avinash – thanks for the great videos. You guys need to start analytics.tv!
Anyway, my question is in the first video, I love the segment of comparing visitor sources to days until purchase. Where in GA is it pulling days until purchase from? Is it as simple as turning on the e-commerce functionality?
January 13th, 2009 at 10:50
Great video! Coincidentally, I just posted an article on my blog about visitor segmentation by motivation or intent on my blog as part of a Model to Improve Traffic Conversion. I'll definitely link to this post!
January 13th, 2009 at 10:51
This is a great post! Thanks for the time and effort you put into this.
January 13th, 2009 at 10:58
[...]
While the business or site goal remains the same, the visitor’s intention differs. So does it make sense to group the experience of those looking for career opportunites with those visiting your site to make a purchase? By identifying intent, you can segment users that visit with the explicit intention of purchasing and evaluate their user experience through the purchase flow. With this information you can more accurately make recommendations to improve the site’s usability for those visitors coming to make a purchase.
(Occam’s Razor by Avinash Kaushik has a great video of his discussion with the founders of ClickTracks on the topic of segmentation at http://www.kaushik.net/avinash/2009/01/actionable-web-analytics-tips.html.)
[...]
January 13th, 2009 at 11:34
Avinash, great post, I am some what pleased to hear you have encountered the same issue where decision makers are not using analytics to base their decisions on. And this post has given me more ideas to help convince the powers that be to listen and learn from there audience. Thanks for another great topic.
Kevin
January 13th, 2009 at 11:39
This is like watching "Big Bang Theory", but without the cute girls.
January 13th, 2009 at 13:03
[...]
Videos: Actionable Web Analytics Tips | Occam's Razor by Avinash Kaushik
Just watch the videos.
I love the 3rd tip (economic value) – something we often talk about in the AdWords Seminars (If you don't have goals – how can your marketing or analytics mean anything at all?)
[...]
January 13th, 2009 at 14:18
Thanks for sharing Avinash,
I'd be curious to know what your analytic results on these videos are after a couple weeks?
What would you see as quantified goal values for video?
January 13th, 2009 at 15:49
Avinash in short blog post shock! :) seriously this is great, excellent tips (and nice plug for market motive – reminds me I must sign up to again!)
cheers
Jon
January 13th, 2009 at 16:45
You guys are having way too much fun :-). Great videos and I like the open-ended variations you left the reader with using these three tips (e.g segmentation x latency/time-between-purchase etc.)
On Segmentation, the only thing I would add is one need not wait to have globs of data to start segmenting and studying insights. Assuming the sampling and other conditions are taken care of (e.g seasonality or cyclicality of customers in your business), you can start the segmentation approach to dive into the data instead of waiting for a site-wide, x-months of data dump
January 13th, 2009 at 16:58
Tal: As a fellow fan of "pan session" metrics I could not agree more with you. I am a huge fan of metrics like Visitor Loyalty, Visitor Recency, Days to Purchase and Visits to Purchase etc. And if you segment those metrics you might simply go crazy with excitement (and insights!). :)
Paul: With regards to clicks on suggested keywords GA will not report on it by default. There are couple ways you could go with this (totally off the seat off my pants – actually 36,000 ft in the sky on a flight to NYC! :)).
If your website can take that suggested keyword click and stuff it into a different parameter then GA can totally track that. For example on my blog S is the parameter where the keyword is stuffed in the URL, and I could configure my site to stuff clicks on suggested keywords in the Q parameter and tracks clicks and customer behavior using GA.
Second I suspect that you could capture that data with Event Tracking in Google Analytics, it was built for you to be able to collect stuff like this (and of course mouse hovers and fly out menus and ajax this and that etc etc etc).
Finally one of the simplest ways to enhance your site search experience it to provide "best bets" on top of the search results. I search for turbotax on intuit.com and on top of the search results pages I get two "paid search" search links that show pages that the site owner wants me to see, say a link to turbotax premier and another to the turbotax youtube site (or whatever).
The searcher gets what they want and the company is able to expose other things ins a very soft and subtle way. Win-win.
