Q: We’re huge fans of your blog, but then, we’re kinda nerdy gals.  Who is your audience, and what is the purpose of your blog as you see it?

Avinash: Are you calling me a nerd? Or maybe a nerd magnet?  I am quite ok with that (note to self: go to amazon to place an order for pocket protectors).

When I started the blog in my mind the audience was for the most part practitioners of “web analytics” and the senior management of websites / companies.

The web is an awesome living breathing organism and it sucks that analytics is still stuck in the stone ages. I am passionate about leveraging the web to get really close to our customers, to merge the worlds of qualitative and quantitative data, and to make it the most responsive lowest cost channel for anything that a company can imagine (ecommerce, lead generation, support, advocacy etc). My hope through the blog is to share that passion with more people and to broaden minds about what “web analytics” really is and how to leverage the web as a learning platform.

I am finding that in reality the audience is very diverse. Vendors, consultants, people interested in investing, product managers and a whole host of web marketing people. I am even more surprised at how much international audience there is for the blog, around 30%.

Q: We know, and you know, that analytics is fun and addictive, But so many website owners don’t actually look at their traffic analysis data – there’s something overwhelming about taking that first step.  What’s the best way to get your feet wet in analytics if you’re not an experienced marketer?

Avinash: My standard quote to a marketer is: “Without exception the web can help you earn a higher bonus.” There are two components of a bonus, usually, either you bring in more (acquisition) or you convert more (revenue). If you are a marketer start with either of those.

If you own the site start with the simplest possible report, Visitors by Referring URLs. If you are running campaigns then start by identifying which campaigns are bringing in the traffic. Really simple. Where is the traffic coming from?

Now marry that up with outcomes. Revenue by referring URLs or campaign codes etc.

In 10 minutes you now know where people are coming from and what and how much they are buying. Now look for surprises. I am positive that even with this simple data you’ll be surprised at what you find. Marry up the data with your spending. So you poured $500k into search campaigns and you got less traffic and revenue than you hoped for (or less than what you are getting from blogs). You are now in business.

Notice I did not say anything about page views or path analysis or average time on site or all those traditional metrics. IMHO those tend to not get marketers anywhere because like the rest of us those metrics really don’t “connect” and actions are hard to figure out.

Q: The pay-per-click ad services make it easy to measure the success of a paid campaign, and I think that’s partly why PPC has enjoyed such rapid growth.  But in trying to measure the success of organic SEO efforts, there’s still a lot of guesswork, expense, and a learning curve involved. Do you think Google Analytics will bring the same level of accountability to organic SEO that PPC efforts currently have?

Avinash: As an aside could I just confide that most people don’t measure PPC well. It is easier to show that you spend $10 and made $11. But most people don’t measure that $9 of that you would have made any ways just with SEO or that just the budget spent on PPC was $10 not including your salary and your agency fees and all that. True measurement of PPC effectiveness has yet to arrive at the scene, but it will soon as more accountability is demanded.

An important disclaimer first: I have absolutely no knowledge of any sort about Google Analytics’s plans or road map. {editors Note: This interview was in 2006. Things have changed since then! -gc} Any thing I say is pure speculation on my part.

Google has brought a lot of transparency and data APIs and standards, etc. to the world of PPC simply because of the amount of money spent there and the pressure on Google’s customers to show clean ROI. I think that kind of pressure will be on Google in the near future as folks increasingly spend more money on SEO and realize three months later nothing much came of it.

SEO measurement is much more complex. The success metrics are “outside” the site: page rank, page strength, results for company key phrases, your landing page quality score etc. Success metrics for PPC are in the site: conversion, purchases etc.

Google Analytics (GA) is a great way to measure website metrics (so PPC), as would be omniture or webtrends or clicktracks. Google has provided API’s now into adwords that any vendor can plug into.

For SEO the challenge that the data is outside your site is a tough one. Most of the data is with the search engines. I think Google, and others, will evolve to give us automated feeds of our key SEO metrics in order to bring more accountability to the SEO business. If they do that they might first roll it out via GA, but given the dynamics of the marketplace every web analytics vendor will have access to the data.

Q: We know you work for a large business (Intuit).  We’re curious if you think there are any analytics strategies that cross over into small businesses.  Do you have any web analytics advice that you can give to those mom-n-pop businesses out there that are hoping to improve the performance of their websites?

Avinash: The strategies or tips or advice that I mention on my blog or in my speeches might be valuable for any size of business. The strategies and approaches stay the same usually, it is just the scale of them changes. So for example a frequent advice is absolutely positively immerse yourself in segmentation. This would apply in either case. The 10/90 rule applies specially well for the small business, don’t put a lot of money into a tool to get very good world class data.

I did a very special post just for small businesses. It has a lot of tips and specific reports a small business can start with and then become more complex with time.

If I had to summarize it in a few words: Get a free tool like Google Analytics or Clicktracks, start measuring search engine traffic and referrers, look at top pages on the site that lead to conversion, measure site bounce rate (especially if you do PPC), do SEO and use the site overlay report. Viola! You are a million dollars richer!!

Q: It seems that everyone is looking for a “silver bullet:” one measurement they can look at that will indicate the overall success of their website. We’ve heard a lot of people talking about bounce rate as being this silver bullet.  Can you define bounce rate for our readers, and do you agree that this an important web metric?

Avinash: Ahh great timing, I just mentioned it above. Let us get one thing out of the way: There is no silver bullet. No matter if I say that or your grandma says that. It is a lie. The web and business on the web is simply too complex (and that is why I love it).

There are many definitions of site bounce rate but the one I have found to be most insightful is:

In % terms Site Bounce Rate = (Site visitors who stay on the site for less than 10 seconds) divided by (Total number of site visitors).

Site bounce rate is a great entry level metric. It is fantastic at identifying all sorts of, shall we say, “bummers”. If you compute it, as defined above, it is a great way to know how much site traffic is engaging with the site as a whole.

Where it becomes fun, remember I am a geek, is when you segment it out. So find out what you are spending money on and then compute site bounce rate for that. If you spending a lot on PPC compute site bounce rate for each campaign, very quickly you will find out which campaigns are sending you wrong traffic. Or you are a small business and you have paid xyz search engine $500 to list you in their directory, compute bounce rate for that traffic, see if that money was worth it.

So this is a great metric to start with. It tells you a lot and identifies some “bummers” very quickly. But then you have to get deeper into understanding the why it is happening and what you can do and how to fix it etc and you’ll graduate to other more complex metrics. By then Site Bounce Rate would have more than paid you back for the investment you made in computing it.

Q: Many business owners have goals that they want their website visitors to complete offline – things like recommending their business to other people, making a phone call, or walking into a store.  Have you worked on measuring these kinds of goals, and if so, what advice can you give to these folks?

Avinash: Multi-channel is a big challenge and everyone is making progress, not as much as one would like. Without speaking about my employer I’ll give out some general tips.

Phone is the easiest one. Simply use a unique 800 number on your website. Getting a ton of phone numbers is a very small amount of cost and it is a awesome way to know which calls are coming from the web. Now there are companies that will run your campaigns (PPC or radio or otherwise) that will put in a dynamic 800 number on your website and route the call from their “switch” to your company and in the process capturing the online to offline call. If you are small you can just do the former, use a unique number and then count.

Store is much harder.

If you own the retail store do online coupons that people can being to the store and you know they found you on the web. And it does not have to be a lot of money, I have seen a promotions for just one single pack of post-it notes if I brought this online printout thingy! Or I am sure you have seen many stores like Best Buy or Circuit City partner with BizRate. You’ll see a small survey code on the receipt in exchange for a $25 raffle (talk about cheap!) you go online and fill out a survey and tell them where you did research (online).

If you don’t own a store and you sell online and via stores this is exponentially harder, because of the “missing link” between you, website, and the selling channel, retail store. In this case user market research studies or online website surveys (that ask for purchase channel preference) have been used effectively to gauge the impact of the online channel on the offline channel.

Q: In one of Gradiva’s favorite posts on your blog, you talk about giving names to certain types of  website visitors (the abandoner, the flirt, and so on).  Why is it helpful to think in terms of “personas” instead of stats?  Do you have any other “nicknames” like these that you use in your analytics work?

Avinash: We often forget that human beings visit our websites and not “shopper_id’s” or “visitors” or “cookie values”.  The post was rooted in my hope to bring the human back into our minds. Humans interact with our website, just kooky ones like me and smart ones like you. And we are not the same.

I could write all day about personas but the main reason I love them is because they help us all step outside the “sanitized” world of numbers and think of our customers as people, atleast groups of people. Personas bring reality to our minds and then when you think of your website or analyze the numbers you’ll do it very differently. You’ll be solving for “Susan Simple” or “Tony Advanced” or “Avinash Nerd” etc.

I find that this specially works wonders with non-analysts, our management or marketers etc.

Q: How are you enjoying being a blogger?  Is it more work than you expected?  And do you blog “on the clock” as a part of your work efforts, or is this more of a hobby?

Avinash: Let me take that in the reverse order……

The blog is not a part of my work effort, it is a personal blog. I don’t want to call it a hobby (if it were I might have given it up after a week ). This is a bit corny: the blog is very much a labor of love.

I don’t have books or consulting services to sell. I am following the path of my “virtual mentor” Guy Kawasaki (he does not know I exist but I take a great deal of inspiration from his blog). His advice is “Eat like a bird, and poop like an elephant.” I am trying to live that Japanese quote. 

Blogging is much more work than I expected. I only post twice a week (my posts are usually quite long) and yet my wife Jennie’s calculation is that I am putting around 15 hours into it each week. That would include thinking and drafting and writing posts and commenting and replying to email and all that. 15 hours a week is not what I expected when I started (and it is a lot with a full time job and two beautiful little children and a wife and family and travel).

I am enjoying it very much I have to admit. I am humbled by the kind words people say on the blog and I am amazed at the reach of my little blog (I am surprised and thrilled to be ranked around 12k in Technorati) and I have met so many wonderful people (you for example) that I would never have met before. I think I enjoy it most of all because is a outlet for something I am deeply passionate about and I like to think I am adding some value in our little ecosystem.

I wanted to thank you both so much for the opportunity to do this interview, I had a lot of fun.

The pleasure was all ours!!