Ryan: Days to "Purchase" is a standard google analytics report (I am sure you'll find it in Omniture, WebTrends, CoreMetrics etc as well). You simple turn "on" ecommerce functionality (and by that I mean you identify your site as ecommerce in settings and implement the ecommerce tag on your outcome page).
That brings me to a important point, with GA you can track any outcome, not just Purchase with ecommerce. So your outcome could be a lead, it could be sign up for something or even a download. Any outcome.
Once you turn it on, in you'll see a ecommerce tab in your reports and if you open that you'll see the Days to Purchase and Visits to Purchase reports.
Jake: Good point. But there is no rule that says that you are stuck with the values you got. So for example I can monitor for a few weeks, figure out immediate conversion value is $10 and go with that. But I can revisit say once a quarter and then revisit that value. Maybe it turns out 50% of the acquired customers quit so now for future conversions I can bake in a 50% churn rate and only use the value of $5.
Just a small example of how you can do a lot with something even simple. Of course if you have a dynamic system then you can get a lot more complicated and sophisticated. And that is kosher as well. :)
Ric: Your comment has made the inner "nerd" / "dork" / "dweeb" / "insert the right American term here" every happy!! I don't get to watch TV but I have seen episodes of The Big Bang Theory on my frequent United flights and it is a great pop corn show. So much I/you can relate to or atleast empathize with! :)
Thank you for making my/our day!
Hershel: Ahhh…. The question you should never ask a blogger with fiver jobs and not enough time to instrument things right for measurement! : )
Ok talking seriously for a minute….
The video itself is not instrumented, I wish I had time to do that. I rely on the comments to reflect value of the videos and what worked. I also benchmark my video posts (four so far) for page views and comments. Tells me if there is right amount of consumption and connection.
The blog is does not have a monetization strategy but if I were to state a goal for this post, along with teaching you something, it would be a "branding" post for Market Motive, my company.
There is just one subtle link to Market Motive in the post. But you'll notice it has a tracking code on it, to track referrals to MM and also, if, any conversions to trails or full signups.
-Avinash.
January 13th, 2009 at 21:33
Thanks for sharing the videos and some pointers!
January 14th, 2009 at 01:25
Thanks for your feedback Avinash. I think your point on enhancing what the business and the searcher gets out of using internal search is fundamental to both the companies bottom line and the visitors experience on your site.
When companies make intelligent enhancements to their sites then it will be hard to find visitors who don't respond positively, and it can be what are perceived as small tweaks to a user experience that help improve a companies credibility and returnability factor – and therefore more analytics data to work with!
January 14th, 2009 at 07:54
Avinash,
Thank you for the videos and for mixing things up with the format. Interesting chemistry between you three. Hmmm..what would happen if we took 3 web analytics gurus and had them live in the same house? I think I sense a new reality show coming on!
I enjoyed and learned from all of the videos. The comments on segmentation and personas has me thinking more about the relationship between segmentation, personas, and maybe the campaigns, channels, or traffic source the users came in under. I've not had the benefit of a formal marketing education. I'm interested in learning more about how you go about creating an online persona and if there is a difference between that and a typical buyer persona. Your site is not an ecommerce site. You have users that come in because of RSS feeds. You have users that find you via a google search. You have users that have you bookmarked. You get traffic from blogs. What would the personas for your site look like and how much influence does the traffic source that they came in under play into this? I did some looking around on the internet and it seemed like many of the personas were directed at offline marketing activities. How do online marketing activities and personas differ?
January 14th, 2009 at 10:46
Avinash,
Again more great content from MarketMotive.
If I had to give one tip it would be: People need to listen to, and do their best to understand advice such as this. It still amazes me the money people will spend on an analytics tool, yet the lack of effort in learning how to get the most out of it. Analytics apps are exactly that: tools, not a collection of finished products.
I picture construction worker staring at a hammer sitting on the ground and thinking to himself "Why isn't that fence build yet?! Isn't that what it does?!"
-Kevin
January 14th, 2009 at 10:48
Above: listed = listen.
I suppose people also need to learn how to proofread. :)
January 14th, 2009 at 23:40
Great post. The videos touched on this a bit but I am very interested to know what the best way is to view the referral source of specific keywords in Google Analytics?
I see the list of keywords but it doesn't seem to show me (likely it does but I'm missing it) how to find out where that person came from for each specific keyword.
If I click on a keyword, it only shows the details of how many visitors used that keyword, bounce rate, time on site, etc. What is the best way to set this up so that I can see the specific location a visitor is coming from and which specific keyword they used to find my website?
Thanks Avinash! The reply would be appreciated.
Daniel
January 15th, 2009 at 05:17
Avinash Ji,
Mujhe aisa kyun lagaa ki aap hansna nahin chahte the par phir bhi zabardasti hans rahe the :- )
Vivek Kedia
अविनाश जी ,
मुझे ऐसा क्यूँ लगा की आप हँसना नहीं चाहतऐ थे पर फिर भी ज़बरदस्ती हंस रहे थे :- )
विवेक केडिया
January 15th, 2009 at 14:45
Fantastic tips! Thanks for these videos!
Was nice the approach on SEO (love it!) that “not Dr only Mr.” John Marshall tip brought.
And well, the Quantifying Economic Value is something fundamental but not that easy to make with lead generation and content sites (calm down, it's not easy for me :)) as on ecommerce ones.
Honestly if I would pick one, that would be the Segmentations. Of course Segmentations would be different for each type of site, once we got to work with diferent KPIs.
Please keep making helpful posts!
January 15th, 2009 at 22:51
Alice Cooper's Stalker: There are many different kinds of "personas" that you can create for your website.
For example you can pop open your web analytics tool and create these kinds of personas: people who come from search, people who come on brand terms, existing customers (who have purchased before), direct traffic, people who only come on offers.
Each of these groups have either a common intent or need or behavior etc etc. That helps you understand them better (even if anonymously).
You can also create personas using survey data. For example if you go to intel.com you get an exhaustive (and I mean exhaustive) survey that asks you for lots of things, are you a student, are you a var, are you a existing customer, are you looking for laptops, are you male or female, how old are you, what's your underwear size (ok just kidding with that one). You can see how that data can easily be used to construct personas of who comes to your site and then execute against what all these people want.
Or of course you can use 4Q and create persona's for various primary purpose visitors (then you have even stronger understanding of intent).
Finally there are dedicate companies that will help you create actual personas of your visitors, of course it costs anywhere from ten to a hundred grand US. But you'll get representative habits, pictures, deep (if a bit aggregated) personal information and so on and so forth.
Now here's the important point: You can't but you don't have to start with spending money, and you don't have to wait for months to start understanding your visitors. You can start tomorrow, you can get sophisticated over time.
Daniel: There are two ways to do this.
1) In the keyword report in the Dimension drop down choose Source, it shows the referrers. Like so. . .
2) Better: I just don't use standard GA reports any more. Don't care for them. :) I simply create a custom report with whatever I want. Like so. . .
I choose the metrics I want and the drill downs I want and the world is good again!
Diogenes: I think with leads you should be able to still quantify goal values. You might have to wait until some leads from your online channel convert. Measure conversion. Compute value of each conversion (offline if you have to). Go back and apply that value in your web analytics tool. "Simple."
Ok maybe not simple, but not as much work as you might think.
Good luck!
Avinash.
January 16th, 2009 at 07:59
[...] Web analytics is a bit like accounting, even if you’re not the web analyst you need to have a basic understanding of what it is and why you need analytics. Avinash Kaushik posted 3 videos from internet marketing gurus on Segmentation, Site Search and Quantifying Economic Value that you should check out even if you’re not the full time web analyst. [...]
January 16th, 2009 at 08:26
[...]
Occam’s Razor – Interessante Videos zum nützlichen Umgang mit Google Analytics:
Vor allem das zur Segmentation und Persona-Bildung finde ich interessant.
[...]
January 16th, 2009 at 08:52
[...] Update: Videos: Actionable Web Analytics Tips – via Occam’s Razor by Avinash Kaushik, see there 3 videos about actionable web analytics [...]
January 16th, 2009 at 10:25
Nice post. I'm fighting with some of my friends to try to show their clients this kind of things. Some of them are using Goal Value as a measure to find the value of the page trough that Index, that GA is calculating. So wrong.. I post about that on my blog trying to teach them. Right now on my post I put this link maybe they will listen better your voice, not mine.. Thanks again for this 3 videos, these are educative tools. Thx.
May you consider visiting someday Romania?
January 16th, 2009 at 13:48
I have really been interested in segmenting traffic lately to help avoid the shortfalls of average user information. I really liked that portion of the video. It ties in with a previous post you had on measuring offline marketing activities online.
I just starting using some "vanity URLs" for radio campaigns with redirect tracking links to segment out that traffic and it was great to see how easy it was and how well it worked. I know that info is old news to many people but it was the first time I was able to use it and see the results.
I also enjoyed the portion on personas and am trying to figure out the best way to segment data out by persona. I am taking courses from UBC on web analytics and they get into some great detail about persona marketing but I have not yet put it to use on our site.
Any suggestions of good ways to segment out personas? Here are some of ours: drive market N. CA visitors, destination visitors (outside CA/NV), PPC visitors, natural search visitors, sale purchasers from email blasts, information gatherers, package buyers.
January 17th, 2009 at 11:33
Avinash,
Thanks for the reply. That explanation and detail is just what I was after. Off to create a new 'custom report'.
Much appreciated!
Daniel
January 17th, 2009 at 11:58
Avinash, setting up that custom report worked beautifully! Thanks again.
Do you know:
If someone is searching for Ripe Tomatoes and our site appears on the second page of the Google listings for that keyword is there any way that we can see that they came from page 2 of Google for that keyword (in Google Analytics)?
Adwords makes this easy but I have yet to find a way to do this with non-paid keywords.
Cheers,
Daniel
January 19th, 2009 at 03:23
WOW!
Another great post from Avinash!
The videos were great!
And always insightful!
And with all this info I am struggling to create the "perfect" report!
And by struggling, I mean where do I begin? Ha ha…
But thanks again!
January 19th, 2009 at 04:02
[...]
Avinash Kaushik wrote a great post last week about “Actionable Web Analytics Tips“, where he talks about Segmentation, Site Search and Quantifying Economic Value. Many Online Marketers knows about the importance of segmentation but usually don’t find the time to define their web analytics tool to help them track and understand their users’ behavior.
[...]
January 20th, 2009 at 06:29
Really insightful clips. It's seems I have a lot to learn about actionable web analytics.
January 21st, 2009 at 15:39
What a breath of fresh air!
I've been doing PPC work for a hotels.com affiliate for nearly 8 years and am amazed just in these three short videos how much of a vision one can have beyond simple conversion rates! That's all anyone ever asks me about around here.
In the past I've tried incorporating versions of this concept of 'segmenting' to try and distinguish visitors, rather than the one size fits all approach. Not all visitors are equal!
Thank you Avinash for sharing your thoughts and creating such a wellspring via this blog! I have subscribed to your rss feed and plan to go through all the posts I can!
Looking forward to all comments as well as I come to see what a vibrant community this is!
All the best,
Tim
January 21st, 2009 at 21:48
[...] I forgot to mention this previously, but wanted to point out that even for someone like myself who is very into web analytics and is always thinking about how to use analytics for my clients, that even I learned a few tips (or inspired some as well) from the videos on Avinash's post about some actionable web analytics video tips. [...]
January 23rd, 2009 at 17:18
Awesome! Want more! :D
Thanks, again. I will be putting plenty of these tips into action.
January 28th, 2009 at 02:39
Hi Avinash
That video was a wonderful piece of information.
It has really helped me think in different angles!! And the tip that you gave on Goals was really awesome.
I run an information website, so I have not created any goals, now I have really thought about configuring goals. Thank you!!
Regs
Jawahar
January 30th, 2009 at 08:44
On my link to my name above I wrote a blog post about what I learned here on this blog and how I found a disconnect that my visitors were having. Thanks for the videos!
Dave
January 31st, 2009 at 03:25
Hi Avinash,
I have my own way of tracking comments in a blog as goals but i was just curious on how you do it. I hope you don't mind if i peak at your Source Code after submitting this comment. :)
Claudiu